"Most of it is the same questions" is the most common sentence in fund operations. It's also, at nearly every firm, an unmeasured belief. Nobody knows if their overlap is 60% or 90%, and the difference determines whether your DDQ process should take a week or an afternoon.

Here's how to measure it in about 30 minutes.

The exercise

Take the last three DDQs your firm completed. Different LPs, ideally different formats — one institutional monster, one fund-of-funds template, one bespoke list from a family office.

For each question in the newest DDQ, mark it one of three ways:

Direct reuse — a prior DDQ answered this question, and the answer is still true. Wording differences don't matter; substance does.
Reuse with refresh — the prior answer's structure holds but a number changed (AUM, team size, track record, this year's ESG language).
Net new — nothing in the prior two DDQs addresses it.

Your reuse rate is (1 + 2) over the total. Your automation ceiling is category 1; your update burden is category 2; and category 3 is the only part that was ever going to need real writing.

What the three buckets tell you

If category 1 dominates, your bottleneck is retrieval, not writing. The answers exist; your team's time goes to finding, reformatting, and re-approving them. That's a filing problem wearing a writing costume — and it's the specific problem a knowledge base of approved, citable answers eliminates.

If category 2 is large, your bottleneck is freshness. The dangerous version of this bucket is silent: an answer that was true at final close and is subtly wrong now. Stale AUM is embarrassing; a stale concentration limit or an outdated valuation policy is an LP-trust problem. Category-2 answers are why "copy from the last DDQ" is riskier than it feels — you're reusing the structure and the staleness.

If category 3 is large, look at why. Sometimes it's a genuinely novel LP. More often it's a documentation gap: the question is standard (cyber posture, BCP, valuation committee mechanics) and your firm simply hasn't written the canonical answer yet. A big category 3 before a fundraise is a to-do list, and finding it now — instead of during live diligence with a $10M commitment waiting — is the entire point of the exercise.

The part that compounds

The exercise is diagnostic. The fix is structural: every approved answer should land somewhere it can be matched against the next DDQ's wording, with a citation to the source document and a record of who approved it and when.

Do that, and the three buckets shift every quarter — category 1 grows, category 3 shrinks, and category 2 becomes a maintenance list instead of a surprise. Don't, and every fundraise starts the count from zero again.

That accumulation is what our DDQ agent is built around: it matches new questions against your approved answers first, drafts only what's missing, and returns everything with citations and a confidence score — green answers you skim, red answers that tell you exactly where your document library is thin. Nothing goes to an LP without your approval.

Run the 30-minute exercise either way. The number you get is the business case — for or against — better than anything a vendor can tell you.