The problem this solves
An Apollo saved search is a starting point, not a list. The rows carry whatever the database last believed: contacts who changed jobs two quarters ago, addresses that bounce, companies whose size band drifted out of your ICP. Sending to it raw is how deliverability dies, and cleaning it by hand means an afternoon of tab-switching between enrichment lookups and a verifier before a single email goes out.
The gap between search and sequence is where most outbound momentum is lost. Someone owns the search, someone else owns the sending, and the handoff happens through an export that goes stale on a shared drive. By the time the list is loaded, the freshest signals that justified it are weeks old and nobody remembers which rows were already vetted.
What teams actually want is one motion: define the segment once, and have everything between definition and first touch - enrichment, verification, deduplication, sequence loading - happen without a human relay race.
How the mission runs
- Pull the saved search through your Apollo connection. The agent reads the saved search from your connected Apollo account - your key, your data, your defined segment - and takes the current result set as the candidate universe rather than a stale export.
- Enrich every contact against live firmographics. Each row is cross-checked with the platform's enrichment and firmographic tools: current company, current title, size band, industry. Contacts whose reality no longer matches the segment definition drop out here instead of polluting the sequence.
- Verify every address before it counts. Bulk verification runs on the full survivor set. Deliverable addresses stay; risky and invalid ones are flagged and excluded, so the sequence starts with a list that protects your sender reputation instead of spending it.
- Deduplicate against what you already touched. The cleaned set is checked for duplicates by email and by domain, so nobody who is already mid-sequence gets a cold first touch twice - the mistake recipients remember and the one that gets sending domains reported. Domain-level checks also catch the subtler version: two contacts at the same small company entering separate sequences with different messages in the same week.
- Load the survivors into a named Apollo sequence. The final list lands back in Apollo as a new sequence named after the segment, ready for your copy. You get the true count, who was dropped and why, and a CSV of the whole audit trail.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
A named Apollo sequence containing only enriched, verified, deduplicated contacts, plus a CSV audit of every drop decision and the honest final count.
Make it yours
- Point it at a different saved search each week and let the sequence name carry the date, so segments stay comparable over time.
- Add an intent filter: only load contacts whose companies show active category research this month.
- Skip the sequence load and deliver CSV-only when the list is for an event invite or a different sending tool.
Frequently asked questions
Does this need my Apollo API key?
Yes - Apollo connects with your own key under Settings, then missions work against your account and your saved searches. Nothing is shared across workspaces.
What happens to contacts that fail verification?
They are excluded from the sequence and listed in the CSV with the failure reason, so you can decide whether to re-find a better address or drop the contact.
Will it send the emails too?
No - it loads the sequence and stops. Copy, cadence and the send button stay yours, in Apollo, where your team already reviews outbound.