The Pilot-to-Partnership Pipeline: How Agencies Convert Small Engagements Into Six-Figure Accounts

Every agency owner I know has a graveyard of one-and-done pilots. A client paid for a discovery, a prototype, a small build — then ghosted, or worse, took the artefacts in-house. The fix isn't a better proposal template. It's treating the pilot as the first step of a deliberate pipeline, not a standalone deal.

This is the system we use to turn a £6k–£15k first engagement into a multi-year account. It's not magic, and it doesn't work every time. But the hit rate is high enough that we now scope pilots specifically to feed it.

Why most pilots die

A pilot dies for one of three reasons, and they're almost never about the work quality.

The first is no narrative bridge. You shipped the thing. It works. Nobody on the client side knows what comes next, because you never told them. The relationship had no second act written into it.

The second is wrong stakeholder. You sold to a manager with a budget for one project, not to someone with a roadmap. When the pilot ends, so does their authority over you.

The third is commodity framing. The pilot was scoped as a deliverable — "build us a dashboard" — not as a hypothesis test. Once the deliverable lands, the engagement is logically over.

Fix those three things upstream and your conversion rate climbs without any extra sales effort.

Scope the pilot to reveal the next three problems

The pilot is not the product. The pilot is a flashlight you're handing the client so they can see further into their own backlog.

When we scope a pilot now, we add a line in the SOW that reads roughly:

Alongside the deliverable, we will produce a written assessment of the next 3–5 technical or product risks we observed, with rough effort bands.

That single clause changes the dynamic. It tells the client we're not just contractors filling a ticket; we're going to leave them smarter than we found them. And it gives us a legitimate, expected artefact at the end of the engagement that is itself a sales document.

What goes in the assessment

Keep it short — two pages, maximum. For each risk or opportunity:

• A one-paragraph description in business language
• The technical reason it matters (one paragraph, plain prose)
• A rough effort band (e.g. "2–4 weeks", "1–2 quarters")
• Whether it blocks, accelerates, or is independent of their current roadmap

No prices. No proposals. Just observations. The proposal comes later, and it comes because they asked.

Price the pilot to qualify, not to maximise

There's a temptation to discount pilots heavily to win them. Don't. A pilot that's too cheap attracts the wrong clients — ones who can't afford the partnership you actually want.

We price pilots at roughly 8–15% of what we'd expect a full first-year engagement to be. So if a typical good account is £120k/year, the pilot sits around £10k–£18k. That's high enough to filter out tyre-kickers, low enough to be an easy yes for someone with real budget authority.

We also fix the price. Not capped T&M, not retainers — a fixed fee for a defined deliverable plus the assessment. Founders especially appreciate the clarity, and it forces us to scope tightly. If we got the scope wrong, that's our problem and our lesson.

Design the hand-off, don't just hope for it

The last week of a pilot is where most agencies lose the deal. The team is wrapping up tickets, writing docs, and emotionally moving on to the next project. Meanwhile the client is realising the relationship is ending and starting to think about what's next — alone.

We now build a hand-off ritual into every pilot. It has three parts.

The walk-through

A 60-minute session with the client's senior stakeholder (not just the day-to-day contact) where we walk through what was built, what we learned, and the assessment document. This is not a demo. The deliverable demo happened earlier. This is a strategy conversation, and we treat it like one.

The decision memo

After the walk-through, we send a short written memo summarising the three to five options the client now has. Something like:

Option A: Stop here. You have a working X. We hand over
  source, docs, and a one-month bug-fix window.

Option B: Address risk #2 from the assessment (auth
  rework). Estimated 4–6 weeks, fixed fee.

Option C: Move to a quarterly engagement covering risks
  #1–#3 and ongoing roadmap support. Capped T&M.

We're happy with any of these. Here's how each plays out
in practice.

Notice what's missing: pressure. We genuinely mean "we're happy with any of these." Option A has to be a real, dignified exit, or the rest of the offer feels coercive.

The 30-day check-in

If the client picks Option A, we still book a 30-day check-in. No agenda, no sales pitch — just a call to see how the thing is performing in the wild. About a third of our biggest accounts started with someone who picked Option A, then came back six weeks later because something in the assessment had started to bite.

Stage the commercial relationship

Clients who go from a £10k pilot straight to a £150k annual contract are rare and usually a sign that you under-scoped the pilot. The healthier path is staged.

A pattern that's worked well for us:

1. Pilot — fixed fee, 4–8 weeks, one team of 2–3 people
2. Follow-on project — fixed or capped T&M, 8–16 weeks, addressing one risk from the assessment
3. Quarterly engagement — capped T&M with a defined team and roadmap, renewed every quarter
4. Embedded partnership — multi-quarter commitment, sometimes with a dedicated pod

Each stage is an off-ramp. The client can stop at any point without drama. That optionality is precisely what makes them comfortable continuing.

If you want to see how the pricing models behind each stage actually shake out, we've written about that elsewhere on the blog.

Sell to someone with a roadmap, not a budget

The single biggest predictor of pilot-to-partnership conversion, in our experience, is who signed the pilot. If it's a project manager with a one-line budget, you'll get exactly one project. If it's a head of product, a CTO, or a founder, you have someone whose job is to think about what's next.

During the sales process, we now ask explicitly: "Who owns the roadmap this work sits on?" If that person isn't in at least one conversation before contract, we either find a way to get them in or we adjust our expectations about what the pilot can lead to.

This isn't about being snobby. Small budget-holders are fine clients for small projects. But if you're trying to build an agency on long-term accounts, you need to be in the room with the people who plan in years, not sprints.

Track the pipeline like it's a pipeline

Most agencies track sales pipeline. Very few track expansion pipeline. We keep a simple sheet for every active and recently-closed pilot with:

• Which stage they're at (pilot → follow-on → quarterly → embedded)
• Date of next decision point
• Named senior stakeholder
• Open items from the assessment that haven't been addressed yet
• Probability of expansion, updated monthly

It's not a CRM feature, it's a discipline. Reviewing it once a month in a 30-minute leadership meeting has done more for our revenue stability than any new-business effort.

Where we'd start

If you're running an agency and your pilots aren't converting, don't rewrite your proposal templates. Do this instead, in order:

1. Add the assessment clause to your next SOW. Just that one line.
2. Block out the hand-off walk-through and the 30-day check-in before the pilot even starts. Put them on calendars.
3. Before signing the next pilot, find out who owns the roadmap. If you can't get them on a call, lower your expectations for that account and move on.

The rest — the staging, the decision memos, the pipeline tracking — can come later. The first three changes are nearly free and will start shifting your conversion rate within a quarter. If you'd like to talk through how this maps to your own engagements, our services page is the easiest place to start a conversation.