Lead Data Trust, end to end

Why bought lists behave the way they do

List buying is routine in Indian B2B, and the failure pattern is consistent. The list arrives, it looks substantial, it gets imported, and for two weeks the team works it hard. Connect rates are poor but not catastrophic, so nobody stops. A month later the list is quietly abandoned and the next one is bought.

What went wrong is usually not that the data was fake. It is that the data was old, and nothing in the system said so. A contact who filled a form eight months ago has no memory of doing it, has often changed roles, and reacts to the call as an intrusion rather than a follow-up. The record looked identical on import to one generated yesterday.

That is the specific failure freshness scoring exists to prevent. Leadkaun bands every lead by age and keeps ageing it, so a list imported today does not present as new work. Where the original collection date is unknown the band is inferred from import date and flagged as such, an honest approximation rather than a false precision.

Reading the record before a rep touches it

The most expensive moment in a bad import is the hour after it, when reps start dialling. Anything you can learn before that point is nearly free.

A useful intake check answers plain questions about a file: how many rows have a valid Indian mobile, how many companies look like businesses rather than individuals, how many rows duplicate each other, which fields are systematically missing, and whether this looks like a B2B or a B2C dataset at all. None of that needs a model. It is arithmetic on formats and completeness.

Leadkaun runs that profile on a sample of the file before import and reports a readiness band rather than a score, because a band is a decision and a number is a debate. High means import it. Low means the problem is the file, and no amount of rep effort will fix that.

The point is not to reject files. It is to know what you are importing while you still have the option not to.

Confidence is not the same as grade

A grade answers how good a lead is. Confidence answers how much we actually know about it. These are separate axes and conflating them is a common, expensive mistake.

A lead with a company name, a designation, a location and a stated budget can be graded with some assurance. A lead with a first name and a mobile number might be the best opportunity in the pipeline, and the honest position is that there is not enough information to say. Scoring that lead low implies it is weak, which is a different claim from admitting ignorance.

Leadkaun keeps them apart. Confidence is a weighted read of field completeness with a prioritised list of what to ask for next, so a rep with a thin lead knows exactly which two questions would move it. A thin lead is not a bad lead. It is an unknown one, and that distinction changes what a rep does with it.

Watching a grade change over time

Scores that move without explanation are scores nobody trusts. If a rep opens a lead that was Grade A yesterday and finds it Grade C today, the useful response is not reassurance. It is showing them what happened.

Every score change in Leadkaun is written to an append-only timeline with the grade, confidence and all three sub-scores frozen as they were at that moment. A drop is traceable to the event that caused it: intent decayed after a fortnight of silence, or a duplicate was detected, or the ICP changed and the whole book was re-graded.

This is also the honest answer to a question buyers ask often and vendors usually dodge: what happens if your scoring is wrong? A frozen history means you can go back and check whether the grades predicted the outcomes, which is the only way to find out.

What to do about it this quarter

Start by measuring quality by source rather than volume by source. Most teams know which channel produces the most leads and few know which produces the most valid records. The gap between those two rankings is usually where the budget is being wasted.

Then set the customer profile properly. Fit is computed from the industries, states, business types, roles and budget bands you configure, and a vague profile produces a confident-looking grade on the wrong leads. This is twenty minutes of work that everything downstream inherits.

Finally, stop treating disqualification as permanent. A lead that scores badly should sink in the ordering rather than leave it, so a later signal can bring it back. Permanent rejection should be reserved for genuine wrong fit and invalid records, not for a lead that happened to be unreachable on a Tuesday.