Skip to content
Book a build

HubSpot CRM implementation: phases, timeline and checklist

HubSpot CRM implementation by phase: timelines, what HubSpot does natively vs a partner vs Jetstack AI, CRM migration services and an acceptance checklist.

Jetstack AI / Working guides

The CRM implementation plan

Plan before building

Approval sequence
  1. 01
    Review the planScope, schema and dependencies
  2. 02
    Approve generationDefinitions to inspect, not live assets
  3. 03
    Approve deploymentA separate decision for this portal
  4. 04
    Read the resultsReceipt, read-back and manual acceptance
On this page

Updated October 29, 2026.

What does a HubSpot CRM implementation involve?

A HubSpot CRM implementation turns a portal into a working system: you decide the operating rules and data model, build the properties, pipelines, lists, workflows and reports, move data in, connect other tools, and test before anything is switched on. On Professional and Enterprise editions, HubSpot’s required onboarding guides the first 60 to 120 days. The build itself is done by your team, a Solutions Partner or a tool like Jetstack AI Projects.

A useful implementation brief describes what the portal must do, not just which assets to create. “Build lead routing” leaves the difficult decisions implicit: what counts as a lead, how regions are assigned, which records are excluded and who resolves an exception. Write those decisions down before asking a person or an AI tool to build the assets.

HubSpot implementation phases and timeline

PhaseWhat happensMatching phase in HubSpot’s sample onboarding plans
1. Goals and operating rulesPick up to three goals; capture triggers, exclusions, fallbacks and a decision ownerGoal setting with your HubSpot consultant
2. Account and tech setupUsers, tracking code, sending domain, inboxes, currenciesPhase 0: account and tech setup
3. Data modelObjects, properties, option values, lifecycle stages, pipeline stagesPhase 0 to 1: custom properties, lifecycle stages
4. Build in dependency orderProperties, then lists, then workflows, forms, emails and reportsPhases 1 to 3: the goal-specific work
5. Migration and integrationsImport or migrate records; connect other systemsPhase 0: importing data, with consulting on migration
6. Test and switch onReview the build, test with controlled records, turn workflows on in stagesNot covered as a separate phase
7. HandoffRecord what was built, what is verified and what is still openPhase 4: next goals and customer success owner

How long it takes. HubSpot’s sample plans say customers typically complete onboarding within 2 months for Sales Hub, within 3 months for Marketing Hub and within 2 months for Service Hub Professional. HubSpot’s services catalog sets delivery periods of 60 days for Sales Hub and Service Hub Professional onboarding, 90 days for Marketing Hub and every Enterprise hub, and 120 days for Customer Platform. Migration and integrations usually decide how far past that a full implementation runs, because they depend on the systems you are leaving or connecting. Fees, waivers and per-hub detail are in the HubSpot onboarding services guide.

What clients pay partners for has shifted toward this work. In Jetstack AI’s 2026 analysis of Solutions Partner reviews, implementation and setup appeared in 45.3% of the reviews that described the work, and migration and data import in 12.0%. These are shares of mentions, not measures of project scope.

What HubSpot does natively vs a partner vs Jetstack Projects

WorkHubSpot nativelySolutions PartnerJetstack AI Projects
Implementation planOnboarding consultant plans up to three goalsScoped in the statement of workJetty turns your brief into a plan you review
Building properties, pipelines, lists, forms, emails, workflowsYou build them in HubSpot’s tools, with guidanceUsually builds themBuilds them in dependency order after you approve the plan, then deploys after a second approval
Reports, dashboards, sequences, lead scoresYou build themUsually builds themBuilds them through the Chrome extension
Data import and migrationCSV import, Smart Transfer, paid migration servicesOften offeredNot supported
IntegrationsMarketplace apps and data syncOften offeredNot supported
CMS pages, themes, custom codeYou or a developerOften offeredNot supported
Evidence of what was builtWhat you can see in the portalVaries by partnerA receipt of what HubSpot accepted, read back after the run

Projects is bound to one portal, creates workflows switched off, and can change only the assets it built. For a workflow or asset built elsewhere, the route is to build the corrected version as a new asset with a Project, then retire the old one in HubSpot. For a side-by-side with building by hand, see Projects vs manual implementation.

1. Confirm the portal and the operating rule

Name the destination portal and whether it is sandbox or production. A Project works with one portal; use separate projects for separate environments. Confirm the HubSpot subscription and permissions required by the intended features. An asset type appearing in a product list does not mean that every connected portal is entitled to use it.

Interview the person who will operate the process. Capture the trigger, the expected action, the exclusions and the fallback. For routing, ask what happens when a contact has no location, belongs to more than one territory or already has an owner. If nobody can answer, keep that ambiguity visible rather than letting generated defaults become policy.

2. Write a small schema contract

Define object and property names, field types and option values before describing automations. Treat internal names as interfaces: a label can be edited while a consumer still depends on the same internal field. Check the live portal for suitable existing definitions before proposing new ones.

This sample brief is a planning example, not an importable Jetstack specification:

process: Enquiry routing
destination: Example Service Portal
object: contact
property:
internal_name: service_region
type: enumeration
allowed_values: [north, south, unassigned]
rule: Route only eligible contacts with a resolved region
exclusions: Existing owner; suppressed contact; missing consent decision
fallback: Put unresolved enquiries in a named manual-review queue
acceptance: Test each region, missing region and existing-owner cases

Agree whether the property can be reused as is. A similarly named field with different semantics is not an automatic match. A change to existing options can affect lists and reports outside the immediate brief, so record those consumers as part of the impact review.

3. Review the dependency sequence

Read the plan from foundations to consumers. In a routing example, a property must be available before a list filters on it; that list must be resolvable before a workflow uses it for enrollment. A dashboard is only useful when its reports and underlying definitions agree.

Contact property: service_region
→ Eligibility list: region is known + suppression rules
→ Routing workflow: explicit branch and fallback
→ Verification: inspect references + controlled test cases

Dependencies are not only assets. An owner assignment needs a valid destination owner and a business decision about who should receive the work. An email step needs a reviewed message and the relevant subscription and sending configuration. Keep manual prerequisites beside the asset plan instead of implying they will be resolved by generation.

4. Separate the two approval gates

First, review the plan and authorize generation of the agreed assets. Inspect the generated definitions against the brief. Confirm enrollment, filters, object types, branches and dependencies; a plausible asset name is not sufficient review.

Then decide whether to deploy the reviewed work to the named portal. Deployment is a separate approval because that is the boundary where portal configuration changes. Projects is currently in limited release; check what Projects can build before you promise a particular asset.

5. Handle partial results deliberately

A run may include completed assets, failed assets and consumers blocked by a failed dependency. Read each result rather than assuming the whole plan succeeded or failed together. Keep the run and receipt references so the next reviewer can see what was attempted and what HubSpot returned.

If a dependency is missing, resolve it before retrying the consumer. If a write may have applied but verification did not finish, inspect the destination before creating another copy. Do not use repeated deployment as a substitute for understanding an ambiguous outcome. Recovery should target the unresolved part of the approved scope.

6. Verify configuration and behavior separately

Read-back verification checks the resulting asset against the intended configuration where the asset type supports it. That is different from proving a business process with representative records. Inspect the created references, then test permitted scenarios in a controlled environment with consent from the portal owner.

Projects workflows deploy switched off; activation is a separate decision. Review enrollment and possible side effects before turning one on. For any asset whose verification is limited or manual, write that limitation into the acceptance record rather than implying an automated receipt proves everything.

HubSpot CRM migration services

HubSpot CRM migration services move records from another system into HubSpot. HubSpot offers a self-service tool, paid migration packages and consulting hours; partners and migration vendors cover the rest. Plan migration as its own phase with its own acceptance test, because it is the part of an implementation most likely to slip.

OptionWhat it coversLimits to know
Smart Transfer (self-service)One-time move from Pipedrive, Zoho CRM, Dynamics 365, ActiveCampaign, Copper, Mailchimp, Keap, Salesforce, Account Engagement, Marketo, Zendesk, Freshdesk or Monday.com. Creates properties, deal pipelines and stages, currencies and usersOne-way into HubSpot; Super Admins only; custom field mappings need Data Hub
CRM Data Migration, BronzeUp to 350,000 records, no custom objectsData must come through an API, not flat files; 60-day delivery period
CRM Data Migration, SilverUp to 1 million records, up to 2 custom objects, merging and filteringAPI data only; 60-day delivery period
CRM Data Migration, Gold1 to 5 million records, up to 4 custom objectsFree dataset assessment first; 60-day delivery period
Migration Consulting, 4 hoursUp to four hours of remote consulting or offline workExpires 90 days from purchase
OnboardingConsulting on importing dataAdvice, not the import itself

HubSpot’s catalog says its migration services are delivered remotely and may be performed by a third-party provider. Before you buy any of them, decide which history you actually need, who cleans duplicates first, and how you’ll check counts after the move.

Jetstack AI doesn’t migrate records from another CRM, and Projects doesn’t import data. If you’re moving configuration from one HubSpot portal to another, Portal reuse copies the assets and remaps their references. You can roll back a Portal reuse deployment: Jetstack shows you what will be removed first, then deletes the assets that deployment created in the destination portal. It won’t delete anything another deployment or Project still uses, and it skips assets that were changed after the deployment.

Implementation acceptance

A reading checklist. Selections stay on this page only and are not saved.

Where Jetstack AI fits

For the build phases, use Projects. You describe the work in a brief and chat, Jetty turns it into a plan you review, approving the plan starts generation, and you approve deployment separately. After the run you get a receipt of what was created, checked by reading it back from HubSpot. Projects doesn’t build CMS pages, themes or custom-code actions, and doesn’t import data, manage users or set up integrations, so plan those phases with your team or a partner. Then write the handoff record, so the next admin knows what was delivered and what is still open.

Sources

The implementation workflow

Plan before building.

A reviewable scope with separate generation and deployment gates.

Explore Projects