Duplicate accounts derailed territories; Clearbit and Looker validation synced to Salesforce stabilized coverage

Overview

A retail tech firm’s sales organization fought over territories because the Customer Relationship Management (CRM) system contained duplicate and stale account records. Reps created new accounts when they could not find the right one, enrichment was inconsistent, and coverage maps changed weekly as merges and reassignments lagged. Intelligex built a governed data pipeline that standardized and deduplicated accounts using Clearbit enrichment and custom matching rules, then synchronized cleansed records back to the CRM. A validation workspace in Looker gave Sales Operations and managers a clear view of suspected duplicates, proposed merges, and territory impacts before changes took effect. Territories stabilized, disputes decreased, and reps worked from a clean view of coverage and whitespace—while Salesforce, Clearbit, and Looker remained in place.

Client Profile

• Industry: Retail technology (omnichannel commerce, store operations, and analytics)
• Company size (range): Global go?to?market with enterprise and mid?market segments and regional coverage
• Stage: CRM hygiene handled ad hoc; enrichment used inconsistently; territory rules maintained in spreadsheets; frequent duplicate accounts and stale ownership
• Department owner: Sales & Business Development (Sales Operations/Revenue Operations)
• Other stakeholders: Field Sales and SDRs, Channel/Alliances, Marketing Operations, Finance/Order Ops, Data/Analytics, IT/Integrations, Legal for data governance, Internal Audit

The Challenge

Account records drifted. Legal names, trade names, and franchises were entered differently by rep and region. Multi?brand holding companies spawned separate accounts for each email domain and storefront, and changes to ownership did not cascade across related records. When a lead matched several accounts, reps picked one by guesswork or created a new account, which split activity history and confused pipeline attribution.

Territory assignments broke under the weight of duplicates and stale data. Named accounts appeared more than once with slight spelling differences, regional coverage assigned by country and postal code clashed with franchise records, and partner conflicts were resolved in email. Sales Operations tried manual merges and email campaigns reminding people to search before creating, but incoming records continued to multiply. Disputes consumed time at the start of many deals.

Data stewardship lacked visibility and controls. Enrichment was applied by some teams using a browser plug?in and skipped by others. There was no governed way to see suspected duplicates, to validate merges with affected reps, or to simulate territory impacts before edits locked in. Leadership lacked a reliable picture of coverage and whitespace, and coaching was based on anecdotes rather than on a defensible map of the market.

Why It Was Happening

Matching and enrichment were fragmented. The CRM did not enforce a single path for standardizing names, domains, and addresses or for applying enrichment before account creation. Without a canonical data model and a rules engine for matching, duplicate prevention relied on search discipline and memory.

Territory logic lived in spreadsheets. Named accounts, regional rules, and partner overlays changed with hiring and partner agreements, but the CRM and enrichment workflows were not tied to that logic. As a result, merges and new accounts triggered conflicts after the fact, and ownership ping?ponged between teams.

The Solution

Intelligex implemented a governed deduplication pipeline and validation workflow that standardized, enriched, and matched accounts before syncing changes back to the CRM. Clearbit provided firmographic enrichment, while custom matching rules reconciled legal names, trade names, web domains, and addresses into clusters. A confidence model categorized candidates for auto?merge, human review, or ignore. A Looker validation workspace showed proposed merges, unresolved conflicts, and territory impacts, with maker?checker approvals for high?impact changes. Once approved, merges and reassignments flowed to the CRM, and territory maps refreshed in step. The design used Clearbit for enrichment, Salesforce as the CRM, and dashboards in Looker, with role?based access to ensure the right people validated changes.

• Integrations: Salesforce for account source and sync; Clearbit for firmographic and domain enrichment; data warehouse as the staging area; Looker for validation dashboards and stewardship; identity/SSO for role?based access; collaboration tools for notifications.
• Standardization and enrichment: Normalization of legal names, trade names, and punctuation; website and email domain parsing; address standardization; Clearbit enrichment for company IDs, industry, and hierarchy when available.
• Matching rules: Multi?step matching using domain, name similarity, address proximity, and parent/child hierarchy; franchise and multi?brand handling; suppression lists for ineligible or test accounts; confidence scoring to separate auto?merge from review.
• Territory awareness: Named?account and regional rules encoded; partner overlays and channel exceptions respected; assignment simulation to show owner changes before merges.
• Stewardship workflow: Looker dashboards listing suspected duplicates, proposed clusters, and impacted territories; one?click approve/deny with reason codes; maker?checker on high?impact merges; dispute queue for rep feedback.
• Sync and audit: Controlled merge and reassignment to Salesforce; ownership and history preserved; immutable logs of matches, decisions, and sync results; rollback plan for edge cases.
• Security and privacy: Role?based access to stewardship views; minimal enrichment fields exposed in broader audiences; logs retained per policy; counsel?only notes for sensitive disputes when needed.

Implementation

• Discovery: Cataloged duplicate patterns by segment and region; inventoried territory rules, named accounts, and partner overlays; sampled merge history and disputes; reviewed enrichment usage and gaps; gathered Sales Ops, regional leadership, SDR management, Marketing Ops, IT/Integrations, and Legal requirements.
• Design: Defined the canonical account model and required fields; authored normalization steps; specified matching rules and confidence thresholds; encoded territory simulation and dispute handling; designed Looker stewardship views and approvals; established change control for rules, suppression lists, and territory logic.
• Build: Implemented data ingestion from Salesforce; configured Clearbit enrichment; built standardization and matching jobs; created the confidence model; developed Looker dashboards for validation and territory simulation; wired approvals, notifications, and sync; enabled SSO and audit logging.
• Testing/QA: Ran in shadow mode to compare proposed merges with historical decisions; validated matching across franchises, holding companies, and regional subsidiaries; exercised territory simulation; piloted with one segment and region; tuned thresholds, rule precedence, and dashboard UX from steward feedback.
• Rollout: Enabled stewardship for high?duplicate segments first; expanded to remaining regions; turned on selective auto?merge for high?confidence clusters; retained manual merges as a monitored fallback; tightened rules as accuracy and trust grew.
• Training/hand?off: Delivered guides for stewards and managers on dashboards, approvals, and disputes; trained Sales Ops on rule maintenance and suppression lists; briefed field teams on expected changes and dispute channels; updated SOPs; transferred ownership of rules, dashboards, and logs to Sales Ops under change control.
• Human?in?the?loop review: Scheduled recurring reviews of false positives, disputed merges, and territory edge cases; recorded decisions with rationale and effective dates; updated rules, thresholds, and suppression lists accordingly.

Results

Accounts became clean and predictable. Duplicates surfaced in one place, merges respected territory logic, and ownership changes were simulated and approved before they landed in the CRM. Clearbit enrichment standardized domains and hierarchies, which stabilized matching and reduced the temptation to create new accounts when a lead arrived.

Territories and forecasting aligned. Named accounts and regional coverage no longer splintered across near?duplicate records, partner conflicts were resolved in the stewardship queue with a decision trail, and managers viewed a defensible coverage and whitespace map. The core tools stayed in place; the new layer provided enrichment, matching, validation, and governance between them.

What Changed for the Team

• Before: Reps created new accounts when they couldn’t find the right one. After: Standardized names, domains, and enrichment made the right account easy to find.
• Before: Duplicates caused territory conflicts. After: Matching rules clustered records and merges respected territory and named?account rules.
• Before: Merges happened ad hoc. After: A validation workspace showed proposed merges and territory impacts with approvals and reason codes.
• Before: Coverage and whitespace were debated. After: Looker dashboards displayed clean coverage and gaps grounded in governed data.
• Before: Disputes lived in email. After: A dispute queue collected feedback, and decisions were logged and searchable.
• Before: Enrichment usage varied by team. After: Clearbit enrichment applied consistently in the pipeline before sync.

Key Takeaways

• Standardize first; normalize names, domains, and addresses and enrich consistently before matching.
• Use layered matching; combine domain, name, address, and hierarchy with confidence scoring to separate auto?merge from review.
• Respect territories; simulate assignments and encode named?account and partner rules so merges don’t break coverage.
• Make stewardship visible; provide dashboards for suspected duplicates, approvals, and disputes with reason?coded decisions.
• Log everything; keep an audit trail of matches, merges, and ownership changes to support coaching and audits.
• Integrate, don’t replace; keep Salesforce, Clearbit, and Looker—add the enrichment, matching, and governance layer between them.

FAQ

What tools did this integrate with? The pipeline ingested and synced accounts with Salesforce, applied firmographic enrichment via Clearbit, staged transformations in the data warehouse, and exposed validation and territory simulation in Looker. Access used SSO with role?based permissions.

How did you handle quality control and governance? Matching rules, thresholds, suppression lists, and territory mappings lived under Sales Ops change control with owners and effective dates. Every proposed match, approval, denial, and sync wrote to immutable logs. Maker?checker applied to high?impact merges and named?account changes, and release notes documented rule updates.

How did you roll this out without disruption? The pipeline ran in shadow mode, flagging suspected duplicates without merging. One segment and region piloted the stewardship workflow first, while manual merges remained as a monitored fallback. Auto?merge enabled only for high?confidence clusters after accuracy and adoption stabilized, then coverage expanded.

How were matching rules designed for franchises and holding companies? Rules combined domain, legal name similarity, and address proximity with parent/child hierarchy signals from enrichment. Franchise patterns and multi?brand holdings were modeled explicitly, and low?confidence clusters required human approval with territory simulation before merge.

How did this affect territory assignments and partner conflicts? Territory logic and partner overlays were encoded in the simulation. Proposed merges showed owner changes and conflicts, and approvals captured reason codes. Partner conflicts routed to a dispute queue with context, and decisions updated the rules to reduce repeats.

How was rep feedback incorporated? Reps submitted disputes directly from the Looker view or via a lightweight form linked in notifications. Stewards reviewed context, adjusted clusters, and recorded rationale. Patterns informed rule refinements and suppression list updates.

How did you protect data privacy and sensitive account details? Role?based access limited who could view detailed enrichment fields and dispute notes. Dashboards displayed minimal necessary information, and all views, edits, and syncs were logged. Retention followed policy, with counsel?only notes restricted when necessary.