What should a CRM audit find?

Most audits I've seen end as a long slide deck. Every slide is true. None of them say what to do on Monday.

That's the trap. Any audit will find problems, because every CRM has them. The value isn't the list. It's knowing which three problems are costing you the most money, and fixing those first.

If you want to know how long a good audit should take, we covered that in how long a CRM audit should take. This page is the field guide: what the audit should look for, what it usually finds, and how to decide what to fix first.

What are the main things a CRM audit should look for?

A CRM audit should look in seven places: the data, the pipeline, lead routing, reports, the sales process, what you pay for, and who has access. Most audits only check the first two. The expensive problems often hide in the last three.

1. Data. Duplicates, missing owners, contacts with no company, wrong industries.
2. Pipeline. Deals with no next step, deals past their close date, too many pipelines.
3. Routing. How new leads reach a rep, how fast, and what happens when the owner is out.
4. Reports. Which numbers leadership trusts, and why the others are wrong.
5. Process. Whether reps follow the playbook the CRM was built for.
6. Cost. Seats, contacts and add-ons you pay for and don't use.
7. Access. Who can export, delete or change settings.

What does an audit usually find in the data?

Data problems are the most common finding: duplicate companies, contacts nobody can reach, and fields filled with guesses. They matter because every other part of the CRM sits on top of the data. Routing, reports and AI tools are only as good as the records underneath them.

Some real examples:

• Duplicates at scale. An IT services company had 18,000 duplicate companies. Reps were working the same account without knowing it.
• Contacts nobody trusts. At Worldwide Logistics (WWL), reps estimated their contact data was only 50 to 60% accurate and spent about 30% of their time checking contacts before calling.
• The wrong market. At an Esko business unit selling to consumer brands, 53% of leads came from companies under $50M, while the target customer was $50M to $1B. That's not a lead problem. It's a data and targeting problem.
• Billing contacts. WWL found that whether the billing contact was right was a coin flip. A wrong billing contact can mean months of invoices going to the wrong person.

If your data needs enrichment, not just cleanup, our answer on what revenue potential is and how reps use it shows what good account data makes possible.

What does an audit usually find in the pipeline and process?

Pipeline findings are usually deals with no next step, too many pipelines and stages, and a sales process that reps quietly skip. These problems make the forecast unreliable. Leadership starts asking for spreadsheets, which tells you the CRM has already lost their trust.

• No next step. At Terraboost, 77 of 82 open deals had no next step. You can't forecast deals nobody is working.
• Too many pipelines. Terraboost also had about 20 pipelines. We simplified them to 3: Local, National and Customer Success. Fewer pipelines meant reps could keep them current.
• A playbook nobody runs. At the IT services company, the playbook called for 5 to 6 touches. Reps skipped most of them, and one leader was still qualifying every lead by hand. Lead-to-opportunity conversion sat at 10 to 13%, with 12.5% of opportunities closing.
• Work outside the CRM. At another company, contracts were requested in Slack, so the CRM never knew a deal was close.

How the pipeline is set up matters as much as how it's used. If territories feel unfair, reps work around the CRM. Our answer on designing sales territories that scale covers that side.

What does an audit find in reports, cost and access?

Report findings are numbers that disagree or mislead. Cost findings are seats, contacts and add-ons you pay for and don't use. Access findings are too many people who can export or change everything. These are the findings leaders often don't expect, and some of them pay for the whole audit.

• Reports that mislead. At InCheck, the CRM measured response time on a 24/7 clock, so a 2-hour goal looked like 7.9 hours. The team wasn't slow. The report was wrong.
• Paying for nothing. SC&H, an accounting and advisory firm, was paying for 170,000 contacts nobody used, and its renewal price had doubled. Cutting billable contacts from about 300,000 to about 20,000 changed the math on the whole system.
• Fields nobody uses. DealRoom's CRM had about 8,000 properties, and about 6,000 of them were redundant. Every unused field makes the CRM harder to use and harder to trust.
• Access. Who can export all your contacts? Who can delete a pipeline? In most audits, the answer is "more people than you'd guess."

How do you decide which findings to fix first?

Rank findings by what they cost in revenue, how hard they are to fix, and what they block. Fix the cheap, high-impact items first, like routing gaps and deals with no next step. Then fix the data that everything else depends on. Leave cosmetic cleanup for last.

I use three questions for every finding:

1. What does it cost? Lost deals, wasted rep time, wrong invoices, or money spent on unused software.
2. How hard is the fix? Hours, days or a project?
3. What does it block? Bad data blocks routing, reports and every AI tool you want to add. Fix blockers early.

Then I group the list the way leaders can act on it:

• Fix this week. High cost, easy fix. Routing gaps, deals with no next step, a misleading report.
• Fix this quarter. High cost, real work. Duplicates, enrichment, pipeline redesign.
• Fix when convenient. Low cost. Old fields, naming, small cleanups.

The order matters as much as the list. My framework is visibility first, then intelligence, then activity. Fix what you can see. Then add the data that tells you where to look. Only then add more outreach or automation. Teams that skip to activity end up automating the mess.

Sometimes the audit shows that the tool itself is the problem. If that's where you land, read should we move from Salesforce to HubSpot before you sign anything.

Frequently asked questions

What is the most common CRM audit finding?

Duplicate records and deals with no next step show up in almost every audit. They're common because they build up slowly and nobody owns them. They're also the most damaging, because duplicates break routing and reports, and deals with no next step make the forecast a guess.

Should a CRM audit include software costs?

Yes. Seats nobody uses, contacts you pay for but never market to, and add-ons that duplicate each other are easy to miss and easy to fix. At one firm, cutting unused billable contacts changed the whole cost of the system. A good audit pays for itself here.

What should we do with findings we can't fix right away?

Write them down with an owner and a target date, then review the list monthly. Some findings, like a full data cleanup or a new pipeline design, need a project. Others can wait. What matters is that nothing important sits unowned, and that the team can see progress.

Can AI find these problems on its own?

AI is very good at the reading: scanning every field, deal and record, and grouping duplicates. It's not good at deciding what matters to your business. The best audits let AI do the heavy reading and have people who know the customers decide what to fix first.

How often should we check for these problems?

Run a full audit when something big changes, like a new sales leader, a merger or a new CRM. In between, check the same few views every week: new duplicates, deals with no next step, untouched leads and reports that disagree. Small, regular checks stop the list from growing back.

Sources

• RevOps Shop CRM audits, 2026: an IT services company with 18,000 duplicate companies, a skipped 5 to 6 touch playbook, lead-to-opportunity conversion of 10 to 13% and 12.5% of opportunities closing; and another company with contracts requested in Slack.
• Worldwide Logistics (WWL), 2025 to 2026: contact data judged 50 to 60% accurate, about 30% of rep time spent checking contacts, and billing contacts that were a coin flip. See the WWL case study.
• Esko business unit, 2026: 53% of leads from companies under $50M against a $50M to $1B target customer.
• Terraboost: 77 of 82 open deals with no next step, and about 20 pipelines simplified to 3. See the Terraboost case study.
• DealRoom, 2026: about 8,000 properties with about 6,000 redundant. See the DealRoom case study.
• InCheck, 2026: response time measured on a 24/7 clock at 7.9 hours against a 2-hour goal.
• SC&H, 2025 to 2026: 170,000 unused contacts, a doubled renewal price, and billable contacts cut from about 300,000 to about 20,000.
• HubSpot, Deduplicate records in HubSpot and HubSpot user permissions guide.