Most operators think they have a lead problem. Usually they have a retrieval problem.
The leads are in there. They were paid for. Somebody worked them for a week. Then the system stopped making it obvious what to do next, and the deal quietly stopped moving.
That is not a discipline failure. It is a structure failure. And you can find most of it yourself in about an hour, without buying anything.
Here are twelve checks. Run them in order. Write down what you find.
01
Count the opportunities with no stage
Filter your pipeline for records sitting outside any stage. Every one of them is a lead nobody owns. In most accounts we open, this number is not zero, and nobody knew.
02
Sort by last activity, oldest first
Scroll to the bottom. If deals have sat untouched for ninety days in an active stage, the stage is not a stage. It is a shelf.
03
Search one phone number two ways
Try it with dashes and without. If you get different results, your phone fields are not normalised, and your dialer is calling the same person twice.
04
Ask what each stage actually means
Say it out loud. If two people on your team describe a stage differently, your pipeline report is fiction. Stages need an entry condition and an exit condition, written down.
05
Look for a stage nothing ever leaves
Every messy CRM has one. Deals go in and never come out. It is usually called Nurture, Warm, or Follow Up. It is where deals go to be forgotten politely.
06
Count your tags
If you have more tags than stages, tags are doing the work stages should. Over fifty tags on a small account almost always means nobody can filter anything reliably.
07
Find the duplicates on purpose
Export, sort by phone, look for repeats. Then sort by email and do it again. Duplicates split a contact history in half, so the person calling never sees the last conversation.
08
Check whether calls are logged automatically
Make one call. Then look at the record. If you had to type anything for it to appear, it will not happen consistently, and your call data is already wrong.
09
Open the three oldest open deals
Read the notes. If you cannot tell what the next action is within ten seconds, neither can the person who inherits it.
10
Test one inbound form end to end
Submit it yourself with a real phone number. Time how long until it appears, whether it lands in the right stage, and whether anyone is notified. Most leaks are here.
11
Look at what happens after a no-answer
Nothing? Then a no-answer is a dead end rather than a step. A dialled number that goes nowhere is the most expensive record in the database.
12
Ask for last month’s numbers twice
Once from the CRM dashboard, once from whoever reports them by hand. If the two disagree, the dashboard is decoration and decisions are being made on a spreadsheet.
What the score means
Nought to two broken: your CRM is fine. Spend your money on more leads, not on the system.
Three to six: you are losing deals, but the structure is salvageable. This is where most accounts land.
Seven or more: stop buying leads. You are pouring them into a container with a hole in it, and more volume makes the hole worse, not better.
The part people get wrong
The instinct is to rebuild. New CRM, fresh start, clean slate.
It is almost always the wrong move. A rebuild costs months, and the history you throw away is the only thing that tells you which leads were ever real. Nearly every problem on this list is fixable inside the instance you already have.
Fix the stages first. Then the phone fields. Then the duplicates. In that order, because each one makes the next easier to see.