TPG has cleaned up enough failed HubSpot CRM launches to know exactly what gets skipped. One client declared go-live with 12,000 duplicate contact records. Another went live with a Salesforce sync running in both directions with no source-of-truth logic, creating a data conflict loop that corrupted 6,000 records within 72 hours. These are not edge cases. They happen when teams skip the pre-launch checklist.
What follows is the complete 47-item pre-launch verification list we use before every CRM go-live.
The average cost of a failed CRM launch is not the license fee. It is the time your sales team spends working around a broken system, the pipeline data you lose during the cleanup period, and the six to twelve months of adoption credibility you forfeit when reps decide the CRM can't be trusted. We have seen companies spend $40,000 to $80,000 on post-launch remediation that a proper pre-launch review would have prevented for under $10,000.
"Every item on this list exists because we've seen what happens when it's missing."
Run every check below before you schedule your go-live announcement.
1. Duplicate rate below 5% Run a duplicate report on contacts and companies. Any duplicate rate above 5% requires a deduplication pass before import. Importing duplicates and merging after go-live is a two-to-four-week cleanup project that happens while your sales team is trying to use the system.
2. Required fields are populated on 90%+ of records Identify the fields required for your pipeline reporting and lifecycle stage logic. Run a completeness report. Records with missing required fields will produce incorrect reports and trigger workflow errors on import.
3. Email address format validated HubSpot uses email as the primary contact identifier. Contacts without valid email addresses cannot be deduped reliably and will create association errors. Remove or quarantine records with malformed email addresses before import.
4. Company domain standardized HubSpot's company association logic uses domain. If your domain field includes "www.", trailing slashes, or protocol prefixes inconsistently, association matching will fail and contacts won't link to their companies correctly.
5. Import file field mapping documented and verified Every column in your import CSV must map to a specific HubSpot property. Document the mapping before running the import. Run a test import with 100 records and validate that every field landed in the correct property with the correct data type.
6. Excluded records identified and removed Contacts that should not be in your HubSpot instance (unsubscribed, competitors, invalid domains, records from closed accounts) should be identified and excluded from import. Importing them creates data hygiene work immediately after launch.
7. Historical deal data decision documented Decide explicitly whether you are importing historical deals, what date range you are including, and how historical deals will be staged. Importing all historical deals without a clear staging decision creates pipeline reports that mix historical and current data.
8. Post-import association validation completed After import, run a validation check to confirm that contacts associated to the correct companies and deals associated to the correct contacts. Run a sample of 50 records manually against your source data. Fix association errors before go-live.
9. Naming convention applied consistently Every custom property should follow the same naming convention. A common standard: [category]_[descriptor] in lowercase with underscores. Inconsistent naming makes it impossible to find properties in lists and creates confusion for admins who join after launch.
10. No redundant properties Search for properties that capture the same information under different names. Redundant properties create data fragmentation and split reporting. Consolidate before go-live.
11. Dropdown values standardized Review all dropdown and radio select properties. Confirm that option values are mutually exclusive, cover all valid options, and use consistent capitalization and formatting. "New York" and "new york" are different values in HubSpot reporting.
12. Required properties defined in HubSpot Properties that are required for deal creation, contact creation, or stage advancement should be marked required in HubSpot. Required field enforcement in the system is more reliable than training.
13. Property groups organized logically Custom properties should be organized into property groups that match how your team thinks about the data. A property group called "Sales Process" should contain deal-related fields. A group called "Firmographics" should contain company size, industry, and revenue fields. Unorganized properties create friction for reps entering data.
14. Deprecated properties archived, not deleted Properties that existed in your previous system but are not being carried forward should be archived in HubSpot's property manager, not left active in your property list. Active unused properties create confusion. Deleted properties cannot be recovered if you need the data later.
15. Field-level help text populated on required properties Add help text to every property that requires data entry. Help text appears to reps when they hover over the property label. Clear help text reduces incorrect data entry and reduces questions to your admin.
16. Stage exit criteria documented and communicated Every pipeline stage should have written exit criteria that define what must be true for a deal to advance. These criteria should be in writing and shared with every sales rep before go-live. Without exit criteria, reps make inconsistent staging decisions that corrupt your forecast.
17. Probability percentages reviewed with sales leadership The probability percentage assigned to each pipeline stage drives your weighted pipeline forecast. Review these with your VP of Sales and confirm they reflect actual historical close rates. Default HubSpot probabilities are rarely accurate for your specific business.
18. Closed won and closed lost stages configured correctly Closed won and closed lost are not just stages. They trigger lifecycle stage changes, deal property population, and potentially integration syncs. Confirm that both stages are configured with the correct automation and that the triggers are tested.
19. Multiple pipelines validated if applicable If you have multiple products, sales motions, or customer segments that require separate pipelines, confirm that each pipeline is configured independently with its own stages, probabilities, and automation. Do not assume that automation built for one pipeline applies to others.
20. Deal rotation rules tested If you are using automatic deal rotation to assign new deals to sales reps, test the rotation logic with test records before go-live. Confirm that deals are routing to the correct users and that the rotation resets correctly after each assignment.
21. Pipeline access permissions verified by role Confirm that reps can only view and edit deals in the pipelines they own. If you have separate pipelines for different teams, reps should not have access to pipelines outside their purview. Verify this with a test account for each role before go-live.
22. Sync direction documented for every field in every integration For each integration, document whether each synced field flows from HubSpot to the connected system, from the connected system to HubSpot, or bidirectionally. Bidirectional sync with no source-of-truth logic creates conflicts. Every bidirectional field needs a conflict resolution rule.
23. Test sync run completed with sample records Before activating any integration for your full dataset, run a test sync with a set of 20 to 50 test records. Verify that the correct fields populated in the correct direction with the correct values. Verify that records matched correctly without creating duplicates.
24. Integration error monitoring configured Every HubSpot integration produces an error log when sync failures occur. Confirm that your HubSpot admin has access to integration error logs and that there is a process for reviewing and resolving errors within 24 hours of occurrence.
25. Salesforce sync configuration reviewed (if applicable) HubSpot-Salesforce sync is the most complex integration in the HubSpot ecosystem. If you are running it, confirm that your sync settings explicitly define which system wins on every synced field, that you are not syncing fields you don't need, and that your contact matching logic is set to email address, not name. Name-based matching creates duplicate records.
26. Enrichment tool sync schedule defined If you are using ZoomInfo, Clearbit, or another enrichment tool, define the sync schedule and the fields it is allowed to overwrite. Enrichment tools that overwrite manually entered data without a gating rule will frustrate your sales reps and corrupt clean data.
27. Billing or ERP sync tested end-to-end If HubSpot is syncing with your billing system or ERP, run an end-to-end test: create a test deal in HubSpot, mark it closed won, and confirm that the correct record and data populated in the connected system within the expected time window.
28. Webhook configurations documented If you are using webhooks to push HubSpot data to external systems, document each webhook endpoint, the trigger event, and the expected payload. Test each webhook with a manual trigger before go-live.
29. Integration credentials use a service account Integration credentials should be tied to a service account, not an individual user's account. When individual employees leave, integrations that used their personal credentials break. Use a dedicated HubSpot admin user as the service account for all integration credentials.
30. Every active workflow has a documented owner Each workflow should have a designated owner who is responsible for its performance and maintenance. Workflows without owners get forgotten. They run on stale logic, create incorrect data, and nobody knows who to contact when they break.
31. Enrollment criteria tested with known records For every active workflow, identify at least two records that should enroll and two that should not. Verify that the enrollment criteria correctly includes and excludes the right records. Enrollment logic errors are among the most common causes of incorrect data in HubSpot.
32. Workflow action sequence reviewed for logic errors Review the action sequence of every active workflow for conditional logic errors. Common mistakes include branches that overlap, delay timing that doesn't match your sales cycle, and actions that fire on records that should have been excluded upstream.
33. Re-enrollment rules verified Decide explicitly whether each workflow should allow re-enrollment. If a contact completes a workflow and then meets the enrollment criteria again, should they go through the workflow again? The default answer is no for most sales automation. Make sure re-enrollment settings match your intent.
34. Email sending workflows tested with internal addresses Any workflow that sends emails should be tested with internal email addresses before go-live. Confirm that the correct template populated, that personalization tokens resolved correctly, and that the sending domain and from-name are correct.
35. Conflict between overlapping workflows identified Identify any workflows that could enroll the same contact or deal simultaneously and create conflicting actions. A common conflict: one workflow sets a lifecycle stage to MQL and another workflow enrolls MQLs and immediately advances them to SQL without a qualifying step.
36. Every user assigned the correct permission set Review every user's permission set against the permission matrix defined during architecture. Confirm that no rep has access to properties, pipelines, or reports they should not see. Confirm that every admin has the access they need without unnecessary access to billing or security settings.
37. Sales manager visibility to direct reports confirmed Sales managers need to view activity, deals, and pipeline for their direct reports. Confirm that team hierarchy is set up correctly in HubSpot and that managers can access their team's data without being able to edit another manager's team data.
38. Email integration configured for every rep If reps are logging email activity through HubSpot's Gmail or Outlook integration, confirm that every rep has connected their inbox before go-live. Reps who haven't connected their inbox on day one will log emails manually for two weeks and then stop logging entirely.
39. Mobile app tested on iOS and Android If any of your reps use HubSpot on mobile, test the mobile app experience with your specific configuration before go-live. Confirm that required fields display correctly on mobile, that the deal view shows the properties reps need, and that the activity feed is accessible.
40. Notification settings reviewed with each rep HubSpot's default notification settings generate a high volume of emails and in-app notifications. Review notification settings with reps before go-live and configure them to surface only the notifications that are actionable. Notification fatigue drives reps to ignore HubSpot alerts entirely.
41. Core pipeline report validated against known data Build your primary pipeline report and validate it against data your sales leadership trusts. Use a specific time period and compare the HubSpot report output to a known source: a recent board report, a Salesforce export, or a manually verified pipeline review. Identify and resolve any discrepancies before go-live.
42. Forecast report uses correct probability configuration Confirm that your forecast report pulls probability from your pipeline stage settings and that the stage probabilities were set correctly in item 17 above. A forecast report using default probabilities will produce numbers that don't match your actual business.
43. Activity reports calibrated to your team's workflow Build activity reports for calls, emails, and meetings logged. Run these reports against a period where you have complete data from a previous system. Identify gaps between expected and reported activity counts. Gaps indicate that some activity types are not being logged correctly.
44. Source attribution configured correctly If you are tracking original source and lead source in HubSpot, confirm that the tracking code is installed on every form and landing page, that UTM parameters are being captured, and that your source values match your campaign naming convention.
45. Dashboard access configured by role Confirm that each role's default dashboard displays the reports most relevant to their work. Sales rep dashboards should show their individual pipeline, activity quota, and open tasks. Manager dashboards should show team pipeline and rep performance. Executive dashboards should show company-wide pipeline, forecast, and source attribution.
46. Report sharing permissions reviewed Confirm that reports containing sensitive data (individual rep compensation targets, deal-level financials) are restricted to the appropriate roles. HubSpot's default report sharing is relatively open. Review sharing settings on your 10 most sensitive reports before go-live.
47. Scheduled report delivery tested If you are using HubSpot's scheduled report delivery to email reports to leadership on a weekly or monthly cadence, test the delivery by scheduling a one-time send to internal addresses. Confirm that the report populated correctly and that the email arrived at the correct recipients.
Pre-Launch Checklist Callout Print this checklist. Assign each item to a specific team member with a due date. Do not mark an item complete until it has been tested, not just configured. For items involving reports or integrations, test means you reviewed the output against known data, not just confirmed the configuration exists. A checklist that gets marked complete without testing is not a checklist. It's a list of things someone intended to do.
This checklist takes two to three days to run thoroughly. Budget that time in your project plan. If your implementation timeline doesn't include three days of pre-launch verification, the timeline is wrong.
TPG runs this checklist as part of every CRM implementation we lead. We also offer it as a standalone pre-launch review for companies that implemented HubSpot internally and want an independent verification before go-live or before expanding to additional teams. That engagement takes three to five business days and produces a written remediation report for any item that fails.
Schedule a Pre-Launch CRM Review
How long does it take to run this checklist? Two to three business days for a thorough review, assuming a single mid-market implementation with one to three integrations and a team of 20 to 50 sales users. Larger implementations with more integrations and more complex reporting configurations take proportionally longer. Do not compress this into a single day. Items rushed through in a single review session will have errors that surface within the first two weeks of operation.
Who should run this checklist? The project lead for the implementation should own the checklist. Individual items should be assigned to the team members with direct knowledge of the configuration. Data items go to the person who ran the import. Integration items go to whoever built the integrations. Reporting items go to whoever built the reports. The project lead validates completion, not just assignment.
What happens if we fail multiple items? Do not go live. Fix the failing items, retest them, and reschedule go-live. A one-week delay to fix pre-launch issues costs far less than three months of post-launch cleanup. Sales leadership may push back on the delay. Show them the specific items that failed and what goes wrong if you launch with them unresolved.
Do we need to run this checklist for every HubSpot update? Not the full checklist. For major changes like adding a new pipeline, connecting a new integration, or rebuilding your lifecycle stage logic, run the relevant categories. For minor changes like adding a property or updating a workflow, run a targeted test of the specific change. The full 47-item checklist is for initial go-live and for major architectural changes.
Is there a version of this checklist for HubSpot Marketing Hub? The data, integration, automation, and reporting categories apply to Marketing Hub with minor modifications. The pipeline and user setup categories are specific to CRM. TPG publishes separate pre-launch checklists for Marketing Hub Professional and Enterprise configurations. Contact us for the Marketing Hub version.
What is the most commonly failed item on this checklist? Item 22: sync direction documentation for integrations. Most implementations connect integrations without explicitly documenting which system wins when both systems have a value for the same field. This creates silent data corruption that often isn't detected until a key report shows numbers that don't match reality.
The Pedowitz Group | pedowitzgroup.com | Revenue Marketing Experts Since 2007