Building a call list that is worth working through

A list of five thousand numbers feels like an asset and is usually a liability, because the work of calling scales with the length while the results scale with the quality.

The arithmetic is worth making explicit. Every number on a list costs the same to call: the dial, the wait, the attempt, the note. A number with no plausible relationship to the business consumes that cost and returns nothing. So a long list of weak contacts does not contain the good ones as a bonus — it buries them under work, and the caller reaches the good ones exhausted, or not at all.

There is a second effect that compounds. Working a poor list produces a long run of refusals and dead numbers, and that experience changes how someone sounds. By the time a plausible prospect answers, they get a caller who has been refused eleven times that morning.

So the useful question about a list is not how many contacts it holds but what proportion of them have a genuine reason to take the call.

The sources that produce conversations share one property: something already happened that connects the person to the business.

Existing enquiries that were never followed up are almost always the best available list, and most small businesses have more of them than they realise — messages, missed calls, forms, questions asked in person. These people raised their hand, and the reason they were not converted is usually that nobody called back.

Past customers who have not bought recently are next. There is an established relationship, the business knows what they bought, and the call has an obvious subject.

Referrals from existing customers connect through someone the person trusts, which changes the opening entirely.

Event attendees and people who downloaded or requested something have demonstrated interest in the subject, though the strength varies with how much effort the action took.

What these have in common is a specific first sentence available to the caller. That sentence is most of what determines whether the call continues, and no purchased list can supply it.

Bought contact lists underperform for reasons that are structural rather than a matter of finding a better supplier.

The numbers are usually old, and phone numbers decay: people change providers, businesses close, roles change. A list assembled some time ago and resold repeatedly has a substantial share of numbers that no longer reach the intended person.

The targeting is generally weak, because a list is more saleable the larger it is, which pushes suppliers towards inclusion rather than precision.

The people on it have no relationship with the business, so every call is cold in the hardest sense — there is no first sentence available beyond a description of the caller's own business.

And the list has usually been called already, by everyone else who bought it, which means recipients are primed to refuse.

There is also the regulatory dimension, which is decisive in India: consent does not travel with a phone number, so a purchased list carries no consent basis, and an assurance that it is opt-in or DND-cleaned is unverifiable and does not shift responsibility from whoever dials.

A list becomes workable when it is sorted, and the sorting principle that matters is the strength of the existing connection rather than the estimated size of the opportunity.

Three tiers are usually enough. The strongest tier is people who took an action recently — a recent enquiry, a recent visit, a referral. These should be called first and quickly, because the relevant property is recency and it decays fast.

The middle tier is established but not recent: past customers, older enquiries, people met at events. There is a real basis for the call and no urgency.

The weakest tier is everyone with no prior connection. This tier should be small, worked last, and honestly assessed — if it never produces anything, it is a cost.

Sorting by opportunity size instead is the common error. It puts a large but unconnected prospect ahead of a small one who enquired last week, and the second call is far more likely to produce business. Potential value matters only after the probability of a conversation is accounted for.

The point of segmenting is that each tier permits a different opening, and using the right one is most of what a good list buys you.

For a recent action, the opening references it directly and needs nothing else — calling about the enquiry sent on Tuesday requires no justification and no permission-seeking, because the person initiated it.

For an established but older connection, the opening references the history and then supplies a current reason: acknowledging a previous purchase and saying what has changed since. The history earns attention; the current reason justifies today.

For a genuinely cold contact, the opening has to do all the work — identifying yourself, saying specifically why this person is being called, asking permission, and reaching a question fast.

Using the cold opening on a warm contact is a common and costly mistake, because it discards the connection and tells someone who enquired last week that they are on a list. The reverse — assuming familiarity with a stranger — reads as presumptuous and is spotted immediately.

A list is not a fixed asset; it degrades continuously and needs a small amount of upkeep to remain useful. Three habits do most of it. Record the source and date for every contact when it is added, since that single field determines which tier it belongs in and is impossible to reconstruct later. Remove numbers that are wrong or unreachable rather than leaving them to be redialled — a number that failed three times will fail a fourth. And record and honour every request not to be called, permanently, which is both a regulatory obligation and the thing that prevents the worst kind of error.

Then add to it deliberately. The best source of future contacts is the flow of enquiries the business already receives, and capturing those consistently at the moment they arrive is a more valuable habit than any acquisition exercise.

Automix's Wani is a planned voice follow-up agent, documented at /docs/wani. It is not available to customers and places no calls today, so the list described here is maintained and worked by people.