First-person field note | CRM cleanup

I Tested a 25-Row Lead Queue Before Touching a CRM

When a business says its follow-up is messy, the first step does not have to be a CRM login. A small redacted sample can show the shape of the problem before anyone changes a live workflow.

The boundary I used

I worked with a small sample and separated records into Ready, Hold, Duplicate, Do Not Contact, and Missing Context. The queue did not send anything, call an API, or ask for a CRM password.

Try the local queue Review the next step

I did not start with a CRM login

For a first test, I would rather understand the shape of the problem before changing a system that people rely on every day. A cleanup test should make the next decision clearer, not create a second data-access problem.

The sample used the smallest useful facts: an internal reference, received date, source, owner, status, last-contacted date, next action, a short context note, and permission information. Full names and sensitive details were not needed for the first routing pass.

The first question was not “Which leads are best?”

The first question was: what can we prove about each row? A record did not become Ready because it looked old or valuable. It became Ready only when the minimum facts were present and no stop signal was visible.

That difference matters. A list sorted by age can make a business feel productive while hiding duplicates and missing ownership. A decision queue shows the evidence that is actually there.

What the five outcomes exposed

Ready

Enough information existed for a person to review a draft. The next step was still approval, not automatic sending.

Hold

An unresolved decision or risk signal needed an owner who could read the underlying history.

Duplicate

The lead count was not necessarily a person count. One primary record needed to be chosen before contact.

Do Not Contact

The row stayed out of the message path. No alternative channel or new import should work around a stop signal.

Missing Context

The source, owner, last question, or next action was missing. A draft would only hide the gap.

Why I would not clean the whole CRM first

Full CRM cleanup sounds thorough, but it can become a project with no finish line. A small business may have years of records, inconsistent status names, several owners, and imports from different vendors.

A 25-row audit gives the team something concrete to discuss. Are the five outcomes understandable? Are the required fields available? Does the Do Not Contact rule hold? Are people willing to review a Ready row before a draft is used?

If the sample is not trustworthy, a larger export will not solve that problem. It will only make the review more expensive.

What I would change after the test

Require an owner and next action for new inquiries.
Add a clear suppression field and keep it visible.
Give duplicate records a review path.
Keep a small exception queue visible to the person responsible for follow-up.
Consider assisted drafting only after the evidence and decision boundary are working.

Even then, a draft is a starting point that a person edits and approves. The workflow should preserve the original evidence and the reason a row was held.

What the test does not prove

It does not prove that the business lost a particular amount of money. It does not identify which customer will respond. It does not replace a CRM audit, a privacy review, or the owner’s knowledge of the service area.

It answers a smaller question: can the team make a safer next decision from a limited, redacted sample? That question can be answered without handing over a password or exporting every customer conversation.

Bottom line

For CRM cleanup, start with one small sample and five explainable outcomes. Once the evidence is clear, the team can decide whether a larger cleanup is worth doing.

Start with 10 to 25 redacted rows