Lead list hygiene for outbound teams
What goes wrong with prospect lists — duplicates, unnormalised and dead numbers, collisions, silent depletion — and the import-time controls that stop it.
E.164
normalised
38
duplicates
12
DNC removed
List quality sets the ceiling on everything downstream. The same rep on a clean, well-targeted list and on a scraped one will produce very different numbers, and coaching does not close that gap. Yet list hygiene is usually a one-off cleanup rather than a standing control.
Hygiene problems should be caught at import. A control that depends on a rep remembering is not a control.
Seven ways a lead list goes wrong
Duplicates
Two sources, or one source twice — one prospect becomes three leads.
Unnormalised numbers
One phone number in five formats, so nothing deduplicates.
Dead numbers
Disconnected lines and switchboards that go nowhere.
Suppression by memory
A do-not-call list the rep is supposed to check.
Collisions
Two reps calling the same company for the same client.
Silent depletion
The list runs out and nothing errors.
Invisible errors
An import that reports a count, not which rows failed.
Deduplicate on domain or phone number
A CSV from two sources and one prospect appears as three leads. Reps work all three, and the prospect notices.
Choosing the key is the real work. Company name is unreliable (Ltd vs Limited vs abbreviations). Domain is better. Phone number is good once normalised.
Normalise phone numbers to E.164
| As imported | Normalised (E.164) |
|---|---|
| 040 123 4567 | +358401234567 |
| 0401234567 | +358401234567 |
| +358 40 123 4567 | +358401234567 |
| 358401234567 | +358401234567 |
| +358-40-123-4567 | +358401234567 |
Five strings, one number. E.164 — plus sign, country code, subscriber number, no punctuation — is the only format that deduplicates reliably, and telephony APIs generally require it, so normalising late means failed calls. National-format numbers need a default country, which should come from the account’s configured country rather than a per-row guess.
Track dead numbers as data
Some share of any list is dead. Instead of a vague sense that “the data is rough”, make phone quality a tracked property recomputed as call outcomes come in, and record where each number came from. You can stop reps burning hours on a mostly dead list, and you can see which supplier’s data is worth paying for — usually the bigger saving.
Enforce do-not-call lists at import
Every client has companies you must not contact: existing customers, live negotiations, competitors, and anyone who has objected — the last one a GDPR obligation. A per-client do-not-call list has to filter rows out before they reach a rep, and because it is per client it cannot be one global register.
Stop collisions with expiring reservations
On a shared list, collisions are certain. The fix is a reservation: a claim pinning a lead to the rep working it. Make it expire. A permanent assignment turns the list into a graveyard of leads one rep touched once; an expiring hold, with a background job releasing stale claims, keeps the list circulating.
Alert before a list runs out
A rep on a depleted list produces nothing, activity drifts down, and it can take a week before anyone asks why. Depletion alerting per list pays for itself immediately — and remaining workable leads predict next week’s output better than this week’s dial count.
Report import errors per row
“1,240 imported, 83 failed” is not actionable, and a truncated error list is worse because it implies you have seen everything. The usual cause of a failed batch is one formatting issue across hundreds of rows — fixable in seconds once you can see it.
The import checklist
Every import, in this order
- 1
Normalise
Phone numbers to E.164, using the account country as the default.
- 2
Deduplicate
On domain or normalised number, never company name.
- 3
Suppress
Filter against the client's do-not-call list.
- 4
Validate
Reject or flag rows missing required fields, and report errors per row.
- 5
Tag and watch
Record the source on each lead and set a depletion threshold on the list.
Clean lists remove a ceiling
Clean lists do not make a mediocre rep good. Dirty lists make a good rep look mediocre — and then you coach the wrong problem.