The Discovery Sprint That Pays For Itself: Turning Free Scoping Into a Billable Phase

Every agency we know has the same leak: weeks of unpaid scoping work that either dies in procurement or turns into a fixed-bid landmine. We spent a long time pretending that was just the cost of doing business. It isn't. Discovery is the most valuable thing we sell, and giving it away is what makes the rest of the engagement go sideways.

This is the playbook we use to sell discovery as a standalone, billable phase — what it costs, what's in the box, how to pitch it without losing the deal, and what to do when the client tries to skip it.

Why Free Scoping Is a Margin Killer

The traditional agency motion looks like this: intro call, demo, a couple of "workshops" disguised as sales calls, then someone on your side spends three to five days writing a 40-page proposal with an estimate, architecture sketch, and timeline. You send it. The client ghosts, negotiates, or signs — and if they sign, you're now bound to numbers you produced with half the information.

The problems compound:

• You're guessing. No one shows you the messy CRM, the legacy auth system, or the stakeholder who hates the project. So estimates are padded or wrong.
• You can't say no gracefully. After investing a week of senior time, walking away feels like a sunk-cost failure, so you take projects you shouldn't.
• The client treats the SOW as a contract, not a hypothesis. Every later discovery becomes a "change request" fight.
• Your best people are doing pre-sales instead of delivery. The same architects who should be billing $200+/hour are writing speculative Notion docs.

In our experience, agencies that move to paid discovery see two things flip almost immediately: close rates on discovery proposals are much higher than close rates on full-build proposals, and the build proposals that come out of discovery close at roughly double the rate of cold ones.

What a Paid Discovery Sprint Actually Is

A discovery sprint is a fixed-scope, fixed-price engagement — typically one to three weeks — that produces a decision-grade artifact: enough technical and product clarity for both sides to commit to a build (or to walk away cleanly).

It is not a workshop. It is not a free reskin of your sales process. It has deliverables, a team, a price, and a contract.

Here's the structure we use for a standard two-week sprint:

Week 1 — Map the territory

• Stakeholder interviews (3–6 sessions, 45 min each)
• Access to existing systems: codebase walkthrough, analytics, support tickets, anything that reveals reality vs. the pitch deck
• Competitor and integration audit
• A first-cut user journey or service blueprint

Week 2 — Build the decision pack

• Prioritized scope (MoSCoW or RICE — pick one and stick with it)
• Architecture sketch with explicit assumptions and risks
• Two or three delivery options at different price points
• A go/no-go recommendation, written plainly
• A draft SOW for the build phase, ready to sign

The deliverable is the decision pack — not a Figma file, not a Jira backlog, not vibes. Something a CFO can read in 30 minutes and approve.

How to Price It Without Scaring Anyone

The number we land on for most mid-market clients sits in the range of 3% to 8% of the projected build cost, capped at a fixed fee. For a £150k build, that's a £6k–£12k sprint. Enterprise work runs higher, both because the stakeholder map is bigger and because procurement expects it.

Three pricing rules we don't break:

1. Fixed fee, never T&M. The whole point is to give the client certainty during the phase where they have the least of it.
2. Credit it against the build, partially. We commonly credit 50% of the discovery fee against the first invoice of the build phase if they proceed within 60 days. This removes the "why am I paying twice" objection without giving the work away.
3. No discounts for skipping deliverables. If they don't want the architecture review, the price doesn't change — you just have less to do. This protects the perceived value of every component.

Selling It to a Client Who Expected a Free Proposal

This is the part most agencies get wrong. You can't just announce "we charge for discovery now" and expect the pipeline to behave. The pitch matters.

The framing that works for us:

"We've stopped giving fixed quotes off a 60-minute call because they're wrong about a third of the time, and that's bad for both of us. Our discovery sprint gives you a real number, a real plan, and the option to take it to another agency if you want a second opinion. You own the output."

Three things to notice:

• "You own the output." This is non-negotiable. If the client pays, they get the artifacts — including the architecture and the SOW. Yes, they can technically shop it. They almost never do, because by the end of the sprint, you've built the relationship that wins the build.
• "Real number, real plan." You're selling certainty, not hours.
• The exit is built in. Clients commit more easily when the door is visibly open.

If a prospect refuses to pay for discovery, that's signal. Not always a no — sometimes it means the budget is smaller than they admitted, sometimes it means they're shopping five agencies. Either way, you've learned something in 20 minutes that used to cost you a week.

A Lightweight Template for the Decision Pack

We keep the structure boring on purpose. Every sprint produces the same skeleton, which means our team can move fast and clients know what they're getting. Here's the rough outline we work from:

# Decision Pack — [Client] [Project]

## 1. Problem statement (1 page)
What we heard. What we observed. Where those disagree.

## 2. Scope recommendation
- Must-have (MVP)
- Should-have (Phase 2)
- Won't-have (this engagement)

## 3. Architecture sketch
- System diagram
- Key decisions + alternatives considered
- Explicit assumptions
- Top 5 risks with mitigations

## 4. Delivery options
| Option | Scope | Timeline | Price band | Team |
| ------ | ----- | -------- | ---------- | ---- |
| A      | MVP   | 10 wks   | £X–£Y      | 3 FTE |
| B      | MVP+  | 14 wks   | £X–£Y      | 4 FTE |

## 5. Recommendation
One paragraph. Plain English. No hedging.

## 6. Draft SOW
Ready to sign within 14 days.

Keep it under 25 pages. If it's longer, you're hiding behind detail.

A note on the architecture section

This is where senior engineers earn their keep. The sketch doesn't need to be a final design — it needs to be honest about what you don't know yet and what would change the answer. A risk like "if the existing auth system can't issue OIDC tokens, add 3–4 weeks" is worth more to the client than any pretty diagram.

What Changes Inside the Agency

Moving to paid discovery isn't just a sales change. A few things have to shift internally:

• Staffing. Your best architect or principal engineer leads discovery, not a junior PM. The ratio we aim for: one senior tech lead, one product person, one designer, with a delivery lead floating across multiple sprints.
• Calendar. Discovery sprints need to start on Mondays and end on Fridays. Don't let them drift; the urgency is the product.
• Metrics. Track discovery → build conversion separately from cold → build conversion. If discovery converts below ~60%, something is wrong with either your qualification or your output.
• Sales comp. If your salespeople are paid only on build revenue, they'll bypass discovery to chase bigger deals. Pay a smaller commission on the discovery itself, and a larger one on the conversion.

We've also found it helps to have a written internal rule: no fixed-bid build proposal goes out without a discovery sprint preceding it. The exceptions are small (under £25k) or repeat clients. Everything else gets the sprint or gets a polite no.

When Not to Do This

Paid discovery isn't a universal answer. Skip it when:

• The work is genuinely small and well-defined (a Shopify theme, a single integration).
• It's a repeat client whose systems you already know inside out.
• You're augmenting a team on T&M and the client owns the roadmap.

Forcing discovery onto these engagements just adds friction. The point is to use it where it actually de-risks the build — which is most net-new product work for clients you haven't worked with before.

Where we'd start

If you're running an agency and still doing free scoping, pick the next three qualified leads and pitch them a paid discovery sprint instead of a proposal. Price it at the low end of the range. Use the decision-pack template above. Track what happens to close rate, build margin, and change-request volume over the following quarter.

If you want to see how we structure this for product builds specifically, our product engineering services page outlines where discovery sits in the wider engagement. And if you're earlier in the agency-to-product arc, the internal spin-out playbook on our blog is the natural next read.

The shift is uncomfortable for about a quarter. After that, you stop missing the free proposals.