Pre-meeting people research that actually helps

DeepSearch Name is public-web research — not a background check and not FCRA-compliant. Use it for professional context with sources you can verify; use a regulated consumer reporting agency for employment, housing, or credit decisions. Read the full distinction.

The goal of pre-meeting research is not to memorize a biography. It is to avoid wrong-person mistakes, show respect for someone’s public work, and walk in with one useful question. This workflow fits founders, sellers, recruiters, investors, and journalists — using public sources only. It is not a background check.

Most meeting prep fails in one of three ways: you research the wrong namesake, you drown in tabs without forming a question, or you cross into creepy personal digging that poisons rapport. The fix is a timed, source-first workflow with a hard stop when evidence is thin.

Stay in the right category

If your “research” is really screening for employment or housing eligibility, stop and use a compliant process — not a background check, people search vs background check. For call prep, continue. Keep prep notes labeled as prep. Do not file a sourced AI brief into an HR screening packet and call it diligence.

A useful self-check: if the artifact you are building would be awkward to describe to the other person — “I prepared from your conference talk” vs “I compiled a risk file” — you are probably in the wrong category. Pre-meeting research should be something you could admit in the first five minutes without damaging trust.

The 12-minute workflow

If you only have five minutes, compress to: confirm identity (two signals) → one public artifact → one question. Skip biography completeness. Completeness is how people waste the entire buffer reading LinkedIn endorsements.

Minute 0–3: confirm the human

Use company and city when the calendar invite is sparse. Calendar names lie in small ways: nicknames, missing middle initials, or a contractor listed under a client brand. Recover clues from the email domain, Zoom display name, shared Slack channel, or the person who introduced you.

For common names, follow common name disambiguation. Product entry points: find someone by name, people search. If two plausible people remain when the call starts, your first spoken question should clarify role or team — not pretend certainty.

Minute 3–8: scan public work (not gossip)

Your skim has one job: find a public artifact that earns you a relevant question. You are not collecting a life story. When you catch yourself opening the fifth tab about an old job from a decade ago, you have left the meeting-prep lane.

High-yield sources

• Company bio / team page
• Recent interview, podcast, or conference talk
• Product changelog, GitHub, or portfolio relevant to the meeting
• A thoughtful post they authored (not a like or reshare)
• Press quote or bylined article tied to their current mandate

Low-yield / avoid

• Distant relatives and personal address hunting
• Old controversy rabbit holes unrelated to the agenda
• Private social content
• Uncited AI summaries you did not open
• Comment-section archaeology and likes history

The quality test is simple: would you be comfortable saying “I saw your talk / post / launch about X” in the first five minutes? If not, you are probably researching the wrong layer. Longer article: research someone before a meeting.

Minute 8–11: turn research into conversation

Good talking points

• “You mentioned X in your talk on Y — we’re seeing Z; how are you thinking about it?”
• “Your team shipped A last quarter — does that change priority B?”
• “I saw you moved from role C to D — what’s continuous vs new in this mandate?”
• “Your post on X resonated — we’re stuck on Y; what broke for you first?”

Bad talking points

• Personal life trivia that signals surveillance
• Gotcha questions from unverified blogs
• Reciting their entire résumé back to them
• Correcting them with a stale title you found somewhere thin

Aim for one primary question and one backup. Two strong questions beat eight shallow ones. Note one assumption you might be wrong about (for example, “still owning partnerships?”) so you can revise live if they correct you.

After you have a brief: useful follow-ups

If you use DeepSearch Name chat after a sourced profile, keep prompts evidentiary:

Reject chat answers that cannot point to a URL. Paste the URLs you trust into the agenda doc so you are not relying on memory mid-call. More on the AI layer: AI people search.

Role-specific variants

Founders meeting investors or partners

Confirm firm + partner vs associate; skim a recent thesis post or portfolio note. Reference something they publicly care about, not a rumor about fund performance. See for founders, for investors.

Sales discovery calls

Confirm buying role; reference a public initiative, hiring page theme, or product launch — not a guessed personal detail. Your research goal is relevance, not intimidation. See for sales.

Recruiting screens

Confirm identity and public work for conversation — not covert screening. If your company needs a background check later, that is a separate CRA step with notices. See for recruiters and people search vs background check.

Journalist interviews

Map affiliations and prior on-record quotes with citations. Note what is on the record vs what is only implied by a bio. See for journalists.

Customer success / renewals

Confirm whether your contact still owns the account relationship. Title drift is common after reorgs; a stale CRM contact is a classic wrong-person failure mode before a QBR.

Worked example: 10 minutes before a customer call

1. Calendar says “Avery Ng.” Email domain reveals employer Acme Robotics.
2. Search name + Acme; confirm on team page and a conference listing.
3. Skim Avery’s talk abstract on fleet diagnostics; prepare one question about rollout blockers.
4. Paste two URLs into the call doc; enter the meeting without stalking LinkedIn comments.

Worked example: investor intro with a common name

1. Invite lists “Jordan Kim.” Firm is clear from the email domain; seat is not.
2. Search name + firm; find two Jordans associated with adjacent firms in the same city.
3. Confirm via a recent portfolio announcement byline and the firm’s team page.
4. Skim one thesis post; ask how that thesis applies to your category — not “I read your whole bio.”
5. If still unsure which Jordan, open with “I want to make sure I’ve got the right context — you’re on the X team at Firm?”

Failure modes (and how to recover)

Team template (copy/paste)

• Person:
• Employer / role (source URL):
• Second source URL:
• Public work noted:
• Question to ask:
• Open risks / unknowns:
• Label: meeting prep (not screening)

Identity checklist: verify public web identity. Method depth: how to find someone online, AI people search.

Depth by stakes

Multi-attendee meetings

When three external people join, do not research all of them to biography depth. Triage:

1. Identify the decision-maker or primary counterpart first.
2. Confirm identity for everyone at a lightweight level (name + employer).
3. Do full public-work skim only for the one or two people who will drive the agenda.
4. Note reporting lines if publicly obvious — but do not invent org charts from thin titles.

This keeps you from spending forty minutes before a thirty-minute call. If roles are unclear from the invite, ask your champion who owns the decision — that social clue beats another round of common-name search.

What “good enough” looks like by outcome

Pre-meeting research is successful when it changes your behavior in the room — not when your notes look encyclopedic. Use these outcome tests:

• Identity test: you can state employer and role without hedging, or you know exactly which clarifying question to ask first.
• Relevance test: your question could only be asked of this person, not a random peer in the same job title.
• Source test: you could cite the public artifact if they asked “where did you see that?”
• Comfort test: nothing in your notes would feel invasive if read aloud.

If you fail the identity test, stop polishing talking points. Fix identity first. If you fail the comfort test, delete the line — it will not help the meeting.

Calendar hygiene that makes research faster

Many “I only have a name” emergencies are scheduling problems. Small invite improvements cut research time in half:

• Require company or email on external holds when your team controls the invite.
• Paste LinkedIn or team-page URLs into the description when known.
• Note mutual introducers so researchers can ask for a unique clue quickly.
• Flag high-stakes meetings so people allocate 15 minutes instead of three.

When invites stay sparse, DeepSearch Name still helps — but filters from the email domain and introducer context remain the highest-leverage human inputs. Product paths: find someone by name, people search.

After the meeting: close the loop

Research that never updates the CRM decays. Spend sixty seconds after the call:

1. Correct title/employer if they clarified live.
2. Keep the two source URLs that still matter; drop dead ends.
3. Note the public artifact you referenced so teammates do not rediscover it.
4. Never promote prep notes into a screening file after the fact.

Pre-meeting checklist

1. Right person confirmed with two signals
2. One public artifact reviewed
3. One informed question written
4. Sources saved
5. No invasive personal digging
6. Screening decisions routed to proper tools if needed
7. Notes labeled as prep, not as a consumer report
8. Post-call CRM correction if reality differed from sources

Compare research surfaces: vs LinkedIn, vs Google. Try a search above when you want candidates + a sourced brief fast.