Goal: Build a Zapier flow that deduplicates incoming CRM leads, enriches them via API, and posts an actionable Slack alert. Below is a concrete, testable Zap you can implement today.

High-level flow (one Zap or split into two Zaps):
1) Trigger: New or Updated Lead in CRM (e.g., HubSpot "New Contact", Pipedrive "New Person", Salesforce "New/Updated Record").
2) Normalize key fields: Formatter by Zapier — transform email to lowercase; strip non-digits from phone; extract domain from email.
3) Search existing CRM record: Use CRM’s “Find Contact/Person” search by email (exact) and by phone (exact). If your CRM supports fuzzy search, include name+company bucket.
4) Path or Filter logic:
- Path A (Exact match found by email or phone): go to Merge/Update path (auto-merge or update).
- Path B (No exact match but high-similarity results): send to manual review path.
- Path C (No match): create new lead.
5) Enrichment via API: Use Webhooks by Zapier - POST to enrichment service (Clearbit, FullContact, or your enrichment endpoint) sending {email, phone, domain}. Map enriched fields back.
6) Update CRM: Use Update Contact action to write enrichment fields (job title, company, company size, LinkedIn, confidence score). When merging, update surviving record and add a note listing merged record IDs.
7) Slack alert: Send a Slack message to a private #growth-dedupe channel (or DM) with concise info: action taken (merged/created/flagged), lead name, email, link to CRM record, enrichment highlights, and reason (exact email match / fuzzy). Include buttons or emoji for quick ack.

Concrete trigger/filter examples
- Trigger: HubSpot — New Contact.
- Formatter — Email -> Lowercase.
- Find Contact (HubSpot) -> Search by Email. If found -> Path A.
- If not found -> Find by Phone. If phone found -> Path A.
- If phone/email not found -> Use “Find Contact by Company + Name” (if available) and apply a Filter with conditions: name similarity > 0.8 OR company domain match -> Path B.

Auto-merge vs manual criteria (decision criteria)
- Auto-merge if: exact email match OR exact phone match OR enrichment returns a high-confidence score (>= 90%) AND domain match.
- Flag for manual review if: only name similarity, conflicting emails, or enrichment confidence below threshold.
- Create new lead if: no matches and enrichment doesn’t contradict.

Retry and robustness tips
- Use Zapier’s built-in task history to replay failed runs. For transient API errors (429/5xx), add a Delay action and re-call webhook (exponential backoff: Delay 1m -> 5m -> 15m via Paths).
- Use conditional Filters to stop Zap early if required fields missing (no email & no phone).
- Log every enrichment failure to a Google Sheet for manual follow-up.
- Monitor task usage—Zapier charges per task; batching rules reduce calls.

Practical checklist before turning Zap on
- [ ] Identify primary dedupe key (email > phone > domain+name).
- [ ] Build Formatter steps to normalize inputs.
- [ ] Add Find actions for exact and fuzzy matches.
- [ ] Configure Webhooks POST to your enrichment service; map response fields.
- [ ] Implement Paths and Filters for auto-merge vs manual.
- [ ] Create Slack message template and channel privileges.
- [ ] Test with 10 realistic samples (exact match, fuzzy, conflicting, API error).
- [ ] Set up error logging (Google Sheet) and alerting for failures.

Best-for / Avoid-if
- Best for small teams (1–5) with low-to-medium lead volume, limited engineering resources, and need for quick wins.
- Avoid if you have very high volume (tens of thousands/day), strict PII/legal constraints, or need enterprise-grade dedupe logic — move to a dedicated MDM/tooling.

Recommendation: Start simple — exact-match email dedupe + enrichment -> auto-update + Slack alert. After 1–2 weeks of logs, add fuzzy logic and retry backoff. If you want, use Zapier’s “Webhooks” action to connect to any enrichment API and test mapping.

Tool: zapier

Quality score: 95

Compare Zapier and Make