Product Areas in Amplitude Opportunities (Early Access)
This feature is in Early Access. During this time, aspects of the functionality may still be developed, and this documentation may not always be up to date. If you have any questions, contact Amplitude Support.
A Product Area gives Amplitude's agents context on where to focus based on your product surfaces, workflows, and goals. It also lets you customize how agents identify, prioritize, and spec new Opportunities for that area.
A Product Area defines both where the system looks and how it decides what matters. Product Areas don't have to be small. They can cover a large surface or workflow, but they should have clear scope, goals, and boundaries. The more specific those boundaries are, the sharper the resulting Opportunities.
Start with clear boundaries
Amplitude recommends starting with two or three well-scoped Product Areas. A Product Area can be large, but it should have explicit boundaries. "Checkout" is easier to tune than "The whole web app" because the included and excluded work is clearer.
What a Product Area contains
Each Product Area captures the context the discovery agents use to find and prioritize opportunities.
| Field | What it does |
|---|---|
| Title and description | A short, memorable name and a plain-language summary of the product context, goals, and current state. |
| Scope | The boundaries of the area: the primary focus, the events that define its edges, and the relevant dashboards and charts. Scope keeps discovery on-topic and excludes near-neighbor areas. |
| Target metrics | The metrics that define success. Each metric has a role: primary (the main success metric), secondary (diagnostic), or guardrail (must not regress). |
| Goals | High-level outcomes the area targets, tagged by category: activation, retention, revenue, conversion, UX, or performance. |
| Success criteria | A plain-language statement of what good looks like, for example "Improve checkout conversion this quarter." |
| Insight sources | Which connected data sources discovery should analyze for this area. For the full list, go to Connect context. |
| Repositories | Connected code repositories used to ground execution plans and let coding agents draft changes. |
| Custom instructions | Durable guardrails and preferences injected into the discovery agents, for example "Prefer low-effort, reversible changes." |
| Competitors and comparables | Optional products or workflows that agents should monitor or use for inspiration when evaluating improvements. |
Custom instructions are durable, not one-off
Use Product Area custom instructions for lasting preferences, such as "Require evidence from at least two sources" or "Prefer reversible changes for mature flows." For temporary investigations, run the Opportunity Manager against a specific scope, page, or signal category. Those one-off runs can explore a focused question without changing the durable Product Area settings.
Create a Product Area
Open Opportunities and create a Product Area in one of three ways.
Option 1: Use the setup agent
Amplitude strongly recommends starting with the setup agent. It's a conversational onboarding partner that scans your project, lists existing Product Areas, reviews product context and event names, then asks which product area you want to focus on. From there it investigates the area, watches session replays when they're available, proposes target metrics and goals grounded in your real data, and presents a complete draft for you to confirm or edit.
The setup agent pre-populates every field from your data, so you review and adjust rather than fill in a blank form. You can always edit and tune the Product Area settings afterward.
Setting up a Product Area with the setup agent
AI
I looked through your project: 3 dashboards and 47 events in the
purchase.*namespace relate to checkout. Which product area or workflow do you want to focus on first?You
Let's focus on checkout. We're seeing drop-off before payment.
- Agent action
Runs a focused scan of checkout dashboards, events, and metrics, then reviews 3 representative session replays.
AI
Here's a draft Product Area for Checkout:
- Scope:
checkout_started→payment_submitted→checkout_completed - Primary metric: checkout completion rate
- Guardrail: checkout latency
- Goals: Conversion, Revenue
Want to adjust the metrics or scope before I create it?
- Scope:
You
That's right. Create it.
- Agent action
Creates the Product Area and triggers the first full discovery cycle.
AI
Created. I've started the first discovery run for Checkout. Initial runs can take up to an hour. Check back and I'll have a prioritized set of opportunities for you.
Option 2: Use guided setup
Guided setup gives you a focused wizard for choosing goals, signals, and tools without using the conversational setup agent. Use it when you already know the Product Area you want to create, but still want structured help choosing metrics, sources, and scope.
Option 3: Configure manually
You can also create a Product Area directly:
- Open Opportunities for the first time, or open the Product Areas view.
- Create a new Product Area and give it a concise name.
- Add a brief description of the product surface and its top strategic goals.
- Define the scope: the focus area and the events that bound it.
- Add target metrics and assign each a role (primary, secondary, or guardrail).
- Add optional competitors or comparables that agents should monitor.
- Select the insight sources for the area.
- Connect any relevant repositories.
- Save, then run Discover to generate the first set of Opportunities.
After setup
Creating a Product Area triggers an initial discovery cycle. The first full run can take up to an hour as the system researches your connected sources. After that, you can refresh discovery any time with Discover, and edit the Product Area as priorities change, updating scope, metrics, goals, sources, or repositories.
Was this helpful?