This case study is a fictional composite constructed from patterns observed across multiple expert research programs. It does not describe any single firm, transaction, or investment. All figures are illustrative.
A mid-market buyout firm was running a 14-call expert program on a healthcare distribution target. The program had been carefully scoped: former operators, channel experts, procurement leads from large health systems. By any measure, it looked like a serious research effort.
The program ran over six weeks. The investment committee received a polished 12-page research memo. The deal closed.
Nine of the 14 transcripts were never formally synthesized. The thesis broke within 18 months.
The Program
The 14 calls were scheduled across six weeks and covered a wide range of expertise: former executives from companies operating in the target sector, distributor-side operators who understood the channel mechanics, and healthcare system procurement leads who sat on the buy side of the transactions the target depended on.
The calls were assigned to three junior analysts. Their role was to attend calls, take notes, and upload transcripts to the firm's shared research folder. All three did exactly that. The transcripts were uploaded on schedule.
What the program did not have was a claim extraction protocol — no structured process requiring analysts to pull discrete, attributable claims from each transcript within a defined window. There was no rolling thesis document being updated as calls came in. There was no contradiction log to capture cases where expert perspectives conflicted with management's narrative. Each transcript went into a folder and stayed there.
The 12-page research memo that went to the investment committee was built from five calls: the five that the lead analyst had personally attended and reviewed. The remaining nine were not summarized, not cross-referenced, and not represented in the memo.
“We had transcripts from experts who had worked directly at the target and its two largest competitors. They were uploaded and never opened. We found out 18 months later that two of them contained specific warnings about channel dynamics that proved prescient.”
The program, on the surface, looked complete. Fourteen calls is a serious expert research effort for a mid-market deal. The transcripts existed. The analysts had done their part. The failure was structural, not individual.
What the Unread Transcripts Contained
When the investment thesis began to break down — when the revenue ramp lagged projections and the margin expansion story stalled — the research team went back to the original expert call archive. What they found was not ambiguous.
Transcript #7 had never been read. In it, a former regional sales director for a sector-adjacent distributor described a structural shift in distributor incentives that was already underway at the time of the call — a shift that directly undermined the revenue thesis the investment committee had approved. The dynamic was not speculative. It was described in operational detail, with specific reference to how compensation structures were changing and why smaller distributors were beginning to rationalize their channel relationships.
Transcript #11 had never been read. In it, a former operations executive described a dependency on a single distribution partner that represented approximately 34% of the target's addressable channel. This concentration was not disclosed in management presentations. The expert had flagged it specifically, noting that the dependency created pricing leverage that the target could not easily renegotiate.
Transcript #13 had never been read. In it, a former executive from a direct competitor described a pricing dynamic in the sector that contradicted the margin expansion thesis. The expert's framing was direct: the pricing environment had been compressing for two years and there was no structural reason to expect a reversal.
None of these were subtle signals buried in hedged language. They were direct, specific, attributable claims from operators with firsthand sector knowledge. They were exactly the kind of signal an expert research program is designed to surface.
The irony is that the firm ran the calls. They paid for the access. They had experts on the phone who knew exactly what was happening in the target's channel. The intelligence existed. It simply was never read.
The Cost Calculation
The expert program cost approximately $28,000. At a loaded cost of roughly $2,000 per call — covering expert fees, moderator time, scheduling coordination, and platform costs — 14 calls represents a meaningful but entirely routine research budget for a mid-market deal.
The cost of fully processing all 14 transcripts — the synthesis gap — was estimated in retrospect at approximately 40 analyst hours. At a blended rate of $150 per hour, that represents $6,000 in additional analytical investment that was never made.
The math is stark. The firm spent $28,000 acquiring research and declined — not intentionally, but structurally — to spend an additional $6,000 to extract the full value of what they had bought. The total additional investment required to complete the research process they had already started: 21 cents on every dollar they had already spent.
Against the investment impairment that resulted from the unresolved thesis risks — risks that were documented in transcripts that were never opened — the economics of the synthesis gap are almost comically favorable in retrospect.
“The economics of expert research are almost always positive. The money isn't lost on the calls. It's lost on the gap between running the calls and actually using what you learned.”
— Research director, large-cap PE fundThis pattern — substantial spend on call acquisition, minimal spend on synthesis — is not unique to this composite case. It reflects a structural tendency in how research programs are resourced: expert calls are line items with clear vendor costs, while synthesis hours are discretionary and often deprioritized when deal timelines compress.
What Should Have Happened
The failure here was not a failure of effort. The analysts attended the calls. The transcripts were uploaded. The memo was written. The failure was the absence of structural requirements that would have made it impossible to complete the research process without processing all 14 transcripts.
Four structural changes would have prevented the outcome.
First: claim extraction required within 48 hours of each call. Not a summary. Not a memo. A structured extraction of discrete, attributable claims — each tagged to the expert's background, each assessed for confidence level, each categorized as thesis-supporting, thesis-challenging, or novel signal. This requirement cannot be waived for any transcript.
Second: a rolling thesis document updated after each call. By call #7 — the call from the former regional sales director describing the structural shift in distributor incentives — the channel dynamics risk would have been formally logged. The investment committee would have reviewed a thesis document that included that risk, not a memo that omitted it.
Third: a synthesis session before the investment committee, not just a memo. A session in which the lead analyst walks through the full claim set — including conflicting signals — and the investment committee can interrogate the research, not just consume a curated summary.
Fourth: lead analyst accountability for all transcripts in the program, not just the ones they personally attended. If a lead analyst's name is on the research memo, their accountability extends to every transcript in the program — including those attended by junior analysts. This is not punitive; it is structural. It makes it impossible for transcripts to exist in a folder without anyone owning their synthesis.
None of these changes are expensive. None require new technology. They require a documented research protocol and the discipline to enforce it under deal timeline pressure — which is precisely when teams are most tempted to let synthesis slip.
The case described here is composite, but the pattern it represents is not rare. Research teams across mid-market and large-cap PE run expert programs that generate more signal than the synthesis process can absorb under timeline pressure. Transcripts accumulate. Memos get written from the calls that happened to be reviewed first.
The most common version of this story is not negligence. It is a research program that generated genuine signal — signal that experts with direct sector knowledge were willing to provide — but that did not have the infrastructure to hear it.
The fix is not more calls. Running more expert calls does not solve the synthesis gap; it widens it. The fix is a research process that makes it structurally impossible to leave transcripts unread — where claim extraction is a requirement, not a best practice, and where the investment committee reviews a complete claim set rather than a curated memo that reflects only the transcripts someone happened to read.
The $6,000 synthesis gap in this case was not a budget problem. It was a protocol problem. And protocol problems, unlike budget problems, are entirely within a firm's control to fix.