Your team wants something useful to appear right inside the HubSpot CRM record. Maybe a panel that pulls a customer's live account balance, a button that triggers a bespoke approval, or a view of data that lives in a system HubSpot does not natively model. You have two routes to get it there: install an off-the-shelf app from the HubSpot App Marketplace, or build a custom UI extension. This is the classic build versus buy decision, and getting it wrong is expensive in either direction. Buy the wrong thing and you pay a subscription forever for a poor fit. Build the wrong thing and you carry engineering and maintenance you did not need.
This guide gives you a clear way to decide, written for UK B2B teams who run HubSpot and want a defensible answer rather than a gut call.
First, clear up a common misconception
People often assume the dividing line is whether something appears inside the CRM at all. It is not. Both routes can surface interactive functionality directly in the record. Public Marketplace apps can now ship their own app cards powered by UI extensions, so a bought app can render inside the contact, company or deal record just as a custom one can.
So the real decision is not custom interface versus no custom interface. The questions that actually separate the two paths are about data model fit, workflow fit and ownership. Hold onto those three. Everything below comes back to them.
What a UI extension actually is
A UI extension is a custom, contextual interface built into the HubSpot CRM. Engineers build it with React and HubSpot's @hubspot/ui-extensions SDK (TypeScript is supported), and back it with HubSpot-hosted serverless functions when it needs to fetch or write data behind the scenes. It can render in several places, including the record middle column, the record sidebar, CRM preview panels and the help desk ticket sidebar.
One gating fact matters before you go any further. To build UI extensions you need an Enterprise-tier subscription. Since 14 Nov 2024 this opened up to any Enterprise hub, where previously it was limited to Sales Hub or Service Hub Enterprise. The serverless functions behind an extension also require Enterprise (or a free developer test account for building and testing). If your portal is on Starter or Pro, the build route generally means upgrading a tier first, which changes the maths. We cover what bespoke development on HubSpot involves more broadly on our HubSpot development hub, and the build itself on our UI extensions service page.
Why this decision is suddenly more urgent
If your current in-CRM functionality came from an older integration, the ground is moving. The classic CRM cards that powered the old static right-sidebar panels are being superseded by app cards built on UI extensions. As of 16 Jun 2025, no new work can add a classic CRM card, and all existing apps using them must migrate by the final sunset on 31 Oct 2026. HubSpot has released a tool to swap legacy cards for UI-extension app cards without disrupting the customer view, but the direction of travel is settled. Anything you build new in this space is a UI extension by default.
When buying a Marketplace app is the right call
The Marketplace is a genuinely strong option, and for many needs it is the obvious answer. There are well over a thousand integrations available, and the average HubSpot customer uses more than five. Buy when:
- The need is common and well bounded. Syncing an established accounting, calendar, video or e-signature tool is solved territory. Someone has already built it well.
- A certified app already covers it. Certified apps are reviewed by HubSpot's Ecosystem Quality team against security, privacy, reliability, performance, usability, accessibility and value, on a rolling two-year basis that has to be renewed. As a maturity signal, the minimum active installs for certification rose from 6 to 60 on 15 May 2024, so a certified, widely installed app is a known quantity.
- There is no bespoke data model or workflow. If the app fits standard objects and a standard process, you gain nothing by rebuilding it.
- You are happy not to own the code. You want vendor-maintained software that someone else keeps current, and you accept their roadmap.
The hidden costs of buying are real but manageable: an ongoing subscription, a degree of vendor lock-in, dependence on the vendor's roadmap and recertification, and limited flexibility if your process is unusual.
When building a UI extension is the better call
Build when the off-the-shelf market cannot fit the shape of your business. In practice that means one or more of these is true:
- You need custom data or custom objects surfaced. If the information your team needs lives in a bespoke data model, a custom object, or an internal system no app integrates with, a UI extension can fetch and display exactly that, in context, on the record.
- You need a bespoke workflow or action inside the record. A button that kicks off your specific approval, provisioning or pricing logic, modelled the way your business actually runs, rather than the way a vendor assumed everyone runs.
- No app models your logic. When the process is a competitive differentiator, there is rarely a product that matches it.
- Ownership matters. A custom extension means no per-seat app fees stacking up over time and no dependence on whether a vendor keeps an integration alive. You own the asset.
The honest costs of building: the Enterprise-tier requirement, the need for genuine engineering capability, and ongoing platform upkeep. That upkeep is not theoretical. HubSpot now uses date-based platform versions with defined support lifecycles, and older versions reach end of support and stop building, so a custom extension needs an owner who keeps it current. That is exactly the kind of work we handle on our UI extensions service.
The third option people forget: a Marketplace app of your own
If you need bespoke functionality across multiple HubSpot portals, perhaps because you operate several business units or you are a software vendor whose customers use HubSpot, building your own public Marketplace app can be the right structure. It gives you a distributable, installable product that can itself ship UI extensions. That is a different commercial decision from a one-off internal extension, and we set it out on our Marketplace app development service page.
A simple way to decide
Run your need through three questions in order. Is the data standard, or bespoke to your business? Is the workflow standard, or specific to how you operate? Do you need to own the asset, or are you happy to rent it? If all three answers point to standard and rented, buy from the Marketplace and move on. If any point to bespoke or owned, building is likely the stronger long-term decision, provided you are on Enterprise. The honest answer for many organisations is a mix: buy the commodity integrations, build the handful of things that are genuinely yours.
This is where our "Default to Build" principle applies. Where data and workflow are bespoke, or where ownership protects you from per-seat fees and roadmap risk, building usually wins. Where the need is commodity, the Marketplace is the sensible call, and we will tell you so.
Talk it through with engineers, not configurators
SpotDev is the software engineering firm for HubSpot customers: a UK HubSpot Diamond Partner with an in-house engineering team, HubSpot Custom Integration Accredited and Cyber Essentials Plus certified. We build UI extensions from £8,000 and serverless functions from £2,000, and everything we deliver comes with our guarantee: delivered on time, or you get 20% back. If you are weighing a bespoke extension against an off-the-shelf app and want a straight answer, request a quote and we will help you make the call. For more on what bespoke work on HubSpot looks like, see what you can build with custom HubSpot development and how much a HubSpot developer costs in the UK.
Frequently asked questions
Can a bought Marketplace app appear inside the CRM record like a custom UI extension?
Yes. Public Marketplace apps can now ship their own app cards powered by UI extensions, so a bought app can render interactive functionality directly inside the contact, company or deal record. That means the decision is not really about whether something appears in the CRM. The genuine differences are data model fit, workflow fit and ownership.
Do I need a specific HubSpot subscription tier to build a UI extension?
Yes. Building UI extensions requires an Enterprise-tier subscription. Since 14 Nov 2024 this is available on any Enterprise hub, where previously it was limited to Sales Hub or Service Hub Enterprise. The serverless functions that often sit behind an extension also need Enterprise, or a free developer test account for building and testing. Starter and Pro portals generally need to upgrade before the build route is open.
What is happening to the old HubSpot CRM cards?
Classic CRM cards are being superseded by app cards built on UI extensions. From 16 Jun 2025 no new app can add a classic CRM card, and all existing apps using them must migrate by the final sunset on 31 Oct 2026. HubSpot has released a tool that swaps legacy cards for UI-extension app cards without disrupting the customer view.
When does buying a Marketplace app make more sense than building?
Buy when the need is a common, well-bounded integration that a certified app already covers, when there is no bespoke data model or workflow involved, and when you are happy with vendor-maintained software rather than owning the code. The average HubSpot customer uses more than five integrations, and for commodity needs the Marketplace is usually the faster, cheaper answer.
What does a UI extension cost to build with SpotDev?
SpotDev builds UI extensions from £8,000 and the serverless functions that can back them from £2,000. Every engagement carries our guarantee of delivery on time or 20% back. The right figure for your project depends on the data sources, workflow logic and number of render locations, so request a quote for a scoped estimate.
Stay Updated with Our Latest Insights
Get expert HubSpot tips and integration strategies delivered to your inbox.

