Can fail to match a known person at all, returning 0 results for a named-person search instead of a usable prospect record.
What was measured
Relevance
How closely the generated email matches the provided company, prospect, and outreach context.
decisive for this rankingtransformation
If the generated email does not match the company, prospect, and outreach context, it is not doing the core job well. (3 of 3 judges)
What was given, what came back
Test input: Aditya Nichite prospect lookup · text · group: prospect-search
Input — what we sent
The exact prompt
Aditya Nichite
Direct people/prospect search for Aditya Nichite to test known-prospect retrieval and compare how well tools return validated contact records.
Output — unretouched
Snov.io prospect search with one contact result.
Also checked on this input — same tool, 5 other criteria
People Search✓ WorkedCan locate a specific prospect and return validating context such as company affiliation, job title, location, and email, which is sufficient to confirm a match before outreach.People Search✗ FailedCan also fail completely on a fully specified person query, returning 0 matches and no usable prospect record for a known individual.Personalization quality◐ MixedCan swap in recipient- and company-specific details, but the generated email body still keeps the same overall structure, messaging, and value proposition across different prospects.Supporting prospect information✓ WorkedCan return a single prospect record with enough validation context — name, role, location, and company — to sanity-check the match before outreach.Template dependency⚠ StruggledRelies on a largely fixed outreach template, because two generated emails shared the same structure and core messaging with only limited prospect-specific edits.
Provenance
- Observation
- 2bd11adf-c249-4a48-9c09-aa72504c02b1
- Evidence run
- 9e5960d0-1a72-4f29-9e1e-c7fc96120e61
- Study
- AI Outreach
- Research task
- 86ba3udqq
- Tested at
- Jun 15, 2026
- Source
- first-party
- Evidence state
- verified
- Proof shown
- input + output shown
- Cost / latency
- not captured
- Repeat run
- not captured
- Tester
- not captured
The last three rows are honest blanks, not placeholders — our capture has no field for them yet.
Query this
get_evidence({
tool: "snov-io",
scenario: "prospect-search"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Same input, same check — 0 other tools
measured on Relevance
No other tool was measured on this criterion for this input.
This evidence is published in
Real inputs and real outputs, no retouching · every cell queryable via API & MCP · aidemos.com