A CFO at a venture fund described their fix for questionnaire fatigue like this: they wrote a self-DDQ — a single document answering everything an LP might ask, sent proactively to head off the individual requests.

This is a good idea. It is, in fact, the correct idea. Almost every fund that raises more than once arrives at some version of it, independently, because the arithmetic forces it. Here is how that CFO put the arithmetic:

10+ DDQs per fundraise
~80% of the questions are the same, with different phrasing
The worst single one — a large institutional LP's — can run one to two weeks

Eighty percent overlap means the marginal DDQ should be cheap. The self-DDQ is the obvious way to make it cheap. And for a while, it is.

Then it stops working, and it stops working in a specific way that's worth naming.

The document doesn't know when it's wrong

A self-DDQ is a snapshot. The moment you save it, it starts drifting from the fund.

Your AUM changes. A key person leaves. The auditor changes. A side letter gets signed that alters a fee term. Your valuation policy gets updated after an audit comment. The document doesn't know about any of it. It sits in Drive being confidently, fluently out of date.

And here's the part that makes it dangerous rather than merely inconvenient: nothing about a stale answer looks stale. A wrong AUM figure in a beautifully written paragraph reads exactly like a right one. The document gives you no signal. There's no red flag, no "last verified" marker on the specific claim, no link back to the source that would let you check in ten seconds.

So the CFO does the only thing available: re-reads the whole thing before every send. Which is most of the work the document was supposed to eliminate.

That's the trap. The self-DDQ shifts effort from writing to verifying, and verifying an undifferentiated wall of prose is slower than people expect — because you have to check everything, since you can't tell which parts changed.

The second failure: it's an artifact, not a library

The other thing that goes wrong is subtler.

A self-DDQ is written as a document — sections, headings, narrative flow. But a DDQ isn't a document, it's a query against your fund. An LP asks about your valuation policy for early-stage marks in a specific way, and your beautiful paragraph on valuation policy covers it, sort of, in the middle, mixed in with three other things.

So you copy, paste, and rewrite. Every time. Each rewrite is a small new piece of prose that now exists nowhere but in that one LP's questionnaire. The self-DDQ doesn't capture it. The next questionnaire asks the same thing in a fourth way and you write it a fourth time.

The work compounds in the wrong direction. You accumulate answers, but they scatter into sent questionnaires instead of collecting anywhere.

What actually makes overlap pay

The 80% is real and it is genuinely an asset. Capturing it needs three properties the document form can't provide:

1. Answers are addressable, not narrative. The unit is a question-and-answer pair, not a section. When a new questionnaire asks about valuation policy in unfamiliar phrasing, you want to match against the answer you've already approved for that question, however it was worded before — not scroll a document for the relevant paragraph.

2. Every answer carries its source. Not "our AUM is $X" but "our AUM is $X, per the Form ADV filed on this date, at this line." Then staleness is checkable in seconds instead of requiring a full re-read. When the underlying document changes, you know exactly which answers are implicated — because they point at it.

3. Approval is captured, and it's the thing that gets reused. The valuable artifact isn't the drafted answer. It's the answer a human read and approved, with the edits they made. That's the version that should come back next time. A document has no notion of approval state; every paragraph looks equally blessed.

Put those together and the 80% starts working for you rather than needing to be re-earned. Questionnaire eleven is mostly assembled from approved answers to questions one through ten, each one pointing at the document it came from, with the ones that are newly unsupported clearly marked as the places you need to actually think.

The honest part

None of this makes a DDQ zero-effort, and anyone telling you it does hasn't filled one out. The 20% that is genuinely new per questionnaire is often the hard 20% — the LP-specific ask, the thing about your strategy nobody has probed before, the question you don't want to answer in writing. That part is yours. It should be.

What's worth automating is the 80% that is your fund restating facts about itself that are already written down somewhere. That work is real, it's expensive, and it produces nothing but a correctly-transcribed fact.

The self-DDQ is the right instinct with the wrong container. Keep the instinct.

Try it against your hardest one

Fundry's free DDQ tool takes a real questionnaire and drafts an answer to every question, each with a confidence score. No account needed.

Use the questionnaire you dread. Then look specifically at the questions it didn't answer confidently. That set is the most useful output: it's a map of what your fund hasn't written down anywhere yet, which is the same list of things that will slow down the next raise.