The five questions that decide it
Two or more 'buy help' answers and DIY becomes a false economy. Zero or one, and you're better off spending the budget on a pre-launch audit.
- 01Are you migrating records with history you cannot afford to lose? If yes, buy help — reconciliation is a discipline, not a checkbox.
- 02Do you need an integration that isn't a native marketplace app? If yes, buy help.
- 03Are two or more departments going live at once? If yes, buy help — the cross-team data model is the hard part.
- 04Does someone internally have real HubSpot admin experience and 10+ hours a week? If no, buy help.
- 05Is your revenue reporting going to be built on this portal? If yes, at minimum buy a data-model review.
The hybrid nobody talks about: audit-first
Your team builds. An architect reviews the data model, migration mapping and automation before go-live and hands back a prioritised fix list. You keep the knowledge in-house, you keep control of the build, and you don't ship a structural mistake.
It's the cheapest useful engagement we offer, and for a lot of teams it's the right one.
What DIY teams consistently get right and wrong
| Usually fine in-house | Usually goes wrong in-house |
|---|---|
| Forms, templates, email sends | Data model and property governance |
| Basic pipelines and stages | Lifecycle definitions shared across teams |
| Marketplace app connections | Migration mapping and reconciliation |
| Simple notification workflows | Routing logic with fallbacks and edge cases |
| Standard dashboards | Attribution and forecast-grade reporting |
If you do DIY it, do these five things
- Write the property register before you create a single property.
- Agree lifecycle definitions in one document signed by sales and marketing.
- Clean and de-duplicate the source data before import, then reconcile after.
- Apply a workflow naming convention from workflow one.
- Book training and a two-week hypercare window in the calendar before launch.
Common mistakes
- Deciding based on budget alone rather than on what's hard to reverse.
- Assuming the person who set up your last CRM will find HubSpot equivalent — object models differ.
- Buying a full engagement when an audit would have done.
- Buying nothing when a migration is in scope.
- Starting the build before anyone owns the portal internally.