Before you approve more budget for lead generation, run one audit on the leads you already paid for. In most companies it takes an afternoon, and it changes the answer.
The pattern repeats across categories. A company reports a thin pipeline. The reflex is to buy more traffic. The reports look flat, so the assumption is that not enough people are arriving. The assumption is usually wrong. People arrived. The system did not tell anyone which ones mattered.
Volume is not the problem
A CRM records everything: every form fill, every download, every repeat visit, every stalled conversation. That is its job. What it does not do by default is separate the people who are close to buying from the people who filled in a form once and never returned.
So the sales team opens a list of several hundred names with no order to it, works the top of the list, gets a poor hit rate, and concludes that the leads are weak. Marketing sees the poor hit rate and concludes that it needs more leads. Both conclusions are wrong, and both cost money.
The pipeline is not empty. It is unsorted.
The audit
Export the full lead database. Not a filtered view, not a dashboard: the raw export. Then answer four questions.
One. Where are they actually located? Compare every record against the geography you are able to serve. In most databases a meaningful share of records sit outside the service area entirely. Those are not leads. They are noise inflating the denominator on every conversion rate you report.
Two. How many times did each person come back? Repeat visits are the single most honest signal in the database. Someone who returned three times without being chased has told you something that no lead score derived from job title will tell you. This is the group worth calling first.
Three. Who gave you permission to contact them? Every record without consent is a record you cannot email, regardless of how good it looks. Mark them and route them elsewhere. Emailing them is a regulatory exposure, and in several jurisdictions the penalty is not small.
Four. What is obviously junk? Test records, internal submissions, duplicates. They are always there. They always distort the numbers.
What the audit usually returns
Three groups fall out.
- A small set of genuinely warm records: correct geography, repeat behaviour, consent on file. This is the list to call this week.
- A larger set that is real but early. These need a sequence, not a phone call.
- A meaningful remainder that should never have been counted as pipeline at all.
The first group is almost always bigger than the sales team believed and smaller than the marketing report implied. That gap between the two numbers is the actual problem, and no amount of additional spend closes it.
Why this keeps happening
Two reasons, and neither is incompetence.
The first is that lead capture and lead qualification get built at different times by different people. Capture is set up in a hurry at launch. Qualification is meant to come later. Later does not arrive.
The second is that nobody owns the definition. Marketing counts a lead as anyone who filled in a form, because that is what the campaign is measured on. Sales counts a lead as someone worth a call. Both are internally consistent. Neither is written down. So the two teams argue about quality when they are describing different objects.
Fixing the second one is free. Agree in a single meeting what qualifies a record for a call, write the rule into the CRM as a field, and make the field mandatory. The argument disappears within a month.
The decision rule
Do not increase acquisition spend until you can state, in one number, how many contactable, in-territory, repeat-visit records are already sitting in the database. If nobody can produce that number in under a day, the reporting is the problem, not the funnel.
This is not an argument against spending on growth. It is an argument for knowing the base you are growing from. Fourteen years of running categories and launches taught me that the cheapest revenue in any business is almost always the revenue already sitting in a system nobody has sorted.
Start there. The traffic budget will still be waiting next month.