The Pitch That Changed Everything

Chapter 1: A Tight Runway and a Taller Promise

The ClearRoute office was two rooms above a bakery, the smell of fresh bread drifting through the open window as Maya Ellis wrote numbers on the whiteboard. “£62,000,” she said aloud. “Monthly burn eight and a half. That’s seven months.”

Leo Martin sat on the radiator, laptop balanced on his knees. He kept his hands folded, the curl of a man who coded and fixed things rather than sold them. “Seven months if nothing changes. We need customers or a better burn plan,” he replied. His voice had that careful tone of someone who expected the code to be the truth. “And don’t forget product limitations. The MVP doesn’t do automated reorders or full delivery routing algorithms. It gives alerts and suggested routes.”

Priya Nair, who did marketing, flipped through a stack of printed screenshots. “People want a story,” she said. “They want confidence. We can show how shops can reduce missed deliveries and waste. We’ll use George’s numbers as proof—real evidence.”

They had all heard George’s story. He owned a small grocer in Bedminster, three employees, a fridge full of fresh food and a gentle worry that more bread would end up wasted because deliveries were missed or stock management was sloppy. He had been the kind of customer that turned their idea into a product; a good problem, a clear need.

Maya looked at them and smiled in the way of someone who could already see the headlines. At twenty-eight she loved the moment before everything aligned. “We can promise clear cuts in missed deliveries, reduced waste, and a smooth pilot in two weeks. We’ll say: ClearRoute automates smart routing and stock alerts so your shop misses fewer orders and throws away less.”

Leo’s hand rose. “We need to be precise. The system gives route suggestions and alerts when stock is low based on sales patterns. It can’t automatically place orders with suppliers, and it doesn’t connect with all delivery partners yet. If we say ‘automates’ we’ll be misleading.”

Maya’s eyes flicked to the whiteboard. “Investors like certainty. Customers like certainty. We have to sell them the future.”

Priya put the screenshots down. “We can be proud and bold,” she said. “But honest. The message can be simple: fewer missed deliveries, less waste, and affordable pricing. We’ve landed the pricing at £45 a month per shop for the basic plan. That’s competitive, and easy to explain.”

The ownership split was on the bottom corner of Maya’s notes—an uncomfortable but necessary fact. Maya fifty percent, Leo thirty, Priya twenty. It sat alongside firm figures: twelve pilot shops signed up, twelve times forty-five pounds a month, bringing in £540 monthly recurring revenue. Small, but real.

That evening Maya walked to George’s shop on her own. She liked the directness of meetings where you could see the bread on the shelf and count the boxes in the storeroom. George was honest and gruff in the way of someone who had run a shop for thirty years. He needed a solution that would stop him calling suppliers three times a week and losing sales to missed items.

Around midnight, the team left the office. Maya felt the push of momentum. They had seven months, George’s pilot, and an invitation pending from investor Adrian Cole in six weeks. If they could convert pilots to paying shops, they could show traction. If they could show growth and tighten the burn, maybe the runway would stretch. She did not know yet how a single sentence in a sales call would ripple through January and into the investor meeting ahead.

Chapter 2: The Call That Pulled the Thread

The next morning Maya woke to three missed calls and a voicemail from George. Her stomach tightened as she listened. “Maya, we had a missed delivery yesterday,” George said, voice brittle. “I was sure you said it’d handle the reorders and the driver routes. My staff were confused and the tomatoes went off. What now?”

She sat on the edge of her bed and felt the familiar two-pulse thud of panic and resolve. She called him back. “George, I’m really sorry. I over-explained earlier. ClearRoute gives strong alerts and suggested routes, but we don’t yet automate ordering with suppliers. We should have made that clearer.”

There was silence. Then George spoke, quieter: “Honesty is better than promises. But money and food don’t wait for clarity. We can’t have people missing deliveries and me paying for produce I don’t sell.”

When Maya hung up, she told Leo and Priya immediately. Leo’s face grew long. “This is why the language matters,” he said. “We should be able to say ‘we’re building towards automation’ rather than ‘we automate’. George’s trust is part of our evidence. Without him, the case to investors is weaker.”

Priya, who had drafted the pitch slides, ran a hand over her chin. “We can fix this publicly and ethically,” she said. “Say sorry, say exactly what we do, and outline a short roadmap for the feature people expect.”

They called George and offered a short-term workaround: a tighter schedule for alerts, manual confirmation steps, and a weekly check-in to make sure deliveries matched orders. They also offered a fee credit for the month he had lost fresh produce. He accepted the credit and the check-ins, but his earlier warmth had cooled.

Helen Shaw, their mentor, came by a few days later. Helen had sold a business and now advised founders with a no-nonsense approach. She listened to the recording of Maya’s earlier call and to the team’s discussion of finances. At the end she tapped her pen on a notepad. “This startup life is always a negotiation between aspiration and reality,” she said. “You need both. Maya, the instinct to promise comes from wanting to seal the deal. But the truth is a better foundation for long-term customers and investors. Rebuild trust with George by showing him exactly what you will do and when.”

Helen asked about the numbers. The team repeated them: £62,000 in the bank, a burn of £8,500 a month, runway seven months. “And your pricing?” she asked. “£45 per month per shop basic; £120 for multi-branch premium,” Priya answered. “Twelve pilot shops signed up, but only eight are active,” Leo added. Active meant they used the dashboard frequently and responded to alerts. The revenue remained small—£360 monthly recurring revenue from active shops, an alarming reality when burn was £8,500.

Helen folded her hands. “That’s not unusual at this stage,” she said. “But you must treat each customer like an investor. George is not merely a user; he is proof of product-market fit. Damage to that relationship costs you credibility—especially when you have an investor meeting in six weeks. Fix it, and document every step.”

That night, the founders sat with Helen’s advice and the ledger open. Maya felt the weight of a single sentence made in eagerness. The team could recover, but only by aligning their messages with the product they actually had and the honest roadmap they could deliver.

Chapter 3: Roadmaps, Risks, and a Rigged Demo

The week that followed became a blur of technical fixes and message edits. Leo rewrote parts of the dashboard so alerts were clearer and the manual reorder button could be emphasized in the UX. He added logs so the team could show proof of the system suggesting actions. Priya reshaped the pitch deck, replacing the phrase ‘full automation’ with ‘automated suggestions with manual confirmation’, and added George’s week-over-week data as a case study.

Maya pushed for bold visuals. “If we can show a 30 percent reduction in missed deliveries in the pilot, Adrian will be impressed,” she said. She wanted graphs, headlines, and a short customer testimonial video. Priya smiled; she liked the headline. Leo frowned. “We must not rig the demo,” he said. “We should show real numbers, not a best-case scenario.”

They argued like colleagues who had gone too long without sleep. Maya believed boldness would win the investor and then free them to build. Leo believed honesty and technical certainty would carry them through both customer relationships and due diligence. Priya tried to be the mediator, restraining Maya’s shine with facts and helping Leo find a way to present the technical limitations without sounding incapable.

In the middle of the week George’s shop reported improvement: one fewer missed delivery, better ordering concentrations, and staff who found the suggested routes useful. The system had not automated the orders, but the manual checks cut mistakes. When Maya saw the numbers—ten percent reduction in missed deliveries in that first week—she felt vindicated. Ten percent was better than nothing, but she still wanted more for the slides.

They mapped the roadmap on sticky notes. Phase one: stabilize alerts and improve UX by end of month. Phase two: integrate with two regional suppliers to support semi-automated orders in three months. Phase three: begin full delivery routing algorithms once they proved the data quality. The roadmap was honest but ambitious.

Adrian Cole’s invitation sat in the calendar like a blinking light: investor meeting in six weeks. Adrian was a seasoned angel and early-stage investor in Bristol. He had invested in logistics before and had a reputation for blunt, practical questions. The team knew he would test both their numbers and their integrity.

They rehearsed the pitch twice a day. Maya polished the opening: a crisp problem statement, a human story—George—and a slide that compared food waste margins to the cost of their subscription. Priya worked on customer evidence slides and a clear pricing table—£45 per shop for basic, £120 for multi-branch—along with CAC estimates. Leo prepared to answer technical questions and to show product limitations honestly: no automatic reorders yet, no full driver integration, but strong signals from sales patterns.

On Thursday Leo found a bug in the route suggestion algorithm. It occurred in rare cases when the sales data from a shop was sparse, leading the system to suggest suboptimal delivery sequences. He reported it; the fix would take three days. Maya felt the sting of urgency. “We can scope the demo to shops without sparse data,” she said. “We’ll present the best performing case.”

Leo pushed back. “If Adrian asks about data sparsity, we must answer. If he asks why we selected the shops in the demo, we must be ready.”

Maya sat late at the office that night drafting the investor story. She still wanted to promise the moon. But as she typed and erased and typed again, the story that fit reality felt stronger. It was not as shiny, but it was credible. That credibility would be their currency when Adrian leaned in with his questions.

Chapter 4: Ethics, Equity, and the Ask

Week five narrowed the runway until it gleamed. The numbers did not change: £62,000, burn £8,500. The team could count months left like breaths before a swim. The deck had a new slide on traction: eight active shops, twelve pilots, weekly active users metric at forty percent for staff, and a ten percent reduction in missed deliveries in the strongest pilot—George’s.

The big question still waited: how much to ask for? Maya wanted a figure that would give them runway and speed: £300,000. Priya nodded; the marketing plan and hiring a sales lead and customer success person made sense. Leo wanted a smaller, safer ask that would focus the team on product maturity: £200,000, recruit a senior engineer, and fix the technical debt first.

They argued through the logic. £300,000 for twenty percent would imply a pre-money valuation of £1.2 million—ambitious, but possible if Adrian believed in market fit and growth potential. £200,000 for twenty-five percent felt humble but could buy them the engineering time they needed. Equity dilution meant something visceral to Maya—her fifty percent would shrink, but the capital could buy more headroom.

They framed the ask to Adrian as a proposed instrument: seed equity with milestones. Priya prepared a slide showing use of funds: 40 percent product development and hiring, 30 percent sales and customer success to lower CAC, 20 percent marketing and partnership building, and 10 percent reserves. Leo’s technical milestones were explicit—deliver a robust confidence scoring mechanism for route suggestions within three months and pilot integrated supplier connections within five months.

Helen offered a final set of words: “Ask for what you need. But structure the deal to protect both sides. Many angels want milestones because that reduces risk. Be clear how you measure success.”

They also rehearsed an ethical stance. If Adrian pressed them about George’s earlier feelings, they would say: we overpromised in one sales call, we apologised and provided a credit, we reworked onboarding language, and we have measurable improvements in delivery rates through manual confirmation. Honest transparency might dent the shine, but it would show they had the character investors look for.

A week before the meeting Maya felt the old temptation to over-promise return of capital. She sat alone with the revenue numbers and her mental projections. She drafted two asks: A asks for £300,000 for twenty percent with faster hiring timelines; B asks for £200,000 for twenty-five percent tied to delivered product milestones. She wrote out reasons for each and then changed the deck to present both options as scenarios.

Leo, practical as ever, said, “Present one plan and a fallback. Investors hate indecision.” Priya agreed. They chose plan B publicly in the deck—a conservative ask with clear milestones—and prepared a paragraph explaining plan A as a stretch option, contingent on accelerated commitments from strategic partners. Honesty again: no illusion of secrecy.

They practiced the investor meeting until the script felt like a conversation. Adrian’s credentials were in the back of their minds: years in logistics tech, an investor who had both backed companies and sat on boards. They imagined him asking about churn, CAC, pricing elasticity, and technical debt. They imagined him pushing back on valuation.

They left the mock meeting drained but steadier. The argument about equity and ethics had clarified the ask. They would be honest, they would protect the product’s reputation, and they would ask precisely for what they required. The only unknown was Adrian himself, and that would be revealed in person on the day.

Chapter 5: The Meeting with Adrian Cole

Adrian Cole arrived on a damp Bristol morning, umbrella folded in his hand and a presence that suggested both patience and impatience. He sat in the ClearRoute office with a coffee and read their deck. He was in his fifties, had a map of logistics investments behind him, and asked the first question without preamble: “Start with the problem. Tell me why the independent shops are still suffering despite many logistics solutions.”

Maya began: “Independent food shops juggle suppliers, manual order books, and staff with limited time. Missed deliveries happen because routes are inefficient and stock runs out because ordering is manual or based on guesswork. For a small grocer, missing three percent of orders can cut into margins that are already thin.”

Adrian nodded. He studied the slide with George’s data. “Show me proof,” he said. Priya brought up the week-over-week numbers for George: initial missed deliveries at eight percent of orders, reduced to six percent after the first week with manual confirmations; revenue impact measured in saved stock of roughly £120 a week. Adrian’s eyes narrowed on the active shop metric—eight active shops, twelve pilots.

“Customer evidence is solid,” he said. “But scale is the key. Tell me your unit economics and CAC.”

Priya spoke: “Our average revenue per shop is £45 per month. Our current CAC—customer acquisition cost—is about £120 per shop, because pilot setup takes a lot of hand-holding. With the right customer success and sales hires, we project CAC down to £60 in nine months.”

Adrian tapped his pen. “With £45 monthly, LTV matters. What retention are you seeing?”

Leo answered, precise: “Current retention is early—around sixty percent at three months among pilot shops that become active. Our LTV, assuming a three-year average customer lifetime, is about £1,620 at current pricing. There’s room to increase revenue through premium offerings and supplier commissions, but we can only scale when the product handles data sparsity and integrates with suppliers for semi-automation.”

Adrian asked about the technical limitations directly: “You say you don’t automate orders. What would it take to get there, and what risks remain?”

Leo took that one. He explained the bug they had found: sparse data caused the route suggestion algorithm to perform worse, and the system needed a smoothing layer and confidence scoring. He laid out the technical roadmap—hire a senior engineer, dedicate three sprints to the smoothing algorithm and confidence model, then build two supplier integrations for pilot semi-automation. The timeline: three months for the smoothing layer, five months for supplier integrations. He also spelled out costs—£80,000 for the engineering hires and platform upgrades to reach full confidence in suggestions.

Adrian listened, then leaned forward. “Walk me through the numbers of your ask.”

Maya spoke, steady now: “We are asking for £250,000 for twenty-two percent equity. Use of funds: 40 percent product and engineering, 30 percent sales and customer success to reduce CAC, 20 percent marketing and partnerships, 10 percent reserves. This gives us about twenty months of runway if we reduce burn by hiring strategically and increasing revenue through conversion of pilots.”

Adrian calculated quickly on his tablet. “You currently have £62,000 and burn £8,500 a month. Why not ask for the full £300,000? That would give you more runway for experimentation.”

Outside, the bakery clock chimed. The team had a deal that bought time and demanded honesty. They had addressed the consequence of Maya’s early promise by being transparent in public with George and in private with an investor. In doing so, they had not solved every problem—data sparsity still needed fixing, and the integration work would be hard—but they had arrived at a path that respected both ambition and truth.

Chapter 6: Interviews, Insights, and a Tight Script

Maya put the laptop on the long meeting table and looked at the five faces around her. They had fewer than six months of runway left, as they'd agreed in earlier meetings — that fact felt like a steady clock in the room. "We need to be clear about the pitch on Tuesday," she began. "The pilot price stays at £79 per month. Priya, you're on messaging. Leo, the demo has to be bulletproof. George, gather customer quotes."

Priya smiled and wrote 'clarity, credibility, customers' on the whiteboard. She liked big language — big total addressable market, or TAM — because VCs often reacted to scale. She put a figure on the board: "TAM of one billion pounds across the sector," she said, deliberately loose. Leo frowned. "We can say the market is big, but don't conflate TAM with reachable market. That's projection vocabulary — it needs context, or investors will call it fluff."

George tapped the table. "I'll run the interviews. We need evidence of the problem and also of our solution working. Don't forget the limitation: our data import currently only handles CSVs with a single header row. If a customer has nested fields or multiple header rows, it fails. We built a workaround, but it's fragile and slow." He explained the limitation like he might to a user: import works for standard exports, but bespoke CRMs or nested JSON need normalization first. Helen, the finance lead, nodded. "That's support cost we haven't fully modelled. We can't pretend it doesn't exist."

They split tasks. Priya would rework the deck to make the £79 pilot prominent and to frame early-adopter economics. Maya wanted three strong customer quotes on the landing slide. George booked four interviews with prospects who had asked about alerts — the team's feature they thought would be a killer. Alerts would notify managers when a threshold was reached; George had overheard clients saying 'an early warning is priceless.'

The first interview surprised them. A regional HR director on the call praised the concept of alerts but said, "It looks good, but my team can't justify the time required to set them up. Our IT refuses unless it's plug-and-play. We need a one-click import or a connector to our payroll system." George's face tightened when the director mentioned the exact payroll provider they hadn't accounted for. "We thought the alert was the product. The customer really meant setup and integration are the product. That's different."

Later, another customer asked about support. "If we run a pilot at £79 per month, who supports us? Who cleans the data?" Helen jumped in. "Support hours scale with messy imports. If we sell 200 pilots next year at £79, that's £15,800 monthly recurring revenue (MRR). But if even 20% need hands-on setup, support costs could kill margins. We have to include that in our projections." She explained MRR plainly: monthly recurring revenue is what subscribers bring each month — multiply by 12 to get an annualised figure, but don't call that ARR unless it's stable.

Maya felt pressure. Investors liked customer evidence, but they also liked tidy metrics. "We need to show that customers will pay £79 and that support is manageable," she said. Leo suggested a conservative model: price at £79, assume 20% of customers need six hours of custom setup at £50/hour equivalent cost. Priya raised her hand. "That makes us sound small. We could say we'll move to a higher tier for enterprise integrations and keep pilots cheap."

They ended the day with a list of contradictions: promising alerts as a killer feature while admitting setup friction; saying pilots cost £79 but not having a packaged setup plan; and believing customers would adopt quickly, even when customers emphasised integration pain. The meeting closed with an agreement to prioritise honest evidence in the deck. Honesty would preserve credibility — and credibility, they all understood in the language of founders and investors, was a currency worth protecting.

Chapter 7: Numbers on a Whiteboard

They met Wednesday in the small office with the window that let in a slice of grey London sky. Helen spread a spreadsheet and turned the screen so everyone could see. "Runway is six months if burn stays as is," she said, drawing a simple cash runway line. "If we hit 200 pilots at £79 in twelve months, that's £15,800 MRR, roughly £189,600 revenue a year. But that's not profit. Support, hosting, and sales will take at least 60% in year one if we don't automate imports."

Priya tapped the screen. "We should put a slide on TAM. Investors love scale. I'll say we can reach three percent of mid-market UK employers, which becomes a credible subset. That still sounds massive. We'll get attention." Maya smiled. "Make sure the slide ties to a go-to-market plan — that's the GTM. Explain channels: direct sales, a pilot programme, and a referral model."

Leo was quieter. "We need to be practical on capacity. If we promise faster onboarding, we're promising engineering time. There's the technical debt — quick fixes piled onto the system. Each patch saves a pitch but costs future months. We haven't modelled the time to build connectors. If we give a delivery date, we must be realistic. Support load scales non-linearly with integrations. It's not just headcount." He used technical-debt vocabulary and explained it as 'shortcuts that accumulate and later require extra work.'

George stood by the whiteboard and drew a funnel. "Conversion assumptions matter. If we target 2,000 leads and convert at 2% to paid pilots, that's 40 pilots. To get 200 pilots we need a different funnel or channels. Priya — how will you source those leads? Paid ads won't work on such a small price point; it's a relationship sell." Priya had a stack of notes. "Content, partnerships, and a few anchor customers who will vouch for us. Case studies convert."

They argued over claims. Priya wanted language like 'we will dominate the mid-market HR analytics space.' Leo shut it down. "Dominance is a claim about share and capability. We can't dominate when our import feature is fragile and a competitor has a similar roadmap. We need to project confidence without overstating capability." Maya tried to mediate. "Investors need to see a market vision, not just a shipping list. But they also ask for evidence. Let's show our pilot pricing and customer interest and be honest about integrations."

Helen interrupted with sensitivity to the cap table. "We have to keep dilution in mind. If we close a seed at the valuation we hope for, we can't give away too much. Our ask has to match the runway extension we need: twelve months of runway at our current burn means an ask of roughly £450k to £600k, depending on growth spend. That's investor vocabulary — 'ask' and 'valuation' — which needs to be justified by traction." She explained valuation simply: it's how much the company is worth for a given share price — a number investors use to decide what portion of the company they get for their money.

Their financial model now had two scenarios. Conservative: 40 pilots in year one, £3,160 MRR; aggressive: 200 pilots, £15,800 MRR. The difference was staffing. The conservative model needed minimal engineering hires; the aggressive model required a small integrations team and a customer success hire. That difference in spend changed the ask and the story they would tell to investors.

By the end of the day, there was agreement on one point: keep the £79 pilot price visible, but be explicit about tiering and integration charges. They would not hide the import limitation or the support cost. It was a lesson in investor vocabulary turned into practice: runway, ask, valuation, MRR, TAM, GTM — all were now terms on the whiteboard, each tied to numbers and a plan. The numbers didn't feel abstract anymore; they felt like decisions that would determine whether the company survived the next six months.

Chapter 8: A Faulty Import and a Faster Competitor

Friday's rehearsal was supposed to be about narrative polishing. Instead, it became a crisis. They had a live demo planned with a mid-sized customer who had agreed to let them import real data. The goal was to show the alert working end-to-end. Priya was on the call selling the story. Leo and George were at their laptops. Maya took the lead on the demo. "This is the moment we show flow and value," she said to the team and then to the customer on the call.

They uploaded the customer's export. The dashboard spun. The import failed. The screen showed an error: "Header mismatch — multiple header rows detected." George's heart sank. He walked the customer through the workaround: open the CSV, delete the extra header, re-upload. The customer was polite — at first. "We run multiple reports automatically. I won't ask my team to clean files for a pilot. That kills the use case."

After the call, the room flickered with tension. "This is the exact friction customers mentioned in interviews," George said. "We thought it was a small problem. It's not." Maya bit her lip. "We can't show that in the pitch. Investors will hear about this and think it's basic." Leo's voice was measured but firm. "We should have built at least one connector by now, or an import wizard that normalises nested fields. We didn't because we prioritised visible features over plumbing."

While they were troubleshooting, a push notification from an industry newsfeed pinged their phones: a competitor had launched an 'auto-normalise import' feature plus alerts in one package. Their competitor's press release used words some investors liked: 'seamless integrations' and 'enterprise connectors.' The timeline stung — the competitor had shipped a connector to the payroll provider the team had just failed with.

Maya felt a flash of panic. In a heated moment that later everyone called a mistake, she told George and Leo, "Tell me how long to build that connector, and we'll promise it to Adrian on Tuesday." Adrian was the investor they had yet to impress after the earlier meeting — a seasoned angel whose feedback had been blunt. Promising something they didn't already have changed the dynamic in the room. Priya looked at Maya with concern. "We can't promise a specific feature delivery date for a critical integration unless we're sure. Investors test claims with diligence and references."

The team argued. Leo said it would take at least four weeks of focused engineering plus QA and some resources for documentation. He warned about technical debt: "If we rush, we'll build brittle code that needs refactoring. That doubles time later and increases support calls." George suggested a staged approach: ship a hosted connector that required initial configuration by engineers, then automate. That would reduce early support but take the customer through a managed onboarding experience.

Priya worried about the pitch narrative. "If Adrian hears a promise and checks with our customer references, we'll get found out if it's not done. Or worse, we'll get a conditional offer that hinges on delivery. That's dangerous." Helen agreed. "If we commit publicly to a date and miss it, our credibility — which matters more than any one feature — will take a hit. We don't have the cash to pay overtime or hire immediately to hit a deadline unless we secure the funding first."

In the quiet after the argument, Maya felt the familiar, nasty mix of ambition and fear. She had interpreted criticism from Adrian as rejection before; now she was desperate not to be rejected again. That fear made her reach for big promises. The team left the room with a new list: make the import transparent in the deck; prepare a contingency slide showing staged delivery; and call their mentor, because outside perspective was becoming the thing they needed most.

Chapter 9: Mentor's Mirror and the Cost of a Promise

They rang Tom, their mentor — an ex-founder who had been through three rounds and one exit. He listened without interruption as they replayed the demo failure and Maya's promise. After a long pause he said, "Confidence is not the same as overreach. Investors want evidence you can do the work. Promising a feature you don't have is not confidence; it's a liability. They will call your bluff." Maya felt exposed. "But if we don't promise, we look small," she said. "We already lost momentum when the competitor launched."

Tom's voice was gentle but direct. "Momentum is useful, but not at the cost of credibility. There are two ways to show ambition: one is to promise deliverables without evidence; the other is to lay out a clear, verifiable plan. Say: 'We will prioritise a connector, here are wireframes, here's the team allocation, and here's the realistic timeline with milestones.' That is confidence. You don't need to say you'll be integrated on Tuesday if you can't prove the path."

He explained the investor due diligence process in plain terms: "Investors will ask for references, product demos, code access sometimes, and your unit economics. They'll call customers and ask about support responsiveness. They can and will verify claims. If you misrepresent, you risk losing more than this round — you can burn a bridge with investors who talk to each other."

Helen wrote updated financials after that call. She modelled what it would mean to hire two engineers for four weeks to build a connector: salaries and employer taxes, additional AWS hosting, QA time, and some contract work for the customer success team to help on onboarding. The burn increased substantially and shrank the runway by more than a month. Suddenly the 'ask' number for a safe runway extension grew. "If we hire before the round, we must raise an extra £150k to be comfortable," she said. "If we wait and promise a date, we risk delivering late and burning credibility."

The debate turned practical. Leo proposed a Minimum Viable Connector that could be built in phases: phase one would be a semi-manual import with a simple mapping UI; phase two would be an authenticated connector; phase three would automate scheduled syncs. Each phase had support costs. George suggested a pilot service package at £249 for customers who wanted hands-on setup — an upsell to cover initial manual integrations. Priya worried this would change the story around the £79 pilot.

They rehearsed answers to likely investor questions. "How do you handle messy data?" Helen took this: "We acknowledge imports are nonstandard. We sell the pilot at £79 for standardised imports. For custom imports, we have a service package. Our plan is to invest in a connector over three months once we close the seed. The capital raise will fund that work and acquisitions." She said it in terms investors would understand: a bridge between product limitations and financing plans.

Rehearsal exposed more weaknesses. When an imaginary investor asked, "What's your churn rate?" they had no hard number; they'd never run a long-enough pilot. When asked about competition, they knew names but not differentiation. Tom made them practice answers until each reply tied to evidence: customer quotes, demo artifacts, costed engineering plans, and payment commitments. It was tedious but stabilising.

By Friday night, the team made a pact. No more promises without a plan. Maya admitted her instinct to overpromise came from fear. She apologised to Leo and George. They agreed to trade a headline ambition for a credible roadmap framed with conservative numbers. The mentor's mirror had shown them the difference between sound leadership and performance under pressure. The next step was to incorporate that honesty into slides and practice until they could answer uncomfortable questions in plain language.

Chapter 10: Final Rehearsals and the Investor Table

Pitch day felt like a summit. The meeting room at a small VC had glass walls and a long wooden table. Priya wore a navy blazer and a folder of printouts. Maya opened with the company's story — what they did, why it mattered, and how customers reacted. The £79 monthly pilot was on slide two, clearly labelled. Helen, seated with the laptop showing the financial model, watched the investors' faces. "We ask for £550k for 12 months of runway and to build priority integrations," she said when it came to the ask. She explained the use of proceeds: engineering hires, customer success, and marketing for channel partnerships.

The investors were three people: a partner who was direct, an associate who took notes, and a product-focussed angel who asked technical questions. The partner started with the usual: "Walk me through a pilot — what does £79 get a customer?" Maya answered with a calm, practiced cadence: "It gives access to core alerts, standard imports, and a two-week onboarding. For bespoke integrations there is a paid setup package. We highlight this up front so customers understand what the pilot covers and we protect unit economics." She mentioned the staged connector plan briefly and offered a concrete timeline if they funded the work.

The product-focused angel leaned forward. "How will you handle the data import variance?" George took that one. He explained the CSV limitation, showed anonymised screenshots from the failed demo and the error message, and then presented the phased connector roadmap. He was candid about the customer evidence: quotes that praised alerts but asked for a simpler import. The investors appreciated the honesty. The associate scribbled numbers on a notepad.

The partner asked about competitive risk. Priya answered with the reframed narrative: the competitor had launched a similar feature, but their integration required a paid enterprise contract and no onboarding service. Priya emphasised that their advantage was a combination of low-cost pilots and a strong customer success model if funded — an attempt to show defensibility beyond a single feature. She defined jargon when relevant: 'defensible' meant customers would find switching costs or value enough to stay.

Then came the hard questions. "What happens if you scale to 200 pilots and you discover 30% require hands-on setup?" the partner queried. Helen replied with the conservative model they'd rehearsed: additional hires and an adjusted burn rate, explaining how it affected runway and dilution. She used simple arithmetic to show investors the dependency of hires on revenue growth, and how the fundraise would be sized to prevent being forced into bad decisions. This clarity impressed the table.

An unexpected moment arrived when the product angel mentioned the competitor's new connector. "How will you avoid being copied?" he asked. Leo responded, not with tech bravado but with systems thinking: "We won't win on a single connector. We'll win on workflows and data lineage — how alerts map to actions — which require partnerships and deep customer understanding. Plus, our customer success strategy will produce playbooks they can't replicate quickly." He explained 'data lineage' in plain terms: a traceable chain from raw data to alert so errors are easier to fix.

Towards the end, they were asked to summarise risks. Maya took that, framed risks as three: technical (imports), commercial (conversion funnel), and execution (hiring). For each, she showed mitigations: staged connectors, pilot packages at £249 for bespoke setups, and milestone-based hiring funded by the ask. Her honesty paid off. One investor commented, "You know your limits and you have a plan. That's better than a shiny demo with no contingency."

After the meeting, outside the glass, the team exhaled. They had not overpromised. They had shown the failed demo, owned limitations, and presented a realistic plan to fix them. The investors would deliberate, but the team felt they had traded theatrical confidence for credibility — and that had changed the tone of the conversation. As they walked back to the office, Maya turned to the others and said, quietly, "We kept our word to be honest. That will matter, whatever the decision."

Chapter 11: Assumptions on Trial

Maya pushed the conference room door open and felt the tired quiet that lives in start-ups during a crunch. Around the table were Leo, Priya, George and Helen. The whiteboard still held sketches from last night: routes, numbers, a rough cap table.

Maya began, voice even. "We need to test everything again. We built assumptions into the product — who it helps most, how much they will pay, and what work the product must do. We treated those as facts. They are not."

Priya rubbed her temple. "We have pilot data, but small pilots can mislead. We need to separate confirmed facts, estimates and ambitions. That's good practice for investors. Also, it will keep us honest when we talk to Adrian again."

George added, quietly, "And honest with ourselves. I said earlier we should focus on single-shop owners. The pilots show that's not the clearest fit. I worried we were stretching the product to fit a market that doesn't care."

Helen, the mentor who had been blunt before, nodded. "You all promised clarity in 'Ethics, Equity, and the Ask'. This is the place to show it. What do you have that's confirmed?"

Maya walked to the whiteboard and wrote three columns: Facts, Estimates, Ambitions. She wrote underneath Facts: 'Reduces stockouts in multi-branch shops (2-5 branches); manual CSV imports working; basic transfer coordination tool; customer willingness to pay observed.'

"Facts from pilots," Leo said. "Five clients, two of them with 3 branches, one with 4, two with 2. Across these, we measured stockouts and coordination errors before and after using our board. That's not a national sample, but it's real."

Priya added to Estimates: 'Market segment size of 2-5 branch shops in the UK, pricing elasticity around £99 per branch, monthly churn baseline 5-7% if support is slow.'

Maya smiled and wrote Ambitions: 'Automated supplier purchase orders, full POS integrations, rollout to chains of 20+ stores.' She then underlined it and drew a dark line through those ambitions. "Not this month. Not for our next investor conversations."

The team felt the relief of narrowing. Helen spoke softly, "Investors like clarity. They want to see a believable plan. That means showing what you already proved and admitting what you cannot do yet."

They moved to customer evidence. Priya had arranged follow-up calls. "Fitz & Sons, a bakery with three branches, reduced reported weekly stockout events from an average of five per branch to two per branch in six weeks. They said the coordination feature helped, not because it automated orders, but because staff could call transfers and the dashboard flagged them. That's a behavioural fix, not a supply-chain miracle."

George read from a transcript. "Marta at Green Grocer, two branches, told us she would pay £99 per branch if onboarding was simple and support responded within four hours. She said the onboarding fee must be small because initial cashflow is tight. She tried our CSV import and hit the 'bad header' bug — she called it a 'faulty import' and almost stopped. That's evidence against easy onboarding."

Leo, who had stayed close to engineering, said, "The import issue is real. It's reproducible when files come from older POS exports. Fixing it is straightforward but will take development time and careful testing across formats. It's not a reason to say the product is unreliable. It's a limitation we must state openly."

Maya left the room thinking of the whiteboard columns. Facts, estimates, ambitions. It was a simple way to stop hope from becoming a plan without proof.

Chapter 12: Customers Tell the Real Story

Priya organised the customer conversations and scheduled three visits in a week. She believed that firsthand observation was more convincing than a slide deck. "We need to see how they use the board, how staff react and what they call 'urgent'."

On Tuesday, the team visited Fitz & Sons. The bakery's manager, Tom, met them behind a glass case of pastries. He had used their dashboard for eight weeks. He opened his tablet and showed the team the transfer log and the alert feed.

Tom said, "We used to run out of strong flour on Saturdays — every two weeks at least. Customers would walk away when we couldn't make what they wanted. The transfer feature meant I could see stock in our close branch and ask for a transfer before the Saturday rush. It saved us lost sales and the staff time otherwise spent calling around."

Maya asked, "Did it save money directly?" Tom looked thoughtful. "It didn't magically lower our invoice cost from suppliers, but we stopped losing small sales. If you count the extra pastries sold, it covers the cost of three months of the software for the whole bakery."

George took notes on his phone. He recognised patterns he had seen earlier but now could point to them in a customer's workspace. The product was doing what it meant to do in small chain contexts: avoid stockouts by coordinating inventory between branches."

Priya recorded Tom's words and asked for permission to use his quote in their investor deck. Tom laughed. "Use my name if it helps you get this thing right. But fix the imports. Our old POS exports break your CSV parser and it made our assistant panic on the first day."

Leo felt both vindicated and frustrated. He had argued the fix was necessary. Now the evidence was in a real voice. He promised to prioritise the import robustness work and validate more export formats in the next two sprints.

On Wednesday, the team met Marta at Green Grocer. The store had two branches and a small but loyal customer base. She had a simpler operation than Fitz & Sons and was sensitive to cash flow.

Marta said, "I tried other tools. They were complicated, built for big chains. Your dashboard feels lighter and our staff learned it quickly. But you have to keep support fast. We don't have time for a three-day answer when an order is wrong."

Priya asked about price. Marta nodded firmly. "£99 per branch is fair if you give us quick support. For two branches, it's still the price of a part-time staff member saved from making calls and chasing deliveries."

George asked how they measured the improvement. Marta pulled a notebook. "We used a handwritten log before and after. That is not perfect, but in four weeks we saw fewer notes about 'no stock' on the shop floor. Not every product, but enough to be meaningful."

Helen later commented, "Customer evidence doesn't need to be perfect. It needs to be honest and repeatable. You have to show investors that the signal is real, not noise."

As they prepared the materials for the investor update, Leo queued the import fixes and a small analytics improvement that would let them show stockout trends over time in one graph. It was a modest set of product changes but had high value in investor conversations: evidence and transparency.

Chapter 13: Maya's Apology and Leo's Demo

Maya found George at his desk, head down over a messy log of CSV headers and POS formats. He was quiet, and she felt the heaviness of conversations left too late. She pulled the chair opposite and sat for a moment before she spoke.

"George, I want to apologise. I pushed decisions too fast and I didn't bring you into the loop when it mattered. You were right to warn us about compatibility issues and the niche the product serves best. I'm sorry I didn't listen sooner."

George blinked. He had expected a corporate apology, but the simple directness disarmed him. "Thanks, Maya. I felt sidelined, but I also saw what we can do if we focus. Let's make sure these fixes are logged with clear acceptance tests so we don't stumble again."

Maya nodded and wrote 'acceptance tests' on a sticky note. She knew words meant little if not followed by clear action. It mattered to her that the apology came with a plan.

Later that day, Leo prepared a demo for the whole team. He cleared his throat and said, "I want to show the MVP as it works now, and its limits. No tricks. I'm going to use actual data from Fitz & Sons, anonymised, and show the import process and the transfer request flow."

He started with the dashboard. The interface was simple: a list of branches, stock levels, alert badges for low stock, and a transfer queue. He clicked on a low-stock item from Branch B and initiated a transfer. The transfer created a request that showed on Branch A's tablet within seconds, with the suggested quantity and a reason — 'Saturday rush, replace 5kg strong flour.'

Maya watched staff faces on the simulated screen. She liked the plainness of it. The interface did not hide complexity. It simply coordinated people and showed evidence of the request: who asked, when and why.

Leo then demonstrated the import feature and the limitation. He pulled a CSV from an older POS. The import parser threw an error: 'header mismatch'. He explained how the parser works: it expects a certain set of fields and formats. When a POS export deviates, the import fails.

He said, "We can and must make imports more robust. The fix will be to expand the parser to accept alternative headers and to implement a quick mapping step in the UI so a user can tell the system which column is stock quantity, SKU, and so on. That is a two-week sprint for the parser core and another week to build the mapping UI and tests."

Priya leaned forward. "Good. Let's add a support script for manual mapping in onboarding. If support can do it quickly, Marta's concern about fast answers is addressed in the short term."

George added, "And we should log every failed import with the headers attached so we can prioritise the most common formats. We'll get a list of the top ten POS exports and make sure those work automatically."

Helen smiled. "Leo's demo is excellent because it shows reliability in what works and transparency about what does not. Investors will appreciate that — they don't want a polished lie. They want a realistic plan backed by data and timelines."

The team left the office with a clearer plan and a fragile optimism. They had fixed what they could, apologised where needed and prepared to speak plainly with an investor who would test both their facts and their character.

Chapter 14: Ownership, Pride, and the Hard Choice

The topic of ownership returned to the table like an old debt. The founders had to decide how much equity to offer for the £180,000 investment. Maya called the meeting specifically to resolve it because the term would shape control and decision-making for the year.

She wrote the current cap table on the board for transparency: "Before investment: Maya 40%, Leo 30%, Priya 20%, George 10%." She added a note: "We are proposing to offer 12% to the investor for the £180k, which implies a post-money valuation of £1.5M."

Leo asked, calmly, "That would dilute us proportionally, right? After investment, our stakes would be Maya 35.2%, Leo 26.4%, Priya 17.6%, George 8.8% and Investor 12%."

The numbers sat in the room like a beam. George spoke up, voice tight with emotion. "I understand dilution is part of the process. But with those numbers, will we still have control? I worry that if we give a board seat or protective rights, decision-making shifts to someone whose priorities may differ."

Priya interjected. "We have to weigh pride against cooperation. If the investor wants governance protections, we can negotiate terms that allow them oversight without daily interference. If we refuse a reasonable seat, they may walk. We also need the lead to help with introductions to paying customers."

Helen, who had watched similar fights before, said, "Investors will ask for standard protections: information rights, a board seat for larger checks, and veto rights on things like liquidation preferences. You can limit those to major items and keep day-to-day control. It's about writing sensible terms, not about emotional ownership."

Maya felt the old impulse to protect control. But she also recognised they needed the money to build the import fixes and hire a support person. She proposed a compromise: offer the 12% equity, agree to quarterly board meetings, and give the investor one board observer seat rather than a full seat. This would mean the investor sees the data and advises, but the founders retain a majority and operational control."

George studied the compromise. He wanted more guarantees about the investor's behaviour. "Can we specify in the term sheet that the investor cannot veto product hires or change our mission without consent?"

Priya suggested, "We can add a short protective provision list: approval for major transactions, debt above a certain threshold, or sale of the company, but not product-level decisions. Also, we should keep a clause requiring investor introductions to paying customers as part of the agreed support."

Maya said, "And reporting must be monthly but concise. We'll produce an investor pack with defined metrics: monthly recurring revenue (MRR), customer churn, average revenue per branch, number of active branches, and the import success rate. We'll commit to rapid responses to investor queries and to customer references they require for diligence."

The team debated one more issue: preferred shares versus common shares. Leo argued for a simple equity deal with common shares to avoid complex liquidation preferences that could later penalise founders if the company exits at a lower valuation. Priya suggested a remedy: limit investor rights and avoid aggressive liquidation preference language. They'd ask their lawyer to craft a plain vanilla term sheet reflecting those points.

They agreed to have the lawyer draft a plain term sheet reflecting the agreed points before the meeting with Adrian. The next steps were clear and, for the first time in weeks, the team felt aligned.

Chapter 15: The Revised Pitch and Adrian's Decision

Adrian Cole arrived at the co-working space at ten on a damp Thursday morning. He was punctual, polite and direct. He had met them before, but this would be the pitch after the team had narrowed the product and clarified the use of funds.

Maya opened the presentation with customer evidence. She said, "Adrian, we learned our clearest early value: reducing stockouts and coordinating transfers for shops with 2-5 branches. Fitz & Sons saw a 40% reduction in stockouts; Green Grocer reported fewer urgent reorders. We matched that with a price that customers accept: £99 per branch per month, plus a one-off onboarding fee of £249 per branch."

Adrian listened, arms folded. He is an investor who has seen many pitches and knows the usual pattern of overpromising. He asked, "How many paying branches do you have now and what's the cost to support them?"

Priya responded, calm and measured. "We have six paying branches across our pilots: three at Fitz & Sons, two at Green Grocer and one small trial that's not converting. Our cost per branch is currently £10/month in hosting and third-party services on average. We propose hiring one full-time support person at an annual cost of £48,000 to keep response times fast and churn low."

Adrian nodded. "Talk me through the ask and the cap table." Maya showed the numbers on a slide: "We are asking for £180,000 for 12 months of runway. Before investment: Maya 40%, Leo 30%, Priya 20%, George 10%. We propose offering 12% equity, which implies a post-money valuation of £1.5M. After investment, stakes will be Maya 35.2%, Leo 26.4%, Priya 17.6%, George 8.8%, Investor 12%."

Adrian raised an eyebrow and said, "A straightforward number. What are you committing to with this money?" Maya replied, listing the breakdown: "£48,000 for the support hire, £50,000 for development to fix imports, build a CSV mapping UI and the analytics widget, £12,000 for hosting and third-party costs, £20,000 for sales and targeted marketing, £20,000 total minimal founder draw to keep the team afloat, £10,000 operational costs and £20,000 contingency."

Adrian tapped the table. "And your milestones?"

Priya answered, "In six weeks: robust import parser and mapping UI shipped. In eight weeks: analytics widget showing weekly stockout trends. In four weeks: support hire onboarded and trained. In twelve months: 30 paying branches, churn under 7%, import success on top ten POS formats hitting 90%."

Adrian studied the slides and asked about governance. "What are you offering in terms of board involvement?" Maya explained their compromise: the investor would be offered 12% equity, one observer seat at quarterly board meetings and protective provisions limited to major financial decisions and asset sales. Day-to-day decisions, hiring below a named threshold and product prioritisation would remain with the founders.

Adrian considered that and then asked tougher questions: "What's your customer acquisition cost (CAC) and lifetime value (LTV) estimates? What happens if a competitor builds a cheaper tool? How sensitive is your model to price?"

After Adrian left, the team gathered in the kitchen and shared a simple, quiet smile. They had chosen cooperation over pride, admitted limits, shown evidence, and negotiated governance. The pitch had changed everything: not by bringing in money yet, but by refining what they were asking for and how they would use it. Now they had to deliver on those promises.

Chapter 16: The Table Is Set

They met in the small conference room on the third floor of the coworking space. A whiteboard filled one wall, and the startup's logo, a simple wordmark and a quirk of teal, looked small and earnest in the corner. Maya had written 'Adrian - 14:00' on a sticky note and stuck it to the board like a pin. She felt the note under her fingertips as if it were a heartbeat.

Everyone arrived early. Leo carried his laptop in a battered bag, Priya had a folder of customer notes, George brought spreadsheets, and Maya had printed slides. They had agreed, in the last difficult conversation, to do this honestly. No staged data, no trickery. Honesty had cost them trust before, but it had also given them clarity.

Maya spoke first. 'We have two months of runway if we keep spending as-is. We are bringing in £3,000 monthly recurring revenue from three paying customers. Our net burn is about £27,000 a month, so cash runs out in roughly eight to nine weeks unless we change something.' She said the numbers plainly. The room listened to numbers the way sailors listen to a storm's wind.

Priya added, 'We have an incoming pilot opportunity with a mid-size retailer — five stores — who will pay for a three-month pilot at £5,000 total. That would add cash and critical validation.' She tapped the sheet of notes she had prepared, showing the retailer's name and their buying director: Helen. They had been talking to Helen for weeks and she had been sceptical but engaged.

George placed his spreadsheet on the table. 'If we accept outside investment we dilute current ownership. Right now the founders hold 40% for Maya, 30% for Leo, 20% for Priya and 10% for me. We need to be precise about valuation and conversion terms. A convertible note, like the one Adrian prefers, doesn't set a price now but will convert on a future priced round. We need to protect ourselves from a cap that is too low.'

Leo nodded. 'Technically, our import pipeline has an important limitation. We can't reliably reconcile SKU changes older than 60 days without manual mapping. That means if a retail partner exports historical CSVs with inconsistent codes, our forecast accuracy drops. We've been working on it, but it's not fixed.'

They rehearsed how to say that in plain English. Maya practised: 'We integrate with your till system; sometimes the product codes change, and our tool confuses them. We will flag mismatches and work with your team to reconcile. We won't pretend it's perfect.' It felt stilted and yet right. They had lost an investor before because of a staged demo. The scar made them careful.

They planned contingencies. If the demo worked perfectly, they would show real live metrics from Helen's trial data sandbox — anonymised and agreed. If it failed, they would explain the limitation, bring up actual customer evidence and show a step-by-step plan for fixing it. They would emphasise retention, paid pilots, and precise milestones rather than inflated promises.

Then, twenty-seven minutes into the meeting, the chart for one store stalled. Numbers failed to load. A red banner appeared in the interface: 'SKU mapping mismatch: unknown codes detected.' There was a short, uncomfortable silence.

Chapter 17: The Glitch and the Choice

The red banner cut through the room's practiced calm. Maya's throat tightened. It was the exact situation they'd promised not to pretend away. She took a small breath and decided to speak before the silence could invent stories.

Maya: 'This is the limitation we were talking about. The store export has product codes that changed last quarter and our mapping didn't reconcile them automatically. The forecast looks wrong because the system is seeing mismatched SKUs as new products.'

Adrian did not flinch. He only looked more alert. 'So the demo is broken. Explain what that means for a paying customer.'

Priya stepped in. 'It means live decisions could be noisy until we reconcile codes. In practice, we flag records for manual review and our customer success team — that's me — works with store managers to map codes. After mapping, forecasts improve. For example, Store B had a 35% improvement in forecast accuracy after a reconciliation session last month.'

Maya opened a folder with emails. 'Here is direct customer evidence. An operations manager at Store B wrote that the alerts became useful after we resolved a single mapping cluster. We will not hide the process. We will show customers the flag, offer the mapping session, and include it in onboarding.'

Adrian read the email quietly. He looked up. 'You're honest. That is rare. But it's not enough. The investor in me wants to know how you'll scale that manual mapping. Who will do it when you have 100 stores?'

Leo answered. 'We plan to automate mapping with machine learning models that learn alias patterns from multiple customers. But models need training data. The Helen pilot will provide rich, consistent EPOS exports that let us build a high-quality alias set. Also, we'll build a mapping UI that allows store staff or the buyer to confirm mappings quickly. It reduces manual hours and trains the algorithm at the same time.'

Adrian leaned back. 'What do you expect from me if I invest? And what are you asking for now?'

Maya had rehearsed this. 'We are asking for £120,000 in convertible funding to extend our runway to complete technical work and to secure paid pilots. The convertible would carry a standard 20% discount on the next priced round and a £3 million valuation cap, and interest at 5% per annum. We propose that Leo is accountable for the import pipeline fixes and that we secure Helen's pilot as a condition for drawdown.'

George interjected with practical numbers. 'With £120,000, our net burn drops to a runway of about six months after we reduce discretionary spend and prioritise engineering. If we secure the £5,000 pilot with Helen, we can also negotiate a slightly larger pilot in month two, which gives us breathing space and critical development data.'

Adrian watched them. 'Those are sensible numbers. I don't hand out cheques in a meeting. But I will give conditional interest. Due diligence, a signed pilot with Helen at the terms you outlined, and a clear engineering timeline with milestones will be required. If you hit those, I'm prepared to move forward with a £120,000 convertible at the terms you propose. But I want governance too: I need a right to appoint an observer to the board during the note period.'

They divided the responsibilities with the unvarnished practicality of people who have had to survive before. Maya would keep investor communications, Leo would lead product fixes, Priya would own customer delivery, and George would manage the cap table and financial modelling for the notes.

Chapter 18: Terms and Trade-offs

Negotiation is small decisions stacked together. Over coffee at their usual table, the founders went line-by-line through Adrian's conditions. They had to protect their ownership while making the investment attractive. George sketched possible dilution scenarios on a napkin and then transferred them to the whiteboard for clarity.

Maya started, 'We don't want a low cap that undervalues the effort. A £3 million cap is reasonable, but given our traction maybe we ask for £3.5 million?'

Adrian had mentioned the cap in passing during the meeting; now they had to pick a number and justify it. George explained the math. 'At a £3.5 million post-money cap, the £120,000 converts to roughly 3.4% pro forma. That keeps founders above acceptable thresholds while giving Adrian a meaningful stake at conversion.'

Leo chimed in about governance. 'We're okay with an observer, but we should avoid ceding board seats until a priced round. Adrian needs to see progress, not take control.'

Priya agreed: 'We will accept the observer clause and a covenant to hit milestones. But board seats should be for later when there is a priced round. We need to keep motivation high for the founding team.'

They drafted a response email to Adrian with proposed changes: a £3.5 million cap instead of £3 million, a 20% discount unchanged, 5% interest, an observer seat rather than a board seat, and clear milestones. They also attached a customer pilot template and asked for his input on the timeline.

An hour later, Adrian called. His voice was calm but brisk. 'I understand the cap request. Let me be frank: valuation cap is a function of risk. I like your honesty and the Helen pilot. I'm willing to agree to £3.5 million cap if you get Helen's signed pilot within seven days and show initial mapping results from their data within three weeks.'

The team exhaled. It was a tense win. Conditions remained, but they were achievable. Adrian continued, 'I'll also ask you for monthly investor updates with key metrics: MRR, cash, churn, activation and the mapping progress. If you miss the mapping milestone, the terms revert to the original cap.'

Maya asked about the pilot fee and data access. Adrian said, 'Make sure Helen understands the pilot is paid and that you need CSV exports and API access when possible. I will reach out to my contact at that retailer group to reinforce the value and the need for clean data.'

Within two days Priya had Helen on the phone. Helen had been cautious but pragmatic — that combination had made her a valuable early customer. 'We like the way you talk about trade-offs,' Helen said. 'We also like a tool that tells us what it can and cannot do. If you can promise a pilot with clear objectives and a quick mapping session, we will sign.'

Priya sent a draft pilot agreement and scheduled the first technical call. It included the three-month scope, £5,000 total fee, data export access in week one, and a joint acceptance criteria document that listed forecast accuracy, mapping completeness percentage, and retention targets. That clarity made the retailer comfortable.

They signed the conditional term sheet subject to legal checks and the pilot. The term sheet felt like a contract with a ladder: each rung a milestone leading to the money. It also brought clarity to the roles and a shared rhythm. They left the office later that evening with a mixture of determination and relief.

Chapter 19: Pilot, Pressure, and Progress

Helen's retail IT team provided the first data export within forty-eight hours of signing. The spreadsheet was large and messy, as retail spreadsheets often are: changed SKUs, concatenated descriptions, and odd sale dates. For the team it was exactly the kind of messy realism they needed.

Leo dove in. He set up a separate environment and ran the alias extraction script. The script found many potential matches but also many ambiguous cases. He logged those as 'mapping tasks' and set up the UI prototype for Helen's operations manager to confirm. The UI was spare: highlighted rows, suggested matches and a button for the retail manager to confirm or reject.

Priya arranged a mapping session with Helen's operations manager, Julie. They ran the mapping session over a screenshare. Julie clicked through suggested matches and confirmed sixty-five percent quickly, held back on the rest because of bundles and renamed seasonal SKUs. The session took ninety minutes and the team collected valuable feedback about how buyers think about product families.

After the session Leo ran the forecasting engine again. The forecasts improved for most stores. They measured forecast accuracy by comparing predicted sales to actual sales when they had a holdback sample from last month. The initial accuracy rose from an average of 58% to 74% after the mapping confirmations. They logged the percentage and prepared the demo to show Adrian.

Maya prepared the investor update. She included the mapping session notes, the accuracy improvement numbers, the signed pilot agreement, and the new sprint plan. She also included candid notes about the remaining edge cases and the plan to tackle bundles and seasonal renames. She sent the update to Adrian and attached an invitation for a short demo.

Adrian watched the demo carefully. He asked detailed questions about the sample holdback, statistical significance and how the mapping UI handled ambiguous items. The team explained the heuristic thresholds they used and how human confirmation trained the model. Adrian remained quiet for a long time after the demo; then he said, 'This is promising. I'm glad you didn't hide the glitches. I will proceed with the formal convertible documents subject to standard checks.'

They celebrated, briefly. The conditional interest was moving to formalities, and they had measurable improvements to show. But they were aware of a twin pressure: execution and runway. Even with the term sheet moving forward, the money wasn't in the bank yet. The pilot funds of £5,000 helped, but it didn't solve the long-term runway problem.

George modelled a conservative scenario where no additional customers were added in the next three months. With the pilot payment and the prospective convertible, they could extend to five months by cutting non-essential spend and delaying two hires. The team agreed to postpone hiring and to focus on the mapping and customer success efforts that would increase retention.

They also used the pilot to improve onboarding materials. Priya created a short guide for the mapping session and designed an escalation path for ambiguous SKUs. The onboarding guide reduced time spent on calls and helped Julie at Helen quickly find the right team members to confirm mappings.

They filed the papers and prepared for the external review. The next weeks would be about execution: keeping the pilot healthy, improving the mapping, and proving that the product could scale without heroic manual effort. The rhythm of small wins and transparent communication had become their operating system.

Chapter 20: The Pitch That Changed Everything

A month after signing the pilot, Maya stood in a bright meeting room at Helen's regional office. The room smelled faintly of coffee and vinyl. Helen was there with Julie and two store managers. On a screen, Maya had prepared six slides: problem, evidence, what the pilot would deliver, the mapping plan, expected outcomes and what success would look like for Helen.

The pitch was different from their investor rehearsals. This audience cared about day-to-day operations, not convertible caps. Maya spoke plainly: 'We will give you a prioritised ordering list, daily stock alerts and weekly forecasts for your five stores. The pilot fee is £5,000 for three months. Our acceptance criteria include forecast accuracy above 70% and mapping completion for core SKUs at 90%.'

Helen listened and asked targeted questions. 'How fast can we recover after a seasonal rename? What about promotions that change sale patterns?' Maya answered. 'For renames, our mapping UI allows quick confirmation and for promotions we have a seasonal adjustment parameter that the team tunes during the pilot. We will also include a weekly review call so you can see how recommendations translate into actions.'

Julie looked at the mapping interface, then at the store managers. 'I like that the system asks us to confirm rather than decides on our behalf,' she said. 'It respects our local knowledge.'

Maya nodded. 'That's the point. We provide suggestions but the final commercial decisions stay with you. We deliver clarity and time-saving recommendations.'

One store manager, a woman named Hannah, asked about stockouts. 'If I rely on your forecast and still run out of stock, what do we do?' Maya admitted that some early models weren't perfect but pointed to the in-platform alert system that flagged low inventory and provided suggested reorders. 'We aim to reduce stockouts by improving signal quality. The mapping step reduces a lot of early noise, and paired with your local knowledge it should reduce stockout incidents significantly.'

Maya then presented the honesty clause she had written: 'We will be candid about limitations and provide a remediation plan when we identify mismatches. We'll also give you access to the mapping tool so your team can participate.'

Helen paused. 'That clause worries our legal team, but it makes sense commercially. We want partners who will tell us the truth when things go wrong.'

Priya closed the pitch with the practicalities of onboarding. She outlined the mapping session, the weekly acceptance metric reviews, and the exact data feeds required. They agreed a start date and scheduled the first mapping session for the coming Tuesday. The pilot contract was signed by both sides before the end of the day.

Back at the office the team exhaled collectively. The pilot had not only been signed, it had produced a relationship built on clear expectations and shared responsibility. The clarity helped them pivot away from the temptation to overpromise in order to close deals.

Adrian's external reviewer arrived two weeks later. The reviewer audited the mapping tests, the code, and the training data pipeline. She found the approach sensible and the results reproducible. Adrian released the second tranche of the convertible funds after the reviewer signed off on the milestone and after he verified the signed pilot agreement with Helen.

The pitch that changed everything wasn't a single moment of magic. It was the decision to tell the truth, to listen to customers, and to trade quick theatrical wins for slow, durable trust. And that, in a way, made the startup feel steadier, more human and more likely to last.

Chapter 12: Morning Light

The next morning the office smelled of coffee and fresh printouts. Sunlight came through the blinds in strips across the whiteboard where someone had written, in a shaky marker, "Pilot plan." Maya sat at the kitchen table with her laptop open and a tired, satisfied smile. The pitch had ended late the night before. They had answered hard questions. Adrian had not given them a cheque on the spot, but he had given them his attention and a path forward. That, now, felt like real work ahead instead of a dream.

Leo came in with a thermos and two paper cups. He moved slower than usual, as if he were testing the quiet. Priya followed with a printed email that made her fingers tremble even before she put it down. They had patched things up after the tense moment in the meeting room. The repair was not dramatic, but it was honest—an apology from Leo about assumptions, and a promise from Priya to involve him earlier on product messages. They had both slept poorly and arrived ready to keep that promise.

"Read the email," Priya said. "From FreshWay. They want to run a pilot."

Maya's heartbeat quickened. FreshWay was the food-shop group who had watched the demo closely and asked the practical questions. Their operations manager, Carla, had been careful and concrete in her questions. "How many stores? For how long? What's the cost?" Those questions meant the group was serious.

"It says ten stores for eight weeks," Leo read aloud. "Free pilot, but they want performance metrics and a success fee if we raise sales by more than a baseline. And they want a short exclusivity window—sixty days—while we run the pilot. They want a signed NDA first."

"Okay," Maya said. She felt every decision tighten into focus. "This is exactly what we asked for. Free pilot, where we can prove the product without asking them for money up front. Success fee and exclusivity are normal. We'll have to decide the metrics and the timeline and what success looks like. And we must give Adrian his due diligence updates."

They spent the next hour drafting replies. Maya typed while Leo and Priya argued gently about the technical details to include. They agreed on a clear set of pilot terms: installation in ten stores chosen to represent different layouts; an eight-week active data collection period preceded by one week of calibration and staff training; metrics centered on out-of-stock reduction and sales lift for the items covered by ClearRoute; weekly progress reports shared with FreshWay; a success fee set as 5% of incremental sales attributable to the system for a period of three months after the pilot, capped at a contract maximum; and a 60-day exclusivity window during which FreshWay would not onboard a competitor for the same solution. They also agreed to sign an NDA and to set up a shared data room for operations and technical information.

Those terms felt fair. Maya insisted on them being conservative. She had learned her lesson about mixing ambition with fact. "We will label everything," she told Leo and Priya. "Facts, estimates, ambitions. Facts are what we can prove today. Estimates are our best numbers based on the pilot. Ambitions are what we hope to reach if the product scales."

"Adrian's email came in while you were writing," Priya said, handing Maya her phone. The subject line read: "ClearRoute — next steps."

Maya read it aloud to the room. "He says he is prepared to begin due diligence. He wants three weeks to review the team, code, customer conversations, and legal documents. He asks for a preliminary term sheet if due diligence goes well. He wants exclusivity for thirty days during which he would be lead investor, conditional on the FreshWay pilot showing promising results. He also asked for the shared data room and access to our pilot contracts."

"Three weeks is quick, but reasonable," Leo said. "If FreshWay signs, that shows customer interest. Adrian will want to see the pilot setup and early telemetry. We can give him our codebase overview and a demo environment. He won't want surprises."

"No surprises," Maya agreed. "We will be transparent. We'll show what works and what still needs work. That's how we get trust."

By noon, Carla from FreshWay was on the line. She sounded practical, with the calm of someone used to managing risk across many stores. Maya kept her voice calm and clear, remembering the advice she had given herself the night before about distinguishing facts, estimates and ambitions.

"We want to run the pilot," Carla said. "Ten stores, eight weeks. We'll pick two stores from each region. We expect you to handle installation and training. We will assign a single point of contact at operations. We need an NDA and a timeline by the end of today. One more thing: we can't accept any technology that interferes with our POS or inventory systems."

"Understood," Maya replied. "We will not change your POS. Our system observes shelves and uses computer vision and sensor data to estimate stock levels. Fact: we do not write to your POS. Estimate: in past pilots we reduced out-of-stocks by around 25 percent on covered SKUs within the pilot period. Ambition: with full roll-out we hope to reduce out-of-stocks by 40 percent and increase sales by five to eight percent. But those are ambitions; our contract will make success measurable and tied to sales lift in pilot stores."

Carla paused, then asked the question Maya feared: "What happens if your estimates are wrong?"

"If we miss the targets in the pilot," Maya said honestly, "we still deliver the analytics and a plan you can use. We'll walk through what went wrong. We'll propose fixes—maybe additional sensors, or a different training period. We won't charge beyond the success fee if we don't meet the agreed thresholds. We expect there will be learning."

Carla's tone softened. "Okay. Two more things. First: we need a clause about data privacy. Second: we need to know what the biggest risk is, from your side." Maya had expected the privacy question; she had a signed policy ready. The risk question felt like an invitation to be frank.

"The biggest realistic risk is integration with messy, real-world store environments," Maya said. "Old shelving, odd product placements, different lighting, and older point-of-sale records can create mismatches between what our sensors see and what your systems report. That can lower accuracy in some stores. We can mitigate this by an initial calibration week, staff checklists, and manual audits. But it is the main risk on our side—accuracy in non-ideal conditions."

"Thank you for being clear," Carla said. "I prefer we know the risks up front."

By late afternoon, the NDA had been exchanged, the pilot contract drafted with the conservative terms they had agreed, and Adrian had replied with a short list of due diligence items: references from past customers, a treemap of the code architecture, and details of any third-party services. The three founders sat back and exhaled as the office changed from urgent to busy: the work had real stakeholders now.

"This is not the end of anything," Leo said, pouring a second cup of coffee for Maya. "But it's the beginning of concrete work."

Priya spread her hands over the printed contract as if setting down a promise. "We have to be prepared for hard days. The pilot will show everything—our strengths and our weaknesses. But having a real customer and a potential lead investor waiting means we have a stage. We can either fail forward, or we can succeed. Both options teach us."

That evening, Maya drove to meet Carla at a FreshWay store destined to be in the pilot. The store was small, functional, and warm with the smell of baking. Maya walked the aisles with Carla, explaining the product again, step by step, for the people who would use it. She used plain language and repeated the facts, estimates, and ambitions so everyone could separate them in their heads. "We will install cameras and small shelf sensors. We will not touch your cash registers. Our software will tally visible stock, compare it to expected levels, and flag items likely to run out. We will train your shelf staff to check flagged items and log corrections for our models. Fact: we require one week of calibration to learn your store layout. Estimate: you'll see fewer stock-outs in the second week; our experience suggests around a 20 to 30 percent drop in out-of-stocks for the covered items. Ambition: in three months, with iterative fixes, we aim for more."

A store manager named Noor, who had heard the pitch, asked, "Will this mean more work for my staff?"

"There will be some new checks at first," Maya admitted. "But the system is meant to make their jobs easier by pointing them to the problems before customers complain. If we do it right, they spend less time chasing missing items and more time helping customers."

Noor looked at the shelves and then at Maya with a small, honest smile. "Do it, then. But don't make our life harder. Prove it. We'll help you prove it."

Maya left the store feeling tired but not anxious. The pilot was agreed, the investor wanted to look more closely, and the team had a plan that used clear facts and conservative estimates. There was a risky road ahead—technical surprises, logistics slowdowns, the hard math of proving sales lift—but the company had taken the first practical steps. Back at their office, the three founders sat on the couch and watched the first email threads arrive: installation schedules, a Docusign, and a calendar invite for the first pilot planning call. They let themselves smile. It was not victory, but it was earned hope. They had work to do, and now there was a place to put that work where it could change everything.