Year-End CRM Cleanup: The 10-Step Plan to Start Q1 With Data You Can Trust

Why December Is the Only Month This Gets Done

Every sales and marketing team says it will clean the CRM “after the quarter”. The quarter that actually has room for it is the last one, because deals slow down in the second half of December, next year’s targets depend on this year’s numbers being right, and anyone who has tried to report on pipeline in January with 2,000 ghost deals knows the cost of skipping it. This plan is written for a CRM with somewhere between 5,000 and 100,000 contact records and a team that can give it two half-days. The row counts are from a typical mid-sized export and are there so you can sanity-check your own.

Step 1 — Export Everything and Put It Somewhere Safe

Before any cleanup, export contacts, companies, deals and activities to CSV and store the files where nobody will edit them. Every step below is reversible only if this copy exists. Note the counts: a typical export might be 48,000 contacts, 9,500 companies, 6,200 deals. Those numbers are your baseline; every later step should explain how they change.

Step 2 — Archive Deals That Will Never Close

Filter open deals by last activity date. Anything with no activity in 180 days and no scheduled next step is dead; close it as lost with a reason code of “no response, year-end cleanup” rather than deleting it, so the history stays. In the example export this removes about 1,900 of 6,200 open deals from the pipeline and is the single change that makes January forecasts believable. Tell the reps first; nobody likes finding their pipeline halved without warning, even when the deals were fiction.

Step 3 — Dedupe Contacts on Email, Then Review Near Matches

Exact duplicates on a lower-cased email address are safe to merge automatically; keep the record with the most recent activity and merge the other’s history into it. In a 48,000-contact export, 3 to 5 percent exact duplicates is normal, so expect 1,500 to 2,400 merges. Near duplicates (same name and company, different email; same email with a typo) need a human. Run them through the Duplicate Remover or the CSV Cleaner to get a candidate list, then decide row by row. Our guide on removing duplicate CRM contacts covers the merge rules.

Step 4 — Fix Ownership

Every year people leave and their records stay assigned to them. Filter contacts and companies by owner; reassign anything owned by a departed user to their territory’s current owner, or to a holding queue if no one owns the territory. Unowned records are the ones that never get called and never get cleaned, so this step is also what prevents next year’s mess. In the example, 2,700 contacts belong to three former employees.

Step 5 — Standardise the Fields You Report On

Pick the five fields that appear in your dashboards (usually industry, company size, country, lead source, lifecycle stage) and standardise their values. “US”, “USA”, “United States” and “U.S.” become one value. Free-text industry fields become a controlled list. This is tedious and it is where most reporting errors come from. The CSV Cleaner handles whitespace and case; the mapping of values is a spreadsheet job with a lookup table that you keep for next year.

Step 6 — Validate Emails and Phone Numbers

Run the contact export through an email validation step so bounced and role addresses (info@, sales@) are flagged before the January campaign, and standardise phone numbers to one format with country codes. Expect 6 to 12 percent of a two-year-old list to be undeliverable. Flag rather than delete; a dead email on a live account is still a live account. The deliverability checklist covers the validation step.

Step 7 — Fill or Remove Empty Required Fields

Pull a completeness report: what percentage of contacts have a title, a phone, a company, a country? Anything under 70 percent complete on a field your process depends on needs a decision: enrich it, make it optional, or stop relying on it in reports. The Lead Quality Analyzer gives this breakdown from a CSV in seconds. Enrichment is a separate project with its own budget; see the enrichment guide.

Step 8 — Reconcile Companies and Contacts

Contacts with no company, companies with no contacts, and two company records for one organisation (parent and subsidiary, old name and new name) all distort account-based reporting. Merge company duplicates on normalised name plus domain, attach orphan contacts by email domain where it is unambiguous, and leave the rest for sales to resolve in January. The Company Domain Finder helps with the domain side.

Step 9 — Write Down the Rules

Everything you decided above (merge rule, dead-deal threshold, controlled value lists, required fields) goes into a one-page data standard that lives next to the CRM. Without it, the same cleanup is needed next December. With it, the monthly maintenance in step 10 is a checklist rather than a project. Our service standards page shows the format we use for client files.

Step 10 — Set the Monthly Rhythm

Thirty minutes on the first business day of each month: new duplicates since last month, deals past the dead threshold, records owned by anyone who left, completeness on the five reporting fields. The Data Health Score gives one number to track; if it drops more than a few points in a month, something upstream changed. Teams that cannot staff this are what the Data Care Plan is for.

What the Example Export Looked Like Afterwards

Two half-days of work, one lookup table to keep, and a January where the pipeline number means something. The figures are illustrative but the proportions are what we see in real files; if yours are very different, that is itself worth understanding before you plan the quarter.

Frequently Asked Questions

How long does a year-end CRM cleanup take?

For a CRM with 5,000 to 100,000 contacts and one person who knows the data, plan two half-days: one for export, dedupe and dead-deal archiving, one for field standardisation, ownership and validation. Enrichment is a separate project.

Should I delete dead deals or close them as lost?

Close them as lost with a reason code such as 'no response, year-end cleanup'. Deleting loses the history and makes next year's win-rate analysis wrong; closing keeps the record while removing it from the pipeline.

What percentage of a CRM is usually duplicates?

Exact duplicates on email are typically 3 to 5 percent of contacts in a list that has not been cleaned for a year or more. Near duplicates (same person, different email or spelling) add another 1 to 3 percent and need human review.