Your Client Now Arrives With a Draft Answer. That Changes What They're Buying.

The first meeting has changed, and most advisory scoping hasn't caught up.

It used to open at zero. The client had a problem, some scattered internal opinion, and no structured view of it, and the first phase of the engagement was building one. That construction work was a meaningful share of what got billed, and everybody understood the arrangement.

Now a growing share of clients arrive with a point of view already in hand. Their team ran the research, generated the options, built the market map, and drafted a recommendation. It's coherent. It's well-formatted. It reads like something a competent analyst produced, and in a sense it is.

They're not hiring you to produce that anymore. They're hiring you to tell them whether it's right.

The shift, stated plainly

Olivier Delestre-Levai put it about as directly as it can be put in his March 2026 analysis of the consulting model: consultants "will be less frequently hired to create a first point of view and more frequently hired to validate, sharpen, stress-test, or operationalize one."

His explanation for why is the part worth sitting with. Capabilities that once required external support can increasingly be handled in-house – "early-stage research, option generation, meeting preparation, stakeholder messaging, and competitive monitoring." That moves the buy-versus-build boundary and raises the threshold for external spend. So the reasons a client reaches outside change too: not because they can't structure the problem, but for external credibility, cross-sector benchmarks, political air cover, specialist expertise, or acceleration when something is urgent.

I want to be careful not to overstate this. Nobody has measured what share of engagements now start this way, and I haven't seen a credible number. What I can say is that the ones that do start this way feel different from the first conversation onward, and the difference is not subtle.

A confident draft is harder to work with than no draft

This is the part that surprises people who assume a head start is a head start.

When a client has nothing, your first job is construction, and construction is a known quantity. When a client has a persuasive document, your first job is adversarial – and you're now doing it to a thing they are attached to, that their team produced, that reads well, and that has already shaped how they talk about the problem internally.

Three things make that harder than it sounds.

Plausible is much harder to catch than bad. Your reviewers were trained to spot mistakes, not to spot confidence. A weak argument announces itself; a well-structured argument built on one unexamined assumption does not, and the structure actively discourages you from looking underneath it.

The draft has already done political work. By the time you see it, it's been circulated. People have opinions keyed to it. Some of them have staked something on it. Disagreeing with the document is now disagreeing with colleagues, which is a different conversation than disagreeing with a blank page.

The framing is contagious. Whatever question that draft answered is now the question, and the most valuable thing you can sometimes do is point out it's the wrong one. That's much harder after a client has spent three weeks getting comfortable with it.

Delestre-Levai names the broader condition well: executives now face "more machine-generated recommendations, more scenarios, more synthetic benchmarks, and more persuasive but potentially flawed outputs." Persuasive and potentially flawed is a genuinely difficult combination to advise against.

What that means for scoping

If the engagement starts at step three, the shape has to change, and a scope written for a blank page will mis-price the work badly.

Verification is the opening move, not a late gate. If you've scoped validation as a review step near the end, you've put the most valuable work where it can do the least good. Front-load it. The first deliverable should be an assessment of the draft's foundations, not a competing document.

Price the assessment separately, and early. A short, sharply-scoped verification phase is easy for a client to say yes to and gives both sides real information about whether the larger engagement is needed at all. Sometimes the draft is sound and the honest answer is "this is right, go execute" – which is a better outcome for them and a better advertisement for you than three months of work neither of you needed.

Say what you'd have to see to agree. Naming your own falsification criteria before you start reading is how you avoid becoming a very expensive second opinion that agrees with whatever it was handed.

Protect the right to reframe. Build into the scope that the first finding might be "you're answering the wrong question." If the contract only pays for an opinion on the stated question, you've been hired to ratify the framing.

The questions I'd ask about the draft before reading it

Before I engage with the substance of a client's existing analysis, I want to know how it came to exist, because provenance tells you where to look for the weak joint.

Who decided the question? Most drafts answer the question they were given, competently. If one person framed it and nobody challenged the framing, the framing is the least-examined part of the document and usually the most consequential.

What sources did it actually read? Not "what does it cite" – what was in the corpus. A market map built from vendor marketing and analyst summaries will be internally consistent and systematically wrong about how buyers behave, and nothing in its tone will indicate that.

What did it leave out, and does anyone know? A draft's omissions are invisible by construction. Asking the team what they considered and set aside surfaces more than asking what they concluded.

Who has already agreed with it? This one isn't about the analysis at all. It tells you what it will cost, politically, if the finding is that the draft is wrong – and that cost is a real input into how you deliver the answer, though never into what the answer is.

What doesn't transfer

The layer that stays yours is the one Chris Hood identifies: the specific organizational pathology preventing a particular client from implementing what they already know they should do.

That's not a research finding and it can't be generated, because it isn't in any document. It comes from having watched the same failure in thirty different organizations, sitting in enough rooms to hear what isn't being said, and knowing which recommendation this leadership team will actually act on versus which one will be admired and shelved.

A draft can be right and still be unimplementable here, for reasons that have nothing to do with its analysis. Noticing that is the work.

The reframe worth making

I'd resist reading any of this as a downgrade. Stress-testing a convincing but possibly wrong answer is harder than producing a first one, and it's worth more, because the cost of acting on a plausible error is higher than the cost of a slow start.

The uncomfortable part is that it's less legible. A 55-page analysis looks like work. A two-page memo saying "your third assumption is wrong and here's what it costs you" looks like an afternoon, and it may well have been. Pricing that honestly, and getting clients to value it, is the actual business problem this shift creates – and it's the one worth solving now, while the shift is still something you can get ahead of.