How do I present requirements findings to a client?

Present requirements findings as a diagnosis of the backlog, never the people who wrote it: lead with aggregates, rest every finding on a stated standard, keep authors anonymous, and pair every problem with a rewritten example. The moment a finding reads as “your team writes bad stories,” the room stops evaluating the evidence and starts defending itself, and the engagement stalls there.

This matters more for coaches, consultants, and trainers than for anyone else: you were hired to change how the team works, the product owner who wrote the backlog is in the room, and everything you say is heard as an outsider’s opinion until something neutral backs it. Here’s how to deliver hard findings, keep your authority, and still get the follow-on engagement.

Why do findings land badly?

Because a backlog is authored, and criticism of an artifact reads as criticism of its authors unless you actively separate the two. The defensive reaction isn’t irrational: a story that looks vague to you was written by someone under deadline pressure who knew the missing context by heart, and their team has shipped from stories like it for years. Your job in the readout is to make the gap visible without making the author wrong. The framing that works: the story slipped through, the process let it, and here’s what catches it next time.

Why lead with aggregates?

Aggregates are impersonal, and impersonal is what lets a sponsor agree with you in front of their own team. “41% of the backlog scores below the readiness bar, and testability is the weakest dimension” is a statement about a dataset; “look at this story” is a statement about whoever wrote it. Open with the quality posture, the per-dimension breakdown, the score distribution, and the hotspots by epic or team. By the time an individual story appears on screen, it should be obviously an illustration of a pattern the room has already accepted.

Should authors ever be named?

Not in anything the client sees. Anonymize by default (“Author A”), even when the client asks, because the moment a name appears the readout becomes a performance review and you become the coach who got someone in trouble. A quality-by-author view is genuinely useful to an engineering leader deciding where one-on-one coaching helps; deliver it privately to the sponsor if asked, never in the room. The same caution applies to ticket keys on sensitive backlogs: hide them if excerpts could embarrass.

How do I structure the readout?

• One-page executive summary: the posture, the top three findings, and what you recommend. Write it yourself; it’s the only page the sponsor is guaranteed to read.
• The aggregate diagnosis: distribution, dimensions, hotspots. This is where the evidence lives.
• Exemplars with rewrites: a capped set of low-scoring stories, each paired with the improved version. The rewrite is what turns criticism into coaching, and it proves the fix is a paragraph, not a program. In a training engagement, the rewrites become the workshop exercises.
• What good looks like: at least one strong story from their own backlog. It sets the bar with their own material and gives the team an internal example to copy.
• Next steps: the specific epics or teams to work with first, the coaching or workshop plan, and the re-score that will prove the movement.

Keep the full per-story dataset in an attached spreadsheet, never in the deck. Completeness belongs in the appendix; the readout is for the decision.

How do I handle pushback in the room?

Agree with the specific and return to the distribution. Someone will find one story the scoring judged too harshly; concede it cheerfully, because your claim was never that every score is perfect. Your claim is the pattern: 41% below the bar doesn’t move because one story was miscategorized. Explicit criteria are your authority here. When the finding traces to a stated, documented standard (the six INVEST dimensions) applied by a neutral engine, the argument is with the criteria, not with your taste, and criteria don’t get defensive. This is the difference between “I’ve seen a lot of backlogs and this one is weak” and “this one scores here against the standard we agreed.”

How do I close on the follow-on?

End on the re-score. The assessment found the gap; the engagement closes it and measures the closing: the team works the flagged stories with you, you score a second round after the coaching or workshop phase, and the round-over-round improvement becomes the first slide of the next readout. A Vindex report adds that improvement section automatically once a project has two scoring rounds, under your brand; how a project works shows the full engagement loop.