70% of the Way There Is Only Valuable If the 70% Is Right

Ninety source documents, according to the published case study, plus a 16,000-row CRM export. Three weeks. One 55-page competitive analysis, an executive summary, and a slide deck, delivered in time for a board meeting.

That was an engagement with Illumea Advisory earlier this year. Tiffany Haynes, Illumea's principal, had taken on a payments technology company that needed a defensible picture of where it stood in the market. Their pricing was out of step with the competitive set, their positioning was blurred, and the board meeting was already on the calendar. Tiffany had the relationships, the strategic judgment, and the trust of the executive team. What she didn't have was time to read, catalogue, and cross-reference the corpus the analysis required.

Her description of what that bought her is the most precise thing I've heard anyone say about this work:

Fieldway brought the report seventy percent of the way there, and the seventy percent was rigorous enough that I could trust it. That meant the thirty percent I added on top was the strategic work my client was actually paying me for, not a rewrite of someone else's draft.

Read that twice, because the load is carried by a clause most people skip: rigorous enough that I could trust it.

The objection every advisor is now getting

Here's the conversation happening in boutique advisory firms right now. A client, or a junior, or the principal themselves, points out that a competitive analysis can be generated in four minutes. So why is anyone paying for research?

It's a fair question and it deserves a real answer rather than a defensive one. The first draft of almost everything a junior or mid-level consultant used to produce is now effectively free. That's true. Pretending otherwise is how you lose an argument you should win.

The answer isn't that drafts don't matter. It's that a draft is only an asset above a quality threshold, and below that threshold it's a liability that looks like a head start.

Below the line, a draft costs more than a blank page

Think about what you actually do with a draft you don't trust.

You read it. Then you verify it, which means going to the sources anyway – the thing the draft was supposed to save you. Then you find the errors, and finding errors in confident prose is slower than writing from scratch, because plausible is much harder to catch than bad. Your reviewers were trained to spot mistakes, not to spot confidence. Then you rewrite it, and rewriting someone else's structure is harder than building your own, because you inherit their outline and their framing along with their words.

And at the end of all that you carry the risk of the ones you missed. That risk doesn't sit with the tool. It sits with the person whose name is on the deliverable, in front of a board.

So the accounting isn't "free draft versus expensive draft." It's: does this draft reduce my work, or does it relocate my work into verification and then hand me the liability? At 40% quality, a draft is worse than nothing. At 70% with a trustworthy provenance, it's transformative. The gap between those two states is not a matter of degree, and nothing about the output's surface tells you which one you have.

That's the whole question, and almost nobody is asking it. The market is arguing about whether drafts are valuable, which is like arguing about whether ingredients are valuable.

What the last thirty percent actually is

The thirty percent Tiffany added is not polish. It's the part that requires a person who knows this client, this industry, and what a board will actually do with an answer.

It's knowing which of the findings matter to this executive team and which are true but irrelevant. It's the judgment call about what to leave out, which is invisible in the deliverable and is most of the craft. It's the recommendation that the analysis supports but does not contain, and the willingness to sign it.

Eric Brown names the scarce version of this well: what's rare now is knowing where the model is confidently wrong enough to cost you six months if you believe it. That's a capability, not a step in a workflow, and it doesn't come from reading faster.

Consulting analysts have been circling the same conclusion from the build side: analysis is becoming abundant while judgement stays scarce, and buyers increasingly want to pay for the scarce thing. I'd sharpen that. Buyers want to pay for the scarce thing sitting on top of a base they can trust. Judgment applied to an unreliable foundation isn't premium work. It's a person laundering an error with their own reputation.

What actually makes a base trustworthy

Since the threshold is the question, it's worth being concrete about what puts a draft above it. Three things, from the Illumea engagement and every one like it.

Synthesis, not summarization. That report wasn't a stack of summaries stapled together. It was a new analysis built across the corpus: which findings reinforced each other, which contradicted each other, where the data was thin, and what the patterns implied for the client's decisions. A pile of summaries hands the synthesis work back to you, which is precisely the work you were trying to buy.

Traceability. Every claim goes back to a source you can open. Nobody checks all of them. You check some of them, and what you're really testing in those spot-checks is whether the whole thing deserves your trust. One fabricated citation and the correct move is to re-verify everything, which means the draft is now worth less than zero.

Known provenance. Somebody's name is attached to the method. That's the difference between output and work product, and it's the reason "a tool generated it" and "a calibrated operator produced it with tools" are different purchases even when the artifacts look similar.

The division of labour is the product

On that engagement the structure was the point. One operator, calibrated on how advisors think and what client-ready output looks like. Contracted under Illumea's own engagement rather than as a parallel client relationship. Scoped specifically for synthesis – not strategy, not relationship management, not delivery.

Tiffany stayed the face of the work and the author of the recommendations. Her client never needed to know or care how the reading got done, which is exactly right: they hired her for her judgment, and the judgment is what they got, with more of it than they'd have gotten if she'd spent three weeks reading.

There's a version of this that applies inside your own firm, too. If you've got a junior producing first drafts, the same threshold governs whether that's a genuine gain or just overhead – and the honest answer changes as they develop, which is uncomfortable to say out loud but is the actual basis on which you should be assigning the work.

That's the shape I'd look for in any research partner. Not "can you produce a deliverable," which everything can now. Whether the base they hand you is one you'd be willing to put your name on top of – and whether they'll tell you plainly when it isn't.

Seventy percent of the way there is a gift. Forty percent of the way there, delivered with confidence, is a bill you pay later.