How do you read the say/do gap?

The say/do gap is the distance between what management promised on past earnings calls — the timelines, the capacity, the margins, the capex — and what they actually delivered against those promises later. It is the difference between the say-side and the do-side of the same team, measured over years. When the gap is small and consistent, guidance is worth taking at close to face value. When it is wide and repeated, the next confident forecast should be discounted before you build anything on it. Almost every input to credibility flows from this one reconciliation.

Why is the transcript archive an underused credibility ledger?

Because most readers treat each earnings call as a fresh event and forget the last one, when the real value is in the archive read end to end. A single call is a sales pitch — management is confident, articulate, and forward-looking by design. But three years of calls, read together, are a ledger: every quarter management made specific, dated claims about what would happen, and every subsequent quarter is a mark-to-market on those claims. The transcripts are public, timestamped, and free, and almost nobody reconciles them against each other. That is the underused edge — not the newest guidance, but the record of how the last twelve guidances turned out. Read that way, the archive stops being commentary and becomes a track record you can score.

How do you reconcile prior guidance against outcomes?

You extract each concrete commitment from the older calls, then check whether it was actually met. The method is mechanical and that is its strength. Go back several years of transcripts and pull out every specific, dated statement — “we will commission the plant by the second half of next year,” “we expect margins in the high teens once the new line ramps,” “capex of this much is approved for this project.” Ignore the vague mood music; keep only what can be checked. Then walk forward through the later calls, the announcements, and the financial statements and mark each one: delivered on time, delivered late, missed, quietly abandoned, or silently redefined. What you build is a running scorecard of promise versus outcome. The pattern across it — not any single hit or miss — is the credibility read, and it lets you weight the next promise by the team’s demonstrated hit rate rather than by how convincing they sound today.

What are the red flags?

The tells of a wide say/do gap are consistent across companies once you know to look for them.

• Moving goalposts. A target whose date or size is quietly revised almost every quarter — the capacity that was “next year” for three years running, the margin that is always “two quarters out.” The number never lands; it just keeps sliding forward.
• Vanishing timelines. A commitment that simply stops being mentioned once it comes due. Nobody announces a miss; the project just disappears from the narrative and hopes you weren’t keeping the ledger.
• Announced but never funded. A marquee expansion or new plant unveiled with a press release and a big number that never shows up as a sanctioned project, a rating action, an order, or actual cash going out the door. Announcement is cheap; funding is the commitment.

Any one of these, once, can be bad luck. A pattern of them is the company telling you its guidance is aspiration, not plan.

What are the green flags?

The trustworthy signature is the mirror image: specific, dated commitments that later get confirmed by an independent source. A credible team says something concrete — a date, a number, a named project — and then, quarters later, the transcript, the regulatory filing, the rating rationale, or the cash-flow statement confirms it happened roughly as described. The commitments are precise enough to be falsifiable, and they survive the check. Over time this produces a management that under-promises at the edges and delivers in the middle, whose guidance you can lean on because the archive shows they have earned it. The best version is a full triangulation chain — a plan stated on a call, then approved, then funded, then permitted, then showing up in protected, higher-value revenue — each stage confirmed by a different source than the management mouth that first claimed it.

How does Esploro turn this into the largest single weight?

We make the say/do reconciliation the heaviest input in the entire framework, because a company’s credibility is what tells you whether to believe everything else it says. In our method the transcript archive is worked into a credibility ledger: management’s past promises are lined up against delivered outcomes, quote by dated quote, so that future guidance is weighted by track record rather than taken at face value. Credibility carries the single largest weight of any dimension we score — more than the financials, more than any one catalyst — because a migration story from a team with a clean say/do record is a different proposition from the identical story told by a team that has missed for years. The financials are public and lagging; everyone has them. The credibility read, built from primary quotes and traced to the exact dated source, is the part that is genuinely hard to assemble and therefore the part that is worth the most.

For how this reconciliation sits inside the wider framework — the Analysis Bridge and the triangulation chain from stated to protected — see the method.