In a commerce lead-generation pilot, the brief sounded familiar: more leads, lower cost. The platforms were happily obliging — lead volumes up, cost per lead down — and the sales team was drowning in leads that would never convert. The system was optimising for exactly what it was asked to optimise for. That was the problem.
The lead-quality tax
When "a lead" is defined as a filled form, platforms optimise toward people who fill forms. That population skews curious, incentive-driven and mis-typed. Every rupee spent acquiring them is a tax the pipeline pays for measurability.
The uncomfortable truth: a cheaper bad lead is not an improvement. It is the same bad lead with better reporting.
Redefining the unit of value
We ran a working session with the sales team and defined a hot lead behaviourally: completeness signals, intent signals (category, budget, timeline) and validation (contactable on the first attempt). That definition — not the form fill — became the conversion event fed back to the platforms.
- One definition, agreed with sales, written down and versioned like code.
- Validation at the point of capture, not days later in a CRM queue.
- Platform conversion events switched from "lead submitted" to "validated hot lead".
- A weekly feedback loop: sales outcomes flow back into bidding, audiences and creative.
Platforms are extremely good at getting you more of whatever you tell them to count. Choose the count carefully.
Within the pilot, more than 4,000 leads were validated against the new definition. The reported cost of acquiring a hot lead fell by over 50% — not because the media got cheaper, but because the algorithm was finally pointed at something worth chasing, and sales stopped paying the quality tax.
The part most teams skip
The mechanics were not the hard part. The hard part was organisational: getting media and sales to own one definition, and accepting that lead volume would drop while pipeline quality improved. The first month looked worse on the dashboard. It was not worse for the business.
- Expect reported volume to fall — agree that with leadership before the switch, not after.
- Judge the change on pipeline and revenue, never on lead counts.
- Keep the definition under review; markets drift, and definitions should drift with them — deliberately.
Lead quality is not a targeting problem. It is a definition problem. Fix the definition and the machines do the rest.
Figures in this article are drawn from the founders' prior commerce lead-generation pilot experience and are presented as learnings, not as client results.
AS
SS