AI research.
Each newly-imported prospect goes through automated web-search-backed research. Takes about 30–60 seconds per prospect. The output is a structured research profile stored on the prospect.
What’s in a research profile
| Field | What it is |
|---|---|
| Company summary | 1–2 sentences |
| Industry | One value, used in templates as contact.industry |
| Company size estimate | Headcount / revenue bracket |
| Pain points | 2–3 bullets specific to this business |
| Personalisation hook | One specific, recent, verifiable fact (e.g. “Just opened a Toronto Eglinton location in March” or “Closed Series B led by Bessemer last quarter”) |
| Tone signals | Formal vs casual, technical vs business, etc. |
| Recent news | Last 90 days where available |
The personalisation hook is what feeds the P.S. line of your Step 1 email. It’s the difference between “Hope this email finds you well” and “Saw you announced your Eglinton location — congrats.”
Why prospects sometimes have weak research
The AI prefers verifiable, recent facts. If your prospect is a small business with no recent press, no LinkedIn post in 6 months, and a thin website, the research may fall back to industry generalities. You can:
- Help it — fill in
company_urlon import; that’s the strongest research anchor. - Skip it — use sequence templates and bypass per-prospect research entirely.
Research errors
If research fails (network glitch, bad URL), the prospect lands in pending_research indefinitely. The system retries automatically up to 3 times.
Persistent failures usually mean a bad email domain or a private/blocked website — email hello@tenloai.com and Tenlo will diagnose.