The Hidden Cost of Unbillable Discovery

"We're not paying for discovery."

If you run a consultancy, you've heard that line, probably more than once this year. Usually from a procurement lead with their arms folded, followed by: "Show us you understand us first."

And most of us do the same thing. We say yes. We give it away.

I want to talk about what that really costs, and why it matters more now than it ever has, with the whole industry suddenly talking about outcome-based pricing.

What's actually happening

This isn't a LinkedIn trend. It's showing up in the numbers.

When Accenture reported its full-year results on 1 October, fourth-quarter revenue came in above the top of its own guidance. But two other things in that call caught my eye:

Fixed-price work, which includes outcome-based deals, is now more than 65% of its bookings and still growing.
AI productivity is going back to clients. Julie Sweet was open that Accenture is "definitely giving more productivity due to AI", and making up for it with new work and wider scope.

Two thirds of what the biggest consultancy in the world sells is no longer priced by the hour. Some of that is a fixed fee for agreed work, and some is tied to results. Either way, clients expect the savings from AI to show up in the fee.

It's not just Accenture. Around a quarter of McKinsey's global fees now reportedly come from outcome-based pricing, and late last year its UK managing partner talked publicly about doing more performance-based arrangements with clients. Its internal AI assistant, Lilli, is reported to save consultants up to 30% of their time on research and synthesis.

So here's the squeeze. If AI makes the work faster, billing by the hour means billing for less. Clients know it. Procurement knows it. Outcome pricing is coming whether we like it or not.

The question nobody is answering

The moment you propose an outcome-based fee, your client's CFO will ask one question: "How will we know it was you?"

If you can't answer that, outcome pricing isn't brave. It's a bet, and you're the one carrying the risk.

That isn't just my opinion. A new paper from Erik Brynjolfsson of Stanford's Digital Economy Lab and Georgios Petropoulos of USC looks at how AI is changing the way consulting gets priced. Their conclusion is simple: outcome pricing works when results can be measured and attributed. When they can't, a performance-linked contract just piles risk onto the consultant, without telling anyone anything useful.

In plain English: no measurement, no outcome fee. Or at least, not one you'd want to sign.

Where the answer should come from

The answer to the CFO's question is supposed to come from discovery.

Discovery is where you find out where the client really is. It sets the starting point: the "before" that every "after" gets measured against.

And for most firms, it's the part we do for free, in a rush, and without a score.

What free discovery really costs

Here are some simple numbers. Change them to fit your own firm.

A typical pursuit takes four senior people about three weeks: interviews, workshops, analysis, a findings deck.
Each of them spends around a third of their time on it. That's roughly 160 hours.
At a blended senior rate of £200 an hour, that's £32,000 of time on one pursuit.
If you win one in three, every engagement you win carries close to £100,000 of unbilled thinking.

Your finance team calls it cost of sale. And honestly? That's the politest name for it. What it really is, is the best thinking in your firm, handed over before anyone has agreed to pay for it.

And here's the part that hurts most. Free discovery usually produces something you can't use later.

Why free discovery breaks outcome pricing

When discovery is unpaid, it's rushed. When it's rushed, it's inconsistent. A few interviews here, a workshop there, notes in five different formats, a deck built at eleven at night.

What you end up with is an opinion. Often a good one, but still an opinion.

So six months later, when you want to show the client what changed, there's nothing solid to compare against. No score. No agreed starting point. No way to say, "This is where you were. This is where you are now."

That's the moment outcome pricing falls apart. Not in the pitch, but at the end, when you try to prove it.

What outcome pricing actually needs

If you're serious about pricing on outcomes, you need three things in place before the work starts.

A baseline on day one. A real score of where the client is today, agreed by them before the engagement begins. Not a slide: a number they've signed off.
Evidence from their own people. Not your opinion of them, and not one sponsor's view. Every stakeholder who matters, answering the same questions. When the CFO thinks they're a four out of five and the COO says two, that gap is gold. It's also exactly where outcome deals go wrong if nobody spots it early.
The same measurement at the end. Same questions, same scoring, same people. So the change is theirs to see, not yours to argue.

I'll be honest about the limits. A baseline won't prove that every bit of improvement came from you. Markets move. Other projects run alongside yours. The Stanford paper is clear on that too, and it's why most firms are likely to land on hybrid fees: a fixed core, with a share tied to agreed measures.

But without a baseline, you can't even have that conversation. With one, you're negotiating from evidence, and the CFO's question stops being a trap. It becomes a conversation about numbers you both agreed at the start.

Turn discovery into the thing they pay for

This is where it gets interesting. The work that sets the baseline is the same work you're currently giving away.

So stop giving it away. Turn discovery into a scored, benchmarked diagnostic:

Built on your own framework, not a generic template.
Sent to every stakeholder who matters, each answering in fifteen to twenty minutes.
Scored and benchmarked, then turned into a report in your firm's name.

Now discovery isn't a cost of winning the work. It's Phase 1: something the client pays for, because it tells them something they didn't know about themselves.

And it does a second job. It becomes the baseline your outcome fee is measured against. Run it again at the end with the same questions, and the change is right there in their own people's answers.

One piece of work. Paid for. And it protects your fee.

Why we built Score-AI

I spent years in enterprise marketing. In one role I built the region's first attribution model, because nobody would fund what we couldn't prove. Now, running my own consultancy, every client wants the same thing: put your money where your mouth is.

Fair enough. But you can't take credit for a change you never measured.

Score-AI turns your methodology into a branded assessment. Your client's stakeholders complete it through a simple link, with no logins, in fifteen to twenty minutes. You get a benchmarked report within the hour, with the gaps between stakeholders laid out clearly. When the work is done, you run it again and show the progress.

Baseline on day one. Evidence from their own people. The same measurement at the end.

Where to start

If outcome pricing is on your agenda, start with your own firm.

We've built a short assessment for consultancy leaders: the Outcome Pricing Readiness Scorecard. It shows how ready your firm really is to price on outcomes, where you're strong, where the risk is sitting, and how you compare with other firms.

Before you put your fee on the line, know where you stand. Take the Outcome Pricing Readiness Scorecard

Sources

Accenture (ACN) Q4 2026 earnings call transcript, 1 October 2026, via The Motley Fool
"AI is forcing McKinsey, BCG, Bain to rethink consulting fees", 26 May 2026, TheStreet
Erik Brynjolfsson and Georgios Petropoulos, "Pricing Consulting Services in the Age of Agentic AI: A Strategy Map for Executives", Stanford Digital Economy Lab, 23 August 2026 (PDF)