Prospects, CRM, and Pipeline

Audience: Operators reviewing research output and managing customer records.

Goal: Explain prospect review, CRM import, and pipeline / task workflows.

Prospect review

Open Prospects to decide which AI-researched candidates become real CRM records. Use the same Businesses / Contacts toggle as Research so company and person queues stay separate.

Every prospect on your list belongs to your organization. You see prospects your own runs found, plus prospects a super admin assigned or shared to you. An assigned or shared prospect is your own copy with company details only; it has no campaign and does not show who found it or their evaluation, contacts, or outreach. Accept, reject, contacts, and outreach on your prospect never change another organization's view. A rejection updates campaign targeting only when you belong to the organization that owns the campaign.

Research itself is free. By default (no org unit prices), operators see full prospect identity with no purchase step. When a super admin sets lead unit prices for the organization, unpaid results for that kind are locked: identity fields (names, emails, phones, websites, sources/evidence) are hidden. You still see scores, industry/geography context, fit reasoning, and signals explaining why each candidate was selected. Select locked rows and Purchase selected to unlock them via Square checkout; only the IDs you pay for become fully visible.

What you see

Company details and website (Businesses) or person / profile details (Contacts) — after unlock when billing is active
Evidence and fit reasoning (evidence/sources after unlock when billing is active)
Buying signals with confidence
Urgency and priority scores
Contact hints / ranked contacts and contact quality (after unlock when billing is active)
Company channels and locations when available
Qualification facts when the B2B deep pipeline produced them
Employer affiliation on account-consumers person prospects when company context exists
Locked badge when identity is hidden pending purchase

Long prospect lists are paginated. Open Review details for the full prospect screen.

Common actions

Filter by status, campaign, or intent; Recommended only to focus on top-N ranked qualified businesses
Select locked prospects and purchase them (Square hosted checkout) when org unit prices are set
Accept & find contacts (single or bulk) for qualified businesses — queues Contact Resolution (persona-aware provider waterfall); status becomes accepted while it runs
After contacts are ready, Approve / Approve & import to CRM (or Reject poor fits)
Reject asks why. Every reject stores the reason. Choose Reject this prospect only to skip this company without changing targeting, or Reject and apply to future runs for ICP reasons (wrong industry, geography, size, organization type, persona, poor evidence, or other) so the next research run of a campaign built for your organization excludes similar companies. Already a customer / client and Duplicate never change targeting. Prospects sent from another organization’s campaign also skip targeting changes.
Override & approve when status is system-rejected and you disagree after reading the rejection reason
Choose a primary contact; optionally edit name, email, phone, or role before import
Import approved prospects into CRM (requires unlock when billing is active)
Prefer Accept → Resolve over ad-hoc provider panels for the primary contact path
Generate message drafts or outreach variants when buying signals exist
Request the right contact when you only have a company channel
Close the loop — approve & import, send outreach, enroll in a sequence, or create a booking from one screen
Rerun research from a candidate’s campaign when fresher results are needed
Export CSV for offline review (requires unlock when billing is active)

Only approved prospects should be imported. Imported candidates are marked so the same candidate is not imported again as a duplicate. Research does not find contacts during company Deep Research — spend starts after you Accept a business.

There is no automatic approve/reject from interview policy alone — human review is required before CRM import.

Ad leads

A click on a Meta or Google ad opens the Ad leads page for the research campaign named in that link. The page asks for first name, last name, email, phone, and marketing consent. Business name, address, and website are optional. It does not start a call or a text, and it does not write CRM.

The campaign’s product mode decides the prospects. A missing mode is treated as B2B. A B2B submit creates a business and a person as its contact. A B2C submit creates the person, and also a business when the visitor entered a business name, address, or website. A blank business name uses the website host, without a leading www, then the email domain. Free-mail domains such as gmail.com are skipped. If neither yields a name, the person’s name is used. The address is the business location. The person’s company is that business name. Sending the same lead again does not create a second set.

After the visitor sees thank-you, empty business fields may be filled from public research when a company name or website is present. A failure leaves the submitted business as it was. Approve and import is still how the prospect reaches CRM. A Meta Instant Form lead stays a person prospect and does not use this page.

Prospect Repository

The Prospect Repository is Aventora/CGH’s private platform asset. It is not a tenant workspace screen.

Repository Prospect is canonical inventory. Campaign Evaluation scores that organization against a Campaign Specification. After a tenant’s own campaign run qualifies a candidate, that row appears on tenant Prospects immediately. Names and contacts stay hidden until lead-unlock purchase when unit prices are set. Super Admin release remains for allocating repository inventory and for revoke; it is not required to view your own research run.

Tenant Prospects lists that tenant’s campaign candidates. They do not include other tenants’ prospects, leftover matching counts, or repository-internal discovery fields.

Platform Super Admin uses Repository to inspect identity, discovery provenance, campaign evaluations, evidence, and delivery/revoke history.

Enrichment

Primary path for B2B businesses: Accept & find contacts runs a single Contact Resolution job (Perplexity first, ZoomInfo when the add-on is active) and stops when persona + channel quality requirements are met. “Verified” here means provenance and quality tiers — not SMTP/carrier verification.

On the prospect detail screen (when providers are enabled), advanced panels remain available:

Action	Use when
Save enrichment	Manual edits to company / person fields or fit reasoning
Enrich Prospect (Perplexity)	Backfill company phone, email, socials, locations, firmographics
Find Contacts / Enrich Contact (Perplexity)	Manual rediscovery outside the Accept waterfall
Enrich Contact with ZoomInfo	Org has ZoomInfo add-on Active — review Current vs proposed fields, then accept/reject

CRM contacts support the same Perplexity / ZoomInfo contact enrichment panels. Prefer rejecting on weak fit/evidence, not only on a missing mobile number.

CRM import

Import creates or updates:

Organization (company) — Businesses prospects; Contacts may still link an employer when affiliation exists
Contact (primary person)
Opportunity (pipeline item)
Timeline activity linking the research origin

Metadata stores CRM IDs to prevent duplicate imports. Normal CRM lists are tenant-scoped and exclude soft-deleted records.

CRM records

Use CRM to manage:

Record	Typical fields
Contacts	People, emails, phones, language, source, score, status, tags
Organizations	Companies, lifecycle stage, website, industry, location, size, revenue
Opportunities	Stage, value, probability, close date, linked contact and organization
Cases	Service, retention, billing, and technical issues
Timeline	Notes, messages, calls, meetings, tasks, workflow activity

Typical CRM flow

Search or filter records.
Open a record for details and timeline.
Update profile information when new facts are confirmed.
Create or update opportunities as conversations progress.
Use tasks and cases for follow-up work.

Additional CRM operations supported in the product include soft delete / restore and duplicate / merge workflows for cleanup.

Pipeline

Pipeline tracks deal progression:

View opportunities by stage
Track expected value and probability
Update stage as work progresses
Review expected close dates
Keep pipeline rows tied to contacts and organizations

Pipeline values feed executive reporting when analytics and CRM records exist.

Tasks

Tasks track operational follow-up:

Open, overdue, and completed work
Filters by status or overdue state
Priority, due dates, reminders, and owners
Operator metrics such as overdue tasks, urgent tasks, failed jobs, and hot leads waiting

Tasks are most useful when CRM, inbox, and automations consistently create or update follow-up actions.

Changelog

Date	Change
2026-10-08	An Ad leads submit creates a business and its contact when the research campaign is B2B. A B2C submit creates a person, and a business when the visitor entered a business name, address, or website. The form does not write CRM.
2026-09-29	An assigned or shared prospect is now your organization's own copy with company details only. Other organizations' evaluation, contacts, outreach, and disposition are never shown, and your decisions never change theirs.
2026-09-23	Prospects can be assigned to an organization without copying the record. Accept and reject stay on that organization's assignment.
2026-09-18	Campaign owners can reject a prospect without applying the reason to future runs. Already a customer/client never changes targeting.
2026-09-18	Reject reasons are always stored. Campaign spec/brief learning applies only when the rejector’s organization owns the campaign.
2026-09-18	Rejecting a prospect requires a reason. ICP reasons update the campaign specification and research brief for the next run; already-a-customer and duplicate skip that company only.
2026-09-17	Tenants see their own campaign research results on Prospects without Super Admin release. Identity stays redacted until lead-unlock purchase when unit prices are set. Super Admin revoke still hides a row. Repository remains platform-private.
2026-08-31	Super Admin can revoke a release. Tenant Prospects never include repository discovery queries, other campaign evaluations, or unreleased inventory. Delivery history on Repository detail is the commercial audit trail.
2026-08-31	Prospect Repository is platform-private. Tenant Prospects lists show only commercially released/delivered prospects. Super Admin audits delivery history and releases entitlement.
2026-08-30	Repository admin surface shows identity, discovery provenance, campaign-use history, and per-campaign scores. Recommended is campaign-evaluated top-N, not a global repository ranking.
2026-08-13	Pre-approval Engage eligible is not generally available; removed from the public prospects guide until managed-campaign validation completes.
2026-08-12	Pre-approval engagement: optional campaign policy (default off), Engage eligible first-touch, prospect-linked conversation preserved on Approve & import, human lead disposition.
2026-08-12	Accept → Contact Resolution waterfall; recommended vs all qualified; contact-light research
2026-07-28	Default unlimited access when org unit prices are unset; paywall only when prices are set.
2026-07-28	Lead unlock billing: free research with redacted results; cherry-pick purchase via Square.
2026-07-28	Added Businesses/Contacts toggle, person prospects, enrichment panels, Close the loop, override & approve.
2026-07-21	Initial public prospects, CRM, and pipeline guide.

Prospect review​

What you see​

Common actions​

Ad leads​

Prospect Repository​

Enrichment​

CRM import​

CRM records​

Typical CRM flow​

Pipeline​

Tasks​

Changelog​