What free Attio audits keep revealing

I run a free Attio audit. The offer is simple. You add me to your workspace as an Attio expert, which costs you no extra seat and no money. Over 48 hours I read the whole thing and send back a one-page teardown, ranked by impact, with the three fixes worth doing first and the exact setting to change. No call. No pitch.

I started it as a way to meet teams without a sales motion. It turned into the best research I have on how real companies actually use a CRM. After a run of them, across a capital-raising firm, a pay-equity SaaS, an insurance team, and others, the findings rhyme. Here is what keeps showing up.

The workspace is almost never the problem

I expected to find half-built workspaces. I mostly find the opposite.

The teams that ask for an audit have already done the work. Nine objects. A few thousand companies, a few thousand people, a real deal pipeline, custom fields that map to how the business runs. These are not beginner setups. Someone cared, sat down, and built something solid.

So the audit is rarely "your structure is wrong." That is the good news, and it is worth saying out loud, because most teams brace for it.

The data gets captured and then sits there

Here is the finding that repeats hardest. The workspace captures plenty. It just does not use most of it.

One team had recorded over 300 sales calls. Eight of them had been turned into a follow-up task or a note. The other 290-odd sat in the workspace as audio nobody would ever open again. The recording worked perfectly. The step that turns a call into a next action did not exist.

The same shape shows up everywhere. A discovery field set, pain and budget and decision-maker, filled on about 8% of deals. A whole block of fields built and then left empty. A pipeline value that cannot be calculated because the probability field is filled on a fifth of records.

The data is not missing because the team is lazy. It is missing because capturing it is manual, and manual steps lose to a busy week every time.

"Build me X" usually means "wire up the X you already have"

This changed how I scope work.

A founder will ask for auto follow-ups, post-close onboarding, and richer deal cards. It sounds like three things to build. When I go look, three of the four asks already half exist. The follow-up automation is there but switched off. The onboarding list is built but never wired to fire when a deal is won. The enrichment fields exist, just on the wrong object.

So the real job is rarely a from-scratch build. It is connecting things that were started and never finished. That is faster, cheaper, and a much easier yes. It also means a good audit saves the customer money before any project starts, which is the point.

Schema debt is real and it is quiet

Every mature workspace carries debt, the same way every codebase does.

The usual suspects: a select field and a free-text field that capture the same thing, so the data splits in two. A slug with a typo baked in that you now have to type wrong forever. A field duplicated on People. A dead import from a year ago, where 86% of the records are stuck in the first stage because nobody ever moved them and nobody ever will.

None of it is fatal. All of it adds friction, and friction is why the data above never gets filled. Cleaning it is unglamorous and high value, the CRM version of paying down interest.

Ownership rot piles on the founder

Look at the task list and the story is always the same. Most open tasks belong to one person, usually the founder. A chunk are overdue. A few have no deadline. A few have no owner at all. One teammate owns deals despite having left.

This is not a tooling problem. It is what happens when a team grows faster than its hygiene. But it shows up in the CRM first, and the CRM is where you can fix it with a few assignment rules instead of a meeting.

Latency is usually bloat, not Attio

Teams that complain about a slow workspace almost always blame the platform. The cause is usually their own object bloat: bulk API imports mixed in with live records, so an object that should hold a few thousand contacts holds a hundred thousand, most of them dead weight from a one-time load. Split the imported rows from the live ones and the speed comes back. The platform was fine.

Why the audit stays non-destructive

I never change anything during an audit. Not because my access is limited, it is not. As an Attio expert I have full access, the same as an admin. I choose not to write.

That choice matters for the kind of teams that take the offer. A regulated firm holding investor data, a SaaS company with a hard rule that EU salary data never enters the CRM. They are not handing the keys to someone who will start moving furniture. They get a read, a written teardown, and a clear list. What they do with it is theirs.

What I actually take away

The audit keeps teaching me the same lesson, and it is not the one I expected.

The hard part of a CRM is not setting it up. Plenty of teams clear that bar. The hard part is the loop after: making sure the thing you capture turns into an action, keeping the schema clean enough that people fill it, and not blaming the tool for problems that are really about hygiene.

That is also where AI earns its place. Not as a feature bolted on top, but as the thing that closes the loop the team keeps dropping. The call that becomes a task without anyone typing it. The won deal that starts onboarding on its own. The audit is how I find which loop is open. Closing it is the work.

If you want one, I have a few slots a week. You add me as an expert, no seat, and 48 hours later you get the teardown. Even if we never work together, the list is yours to keep.