RevOps playbook · AI Employee: Rev

CRM Data Hygiene & Updates

Duplicate rate < 1% weekly

Install this playbookBack to marketplace
The problem

The CRM slowly rots. Duplicate contacts multiply, phone numbers become disconnected, emails bounce silently, job titles go stale, accounts merge without records reflecting it, contact-account associations drift when people change jobs. Every report reads from data 15-40% stale. Sales trusts nothing except what they personally know. Every quarter someone runs a data-hygiene project that lasts a week and covers 3% of the problem.

At a glance
Trigger
Cron (nightly)
Approvals
Manual review threshold on merges > confidence 0.9
What it does
Writes to your systems
Systems
Salesforce · HubSpot · ZoomInfo · People Data Labs
How it feels in production

An hour-by-hour walkthrough.

Every night, Rev runs data-hygiene passes across the CRM: - Duplicate detection: fuzzy match on email + name + company across contacts + leads + accounts. Merge candidates surfaced with confidence score. - Bounce follow-up: every email bounce (from marketing platform, sales cadence tools) → look up updated contact on LinkedIn / ZoomInfo → propose update. - Job-change detection: LinkedIn Sales Navigator signal on any contact → propose contact-status update + potentially new employer as account. - Company enrichment: missing company size, industry, tech stack → enrich from Clearbit / Apollo / BuiltWith. - Association fixes: contact still associated with old employer 90 days after job change → propose reassign or archive. - Field completeness: missing phone, email, LinkedIn URL, title → enrich where possible. Rev surfaces the day's proposed changes to the responsible owner: for account changes, the AE; for contact changes, whoever most recently touched them; for uncertain cases, the RevOps team. Owner reviews the batch: approve as-is, edit, or skip. Each approval is one click; a good batch of 40 changes reviews in 5 minutes. On approval, Rev writes back to Salesforce / HubSpot with audit trail preserved (who approved, what the prior value was, what evidence supported the change). Weekly digest to RevOps + sales leadership: data quality score, top account gaps, biggest data-loss patterns (bounces per week, disconnected numbers). CRM stays reliable rather than rotting.
How it works

Step by step.

  1. 01

    Nightly data-hygiene scan

    Deduplication, bounce follow-up, job-change detection, company enrichment, association fixes, field completeness. Every pass produces proposed changes with confidence scores.

    Salesforce · HubSpot · LinkedIn · ZoomInfo · Clearbit · Apollo
  2. 02

    Route proposed changes to owners

    Account changes → AE. Contact changes → most-recent-toucher or contact-owner. Uncertain / no-owner cases → RevOps queue. Batch UI for efficient review.

    Slack · Teams · Web UI · Ownership registry
  3. 03

    One-click approval + audit trail

    Owner reviews batch: approve, edit, skip per record. Approvals write back to CRM with audit trail: who approved, prior value, evidence source (LinkedIn URL, ZoomInfo record ID).

    Approval flow · CRM audit fields
  4. 04

    Handle merges + reassignments carefully

    Duplicate merges: preserve all activity history from both records, tie to the surviving ID. Job-change reassignments: preserve prior-employer relationship as history, tag new employer as active.

    CRM merge tools · History preservation
  5. 05

    Weekly digest + data quality trend

    Data quality score trend, top account gaps, weekly change volume, bounce-and-recover rate, deduplication rate. Highlights systemic issues (form allowing duplicates, marketing importing dirty lists).

    Analytics · Slack digest · RevOps dashboard
Systems and wiring

What you connect to make this run.

Salesforce · HubSpot · CRM

read+write

Primary system for records. Writes proposed changes on approval with audit trail. Deduplication merges use platform-native merge tooling to preserve history.

LinkedIn Sales Navigator · ZoomInfo · Apollo · Clearbit

read

External enrichment. Job-change signals, updated titles, verified emails, phone numbers, company data. Rate-limited and cached to avoid API cost spirals.

Marketing platform · Cadence tools

read

Bounce feed from marketing (Marketo, HubSpot, Iterable) and sales cadence tools (Outreach, Salesloft). Every bounce triggers verify + re-enrich workflow.

Data warehouse · Snowflake · BigQuery

read+write

Quality metrics + historical trends. Records marked as "verified fresh" vs. "stale" for downstream analytics + reporting.

What changes

Before and after, honestly.

CRM data freshness (fields verified within 90 days)
Before
30-55%
After
85-95%
Duplicate rate in contacts + accounts
Before
10-25%
After
Under 2%
Email bounce rate on cadence tools
Before
8-18%
After
1-3%
Sales hours per week on data cleanup
Before
3-8 hours per rep
After
10-20 minutes (review daily batch)
Related

Playbooks that pair with this one.

RevOps

Account & Contact Dedupe

Deduplication is one component of hygiene; can be run standalone.

RevOps

Inbound Lead Enrichment & Routing

Enrichment engine shared between new-lead flow and hygiene backfill.

RevOps

Stale Opp Cleanup

Same pattern applied to opportunities (stale, forgotten, mis-classified).

Frequently asked

Answers about this playbook.

Won't approvers rubber-stamp changes they don't fully understand?

Each proposal shows the evidence (source of the update, prior value). One-click approval on high-confidence changes; medium-confidence requires explicit review of evidence. Low-confidence auto-holds for RevOps.

What if enrichment data is wrong (e.g., LinkedIn scraper outdated)?

Multi-source verification for material changes. Job-change requires LinkedIn + ZoomInfo agreement, or a single-source with fresh timestamp. Sole outdated source triggers a hold for manual verification.

How does it handle GDPR / privacy requirements?

Enrichment respects opt-outs + geographic rules. EU contacts require different enrichment path; some data (personal email vs. work) restricted per jurisdiction. Never enriches PII outside legal scope.

Can it fix bad data from historical imports?

Yes — bulk-hygiene mode processes historical records at controlled rate. Prioritises active-pipeline records first; slow-cycle backfill on the tail.

How does it coordinate with the marketing team's data ops?

Shared data ownership. Rev handles CRM records; marketing ops handles campaign engagement data. Overlap zone (contact-status, subscription preference) coordinated through shared change routing.

See it run on your data.

Free plan, no credit card. Connect the systems this playbook needs and run it against a past event first.

Book a demoMore RevOps playbooks