.svg)
Braze Decisioning Studio Go and Pro run on the same decisioning engine, so the choice between them comes down to scope. Go is a self-serve agent that optimises email for clicks, while Pro is a full-service tier for any channel, data source and business metric. This guide gives a three-question test and a worked example so you can settle it before a licence conversation.
Key Takeaways
- Choose by metric first: Go optimises genuine clicks, while Pro can optimise any business metric.
- Go is email only, so push and SMS together, offers or product decisioning point to Pro.
- Go suits recurring email programmes, and Braze advises running an agent for at least 30 days.
- Go needs a dedicated segment, and Braze recommends a minimum audience of 20,000 users.
- Pro needs data readiness, and your team owns the technical implementation.
- Wait until the agent is fully trained before comparing it against your current programme.
What is the difference between Decisioning Studio Go and Pro?
Braze describes Go and Pro as two tiers of the same decisioning engine. Go is the self-serve tier.
You set it up yourself inside Braze, it optimises email for clicks, and it works with the options you configure. Pro is the full-service tier, opened up to any channel, data source and business metric, with support from the Braze AI Decisioning Services team.
Everything you can do in Go, you can also do in Pro. That makes the choice a question of scope, and of whether your use case fits inside Go.
| Capability | Decisioning Studio Go | Decisioning Studio Pro |
|---|---|---|
| Channel | Any channel | |
| Optimises for | Genuine clicks (engagement) | Any business metric, including revenue, conversions and lifetime value |
| Data it learns from | Your audience and its email engagement | Your first-party data as well |
| Offers and product decisioning | Not supported | Supported |
| Constraints and guardrails | Productised frequency, repeat-content limits and quiet hours | Custom constraints and logic |
| Setup | Self-serve in the Braze dashboard | Designed and configured with the AI Decisioning Services team |
Braze Decisioning Studio Go vs Pro: which fits your programme?
We recommend answering three questions in order, and stopping at the first one that points to Pro.
- What will the programme be judged on? If the answer is clicks, Go can carry it. If the answer is revenue or conversions, you are in Pro territory.
- Does the programme need more than email? Push and SMS together, product or offer choice, custom constraints, or first-party data all point to Pro.
- Who will own the work? Go is self-serve inside the Braze dashboard. With Pro, the services team provides design guidance and your team owns the technical implementation.
We put the metric first because it settles everything after it. A programme that your finance team judges on revenue can't be answered with a click metric, however well the agent performs on clicks.
Which programmes suit Decisioning Studio Go?
Braze points Go at a recurring email programme where clicks are a meaningful signal. You supply the subject line, call to action, image and base creative options yourself. The documented fit guidance sorts common programmes into three groups.
| Programme type | Documented fit | What the docs advise |
|---|---|---|
| Always-on calendared campaigns | Best fit | Provide multiple base creatives or variant sets to refresh over time |
| Evergreen programmes (winbacks, re-engagement) | Best fit | Provide variant options that frame the same message for different motivations |
| Multi-email promotions | Supported | Aim for at least 10 campaign days of runway |
| Drip sequences (onboarding tutorials) | Not recommended | Use Canvas for sequencing |
| One-time email blasts | Not recommended | Use Optimize with BrazeAI for A/B testing instead |
The pattern is time. Braze advises you to plan to run an agent for at least 30 days, and ideally keep it always on.
A single send gives the agent no opportunity to improve decisions before the campaign ends. A dedicated two-week promotion agent doesn't finish training before the promotion does.
Two further fit notes matter for journeys. The agent decides once per day, so it isn't suited to emails that must go out within minutes of a trigger.
It also doesn't model sequence requirements. It picks what to send each user based on what's likely to drive a click for that user.
When does a programme need Decisioning Studio Pro?
Braze says to choose Pro when you need a custom metric, first-party data, multiple channels, offer or product decisioning, or custom constraints. The agent in Go does not learn from custom data such as custom attributes or warehouse data. A programme that depends on that data belongs in Pro.
Pro also works differently in Braze. The agent sends the selected message content to Braze in an API payload.
Braze fills your message setup through an API-triggered Canvas or campaign. Engagement outcomes return to Pro through a connected daily data feed, and the loop continues.
The documentation is plain about who does what. The services team provides design guidance. Your team owns the technical implementation, and some teams use a third-party implementation partner for that work.
For design, Braze advises that the more information agents have about your customers, the better they perform. Braze also advises fewer constraints.
CustomerIK's Braze services page covers onboarding, technical integration and AI enablement.
What does Go ask of your team?
Go asks for an audience, a dedicated segment and a creative library. Braze recommends a minimum audience of 20,000 users, with a maximum of 40,000,000. Smaller audiences can run, but they learn more slowly and show less reliable lift.
The segment should belong to the agent alone. If users in it also receive messages from other Canvases or campaigns, those messages affect the engagement the agent observes. The agent also keeps a random control group, with a minimum size of 5%, and those users still receive email.
The creative library is where your team adds value. A handful of distinct subject lines gives the agent more useful choices than a hundred reworded versions. A daily programme also needs more creatives than a weekly one.
Connected Content renders and sends normally, but the agent doesn't see or learn from what it returns. Don't let that be the only difference between two creatives.
What does Pro ask of your team?
Pro asks for data readiness and a set of design decisions. Braze states that high-quality data is the foundation of an effective agent.
Its data preparation section covers assessing your data and connecting warehouse and event sources. It also covers choosing between snapshots and event streams, aligning on Braze external IDs and backfilling historical data.
You make the design decisions with the services team. They cover the success metric, the audience, the experiment groups, the dimensions, the options and the constraints. Braze notes that each of these has implications for how much incremental uplift the agent may generate, and how quickly.
We recommend opening the data conversation before the design conversation. Data readiness is the part of Pro that your team controls, and it shapes which metrics and dimensions are realistic.
Our customer data management service covers identity resolution, CDP integration and data governance.
How do you measure either tier fairly?
Go gives you one comparison by default: the agent against its random control. This group is required, and it is how lift is measured. Comparing against your current programme is something you assemble yourself, because Go does not orchestrate or report on that group.
If you build that comparison, the agent group must contain at least 20,000 users after the split and after allocating its random control. Keep send behaviour equal, and load your current creatives into the agent so both groups choose from the same content.
A frequent reading error is timing. A still-training agent usually trails a mature, already-optimised programme early.
Wait until the agent is fully trained before comparing. Don't judge results, resize the audience or change the control split while it trains.
We suggest writing down the comparison and the read date before launch. A date agreed in advance stops a week-two dip from turning into a decision to switch the agent off.
What does a worked example look like?
This example is hypothetical and illustrative. It isn't client work and it uses no real figures. Take a retailer with three email programmes and walk them through the three questions.
The first programme is a weekly rewards-member email, judged on engagement. Clicks are the agreed signal, email is the only channel, and the team can write several distinct angles. Go fits, and the team can start it self-serve while keeping the segment dedicated and the run always on.
The second is a winback that must choose between a discount offer and a content nudge, judged on repeat revenue. The metric is revenue and the choice involves an offer, so it answers Pro on the first two questions.
The team opens the data conversation with the services team while the first programme learns. That matches Braze's own triage guidance to start with self-serve email personalisation and get expert help later.
The third is a three-email onboarding tutorial that must arrive in order. It fits neither tier, because Braze advises Canvas for sequencing. The team keeps it in Canvas and revisits the question only if it wants optimisation inside the drip.
What mistakes should you avoid when choosing?
Four mistakes are worth avoiding in this decision, and each maps to a documented behaviour.
- Choosing Go and reporting revenue. Go optimises for genuine clicks, so a revenue claim needs a Pro design or a separate measurement plan.
- Running Go on a short promotion. The agent needs roughly 10 campaign days on which it sends, so a two-week burst can end before it learns.
- Reading results while the agent is in training. Performance can look close to random then, and the guidance is to wait for the fully trained stage.
- Sharing the segment. Overlapping Canvases and campaigns add noise the agent can't account for.
How does this connect to the rest of the programme?
Decisioning agents sit alongside the testing and orchestration you already run. Our guide to A/B testing your email campaigns in Braze covers audience-level testing, which finds one winner for everyone.
Go is the individual-level step beyond it. The same creative thinking appears in unlocking hyper-personalization in Braze campaigns.
Evergreen programmes are strong Go candidates. Our posts on building win-back campaigns in Braze and automating loyalty programmes describe programmes you could feed an agent.
When an agent hands email back to a journey, the orchestration in building an automated marketing workflow in Braze still applies. Reading the results belongs with email campaign reporting in Braze.
As a Braze implementation partner, CustomerIK recommends deciding the success metric and the owner of data preparation first. Start Go on the programme that already fits. Open the Pro conversation for the one that doesn't.
Frequently Asked Questions
1. Is Decisioning Studio Go the same engine as Pro?
Braze describes them as two tiers of the same decisioning engine. Go is self-serve and optimises email for clicks, while Pro is full-service and opened up to any channel, data source and business metric.
2. Can Go optimise for revenue?
Go optimises for genuine clicks. Optimising the same programme for conversions or revenue is listed under Pro.
3. How large does the audience need to be for Go?
Braze recommends a minimum audience of 20,000 users, with a maximum of 40,000,000. Smaller audiences can run, but they learn more slowly and show less reliable lift.
4. Can we start with Go and move to Pro later?
Braze's triage guidance lists a path of starting with self-serve email personalisation, then getting expert help later if you outgrow it. That path is Go, then Pro. Check scope and terms with your Braze account team.
5. Does Pro need an implementation partner?
Braze says your team owns the technical implementation and that some teams use a third-party implementation partner for that work. The services team provides design guidance.
Sources
- Braze, BrazeAI Decisioning Studio Go. Retrieved 14 Oct 2026.
- Braze, BrazeAI Decisioning Studio Pro. Retrieved 14 Oct 2026.
- Braze, Get started with Decisioning Studio Pro. Retrieved 14 Oct 2026.
- Braze, Prepare your data. Retrieved 14 Oct 2026.
- Braze, Set up your Decisioning Studio Go agent. Retrieved 14 Oct 2026.
- Braze, Creative setup and management. Retrieved 14 Oct 2026.
- Braze, Measuring Decisioning Studio Go. Retrieved 14 Oct 2026.
- Braze, What to expect after launch. Retrieved 14 Oct 2026.
- Braze, Examples for Decisioning Studio Go. Retrieved 14 Oct 2026.
- Braze, Decisioning Studio Go FAQ. Retrieved 14 Oct 2026.
- Braze, Use Decisioning Studio Go with Canvas. Retrieved 14 Oct 2026.
- Caveat: none of the pages read gives pricing, packaging or a release status for either tier, so this post states none. Confirm both with your Braze account team.
- Written by CustomerIK, a Braze implementation partner.
CustomerIK is a Braze implementation partner working across onboarding, technical integration, marketing operations and customer data management.
If you're deciding between Go and Pro for a programme and want a second view before you speak to Braze, let's talk.
.svg)
