Outbound Lead Engine
Synthetic data only: fictional companies on .example domains, mock data providers,
a simulated sender. Source and measured results.
250
companies
576
verified contacts
375
scored 3+
203
with a theory
119
emails sent
4
replied
1
positive
Theories
95% Wilson interval of the positive reply rate. A theory is paused when the high end drops below 1%.
| name | status | sent | replied | positive | positive pct | 95% low, % | 95% high, % |
|---|---|---|---|---|---|---|---|
| new office, new market | active | 82 | 4 | 1 | 1.22 | 0.22 | 6.59 |
| fresh funding, hiring for pipeline | active | 6 | 0 | 0 | 0.00 | 0.00 | 39.03 |
| new tool, broken attribution | active | 31 | 0 | 0 | 0.00 | 0.00 | 11.03 |
Last 14 days
By event day.
| day | sent | opens | replies | positive | bounces | unsubscribes |
|---|---|---|---|---|---|---|
| 2026-10-05 | 60 | 29 | 4 | 1 | 2 | 0 |
| 2026-09-24 | 20 | 9 | 0 | 0 | 0 | 0 |
| 2026-09-23 | 39 | 12 | 0 | 0 | 0 | 1 |
Positive replies
Classified by the LLM when a Groq key is configured, by keyword rules otherwise; a failed LLM call falls back to the rules and is counted below.
| day | first name | title | company | theory | variant | reply text |
|---|---|---|---|---|---|---|
| 2026-10-05 | Anna | CMO | PrimeLabs | new office, new market | B | We're rebuilding reporting right now, so happy to take a look. How long does setup take? |
Where the money went
Mock provider prices from the design's cost table; LLM at Groq list prices.
| item | usd | calls |
|---|---|---|
| enrichment: phantombuster | 17.0000 | 740 |
| signals: apify_news | 16.6000 | 166 |
| signals: apify_jobs | 16.6000 | 166 |
| enrichment: findymail | 8.6520 | 721 |
| signals: builtwith | 8.3000 | 166 |
| signals: crunchbase | 8.3000 | 166 |
| enrichment: apollo | 7.3280 | 708 |
| enrichment: hunter | 2.0300 | 203 |
| enrichment: snov | 1.9680 | 165 |
| llm: copy (openai/gpt-oss-120b) | 0.0758 | 218 |
Reply classes
| reply class | replies | by llm | by fallback |
|---|---|---|---|
| ooo | 3 | 3 | 0 |
| negative | 2 | 2 | 0 |
| positive | 1 | 1 | 0 |
| referral | 1 | 1 | 0 |
Mailboxes
| status | daily limit | assigned | bounced | unsubscribed | bounce pct | |
|---|---|---|---|---|---|---|
| mira1@tallybird-mail.example | active | 20 | 60 | 1 | 1 | 1.67 |
| mira2@tallybird-mail.example | active | 20 | 39 | 1 | 0 | 2.56 |
| mira3@tallybird-mail.example | active | 20 | 20 | 0 | 0 | 0.00 |
Errors
Dead letters, rejected webhooks, rejected LLM answers.
| source | n | last |
|---|---|---|
| llm answer rejected: copy | 15 | 2026-10-05 |
| webhook rejected: stale or future timestamp | 1 | 2026-09-23 |
| webhook rejected: missing signature | 1 | 2026-09-23 |
Draft theories
Proposed by the LLM from segment reply rates; a human switches them on.
nothing yet