ZapBounce and Copper

Copper creates and enriches records from Gmail activity. That is its selling point and it has a side effect: every address a rep ever typed into a To field can become a record.

Including the ones typed wrong. A mistyped address that bounced still produced an outbound message, and Copper saw it.

Automatic capture records mistakes as contacts

A rep sends to a mistyped address, it bounces, the rep corrects it and sends again. Copper now holds two records: the real person and a contact that never existed.

Over a year of activity that accumulates quietly, and nobody notices until someone builds a marketing list from the contact base and a chunk of it bounces.

Copper's own deduplication matches on address, so the phantom and the real contact look like different people and stay separate.

How the data moves

  1. Verify the whole contact base once

    The phantom records are spread across the history, not concentrated in one import, so a full pass is the only way to find them.

  2. Use a custom field for the verdict

    Copper custom fields are filterable, so the verdict can drive a view without changing any record.

  3. Check whether an invalid address has a near neighbor

    An invalid address often sits next to a near-identical valid one the rep sent to afterwards, which identifies the pair to merge.

  4. Verify at the point of capture where you can

    Copper's API and its automations can call out. Checking before a record is created stops the accumulation restarting.

Setting it up

  1. Export contacts from Copper as CSV.
  2. Verify the file and look at the reason codes alongside the invalid verdicts.
  3. Create a custom field for the verdict and import the results back.
  4. Review invalid records that have a near-identical valid record, which are usually merge candidates rather than deletions.
  5. Build a filtered view on the verdict for any campaign list.
  6. Add a verification call to whatever creates contacts through the API.

Copper: common questions

Why does Copper collect bad addresses?

Because it records what reps actually sent, including the message that bounced. That is accurate capture producing an inaccurate contact.

Should I delete the phantom records?

Merge them into the real contact where an invalid address matches a valid one as a typo pair. Deleting loses the activity history attached.

How often should this run?

Quarterly. The accumulation is slow and steady rather than sudden.

Check a Copper export today

100 free checks a month, no card. Unknown results are refunded.