Growth
AI Marketing Analytics for HubSpot
Answers marketing performance questions across your analytics and ad platforms, connected natively to HubSpot so it reads marketing emails, campaigns, forms, contacts…
Growth
AI Growth Experiment Agent
Every growth experiment gets logged, tracked against your own baseline, and closed with a clear keep-or-kill decision.
Growth teams run plenty of experiments, but results get lost in spreadsheets and the same ideas get tested twice. This build logs every experiment with its hypothesis and success metric, tracks it against your own baseline data, and closes each one with a clear keep, kill or iterate call. A regular digest keeps everyone on the same page.
- Every experiment logged with a hypothesis and success metric
- Results checked against your own baseline, not a guess
- A clear keep, kill or iterate call for each test
$2,900
20–30 working days
Optional maintenance $249/mo
What it is
AI Growth Experiment Agent plans, tracks and summarizes the growth experiments your team is already running, checking every result against three months of your own baseline data and closing each one with a clear keep, kill or iterate decision. Results land in a recurring digest your team can act on.
Why businesses buy it
Experiments only pay off when someone makes a decision at the end. Without a shared log, tests run too long, winners never get rolled out and lessons leave with the person who ran them. This build gives your team one clear record of what was tried, what it measured against and what was decided, checked against three months of your own data. Your budget goes into ideas that work. The scope sets out data sources, experiment limits and digest schedule before you order.
Problems this solves
Growth experiments run on and nobody formally closes them
The agent tracks the experiments your team is already running and closes each one with a keep, kill or iterate decision. Tests stop drifting along without an outcome.
Results are judged without comparing to a proper baseline
Every result is checked against three months of your own baseline data. A lucky week is less likely to be mistaken for a real win.
Lessons from past tests are forgotten and repeated
Results land in a recurring digest by email, Slack or dashboard. Your team builds a running record of what was tried and what happened next.
A good fit if
Your team already runs growth experiments but tracks them loosely across spreadsheets, Slack threads and memory.
You want every test judged against your normal variation, not called a win or loss by gut feeling.
You want one recurring digest showing what's running, what finished and what's next, instead of chasing updates.
You have at least three months of baseline data to judge new experiments against.
Not the right fit if
You want the tool itself to find where the funnel is leaking and design the experiments for you. See AI CRO Agent, which does the finding and designing.
You want a full weekly briefing across acquisition, activation and retention, not just an experiment log. See AI Growth Intelligence, which is the broader system.
Your main need is recurring performance reporting, not experiment tracking. See AI Marketing Analytics, which is built for that.
You need more than three connected data sources or more than five live experiments tracked at once. Add a note at checkout for a Change Quote, or use our Custom AI enquiry form.
How it works
- Experiment ideas are logged in one place. Whatever your team is already testing, a pricing change, an onboarding tweak, a new channel, captured with its hypothesis and success metric.
- Each experiment gets a tracking plan. The data source it reads from, the sample size or time window needed, and the metric that decides the outcome.
- Progress is checked against three months of your baseline data. So a result is judged against your normal variation, not treated as significant by chance.
- Results are summarized as they land. In plain language: what changed, by how much, and whether it beat the bar you set.
- A clear call closes every experiment. Keep it, roll it back or run a follow-up test, with the reasoning behind the call.
- A recurring digest keeps the whole team aligned. Delivered to email, Slack or a dashboard, covering what's running, what finished and what's next.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- Read-only access to each data source your experiments should be tracked against.
- Your own definitions of the metrics that decide a win.
- At least three months of baseline data to judge new results against.
- The digest recipients and the channel they should receive it on.
- Admin access and API keys for every connected system, shared by adding our team as a user or creating a key you control, never by email.
- Your business information (company basics, hours and brand voice), shared during intake.
- One decision-maker who can approve keep-or-kill calls.
- An API key for your own AI model account (for example OpenAI or Google Gemini), created for this build with a spending limit.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake verified, current experiments and baseline data agreed.
Weeks 1–2
Data connected and the experiment tracking pipeline built.
Week 3
Digest format and alerts configured.
Weeks 3–4
Tracking checked against three months of your baseline data.
Week 4
Your review and two rounds of changes.
Go-live
Live, recorded handoff session, 30 days of support starts.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $2,900 |
| Express | 15–25 working days | $3,400 |
| Rush | 10–20 working days | $3,900 |
What you receive at handover
- AI Growth Experiment Agent running with read-only access to up to three agreed data sources
- A recurring digest or dashboard summarizing experiments, with up to five alerts
- Keep, kill or iterate decisions for each tracked experiment (recommendations for your team to confirm)
- A test report showing results checked against three months of your historical data
- An admin guide and recorded handoff session
- 30 days of post-launch support
Expected benefits and how to measure them
- Can help your team close more experiments with a clear decision
- Helps avoid acting on false wins
- Makes past learning easy to find
- Experiments closed with a decision
- Count experiments closed each month in your tracking sheet or tool, before and after launch.
- Time from test end to decision
- Log end date and decision date for each experiment and compare averages either side of launch.
- Repeated tests
- Count experiments that repeat an earlier idea without new reasoning, before and after.
Before you start: List current and recent experiments, noting which have a recorded decision and how long decisions took.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build price is one-time. Running it costs roughly $20 to $150 a month in your own accounts, covering AI model usage and the hosting or scheduler behind the tracker.
AI usage during the build and testing is billed the same way, in your own accounts, so every provider relationship stays yours.
Ask about running costsCustom requirements
If you need a fourth data source, more than five experiments tracked at once, or a digest in more than one language, add a note at checkout for a written Change Quote. Larger or bespoke projects go through the Custom AI request form instead.
Request a custom versionKeep it running
After launch, monthly maintenance keeps the build current as your business changes. It covers monitoring, fixes and change hours for things such as a new experiment type or an adjusted baseline. Maintenance starts after the post-launch support period, bills monthly and can be cancelled any month. Unused hours do not roll over.
$249/mo (the minimum fee, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | An experimentation platform charges a monthly fee per test or seat | A growth agency quotes a monthly retainer | Ours is one fixed price, shown on this page before you buy |
| Scope | A platform tracks whatever you set up yourself | An agency agrees scope during the project | Ours is written on the Scope Card, up to five tracked experiments |
| Delivery | A platform installs quickly but needs configuring test by test | An agency usually takes several weeks, set per project | Ours has a fixed delivery time shown before you buy, with faster options |
| Decisions | A platform shows the numbers, you still call the result | An agency build depends on what was agreed | Ours closes every experiment with a stated keep-or-kill call |
Questions about this product
Does AI Growth Experiment Agent decide which experiments to run?
No. It plans, tracks and summarizes the experiments your team already wants to run, giving each one a tracking plan and a clear decision at the end. Choosing what to test next stays your team's call; if you want that leak-finding done for you, see AI CRO Agent instead.
What does AI Growth Experiment Agent cost to run each month?
Roughly $20 to $150 a month in your own accounts, covering AI model usage and the hosting or scheduler behind the tracker. You pay that directly to the provider. Monthly maintenance is optional. We charge nothing extra per experiment tracked.
How fast can AI Growth Experiment Agent go live?
Delivery follows the fixed timeline shown in the delivery panel, counted from a verified intake. Express and rush options are available for a fee. The clock starts once your baseline data and current experiments are confirmed, so gathering them in the first days of intake is the best way to stay on schedule.
How does it decide if a result is significant?
It checks each experiment's outcome against three months of your own baseline variation before launch, so a small week-to-week swing that's normal for your business doesn't get called a win by mistake. That baseline is set during intake and reused for every experiment that follows.
Can it track experiments across more than one channel?
Yes, up to five live experiments at a time, drawing on up to three connected read-only data sources such as GA4, your CRM or an ad platform. Each experiment is tracked independently, so results from different channels never get blended together.
Can I add more experiment slots or data sources later?
Yes. Small adjustments, such as a new metric or a revised baseline, fit inside the monthly change hours included with maintenance. Tracking more than five experiments at once or adding a fourth data source is scoped and quoted separately through a Change Quote.
What if the digest doesn't match how my team talks about results?
You get two rounds of changes built into the build price, using the format and terminology your team already uses. After delivery you have five working days to flag anything that doesn't match the Scope Card, and that gets corrected as part of the original price.
Who owns the experiment history it builds up?
You do. Every logged experiment, its result and the decision reached belong to your business, stored in your own connected accounts. There is no license fee for keeping that history after launch, only optional monthly maintenance if you want ongoing changes made.
Stop losing track of what you already tested.
Order AI Growth Experiment Agent and close every experiment with a clear decision your team can act on. Data sources and limits are agreed in writing before you pay.
Works well with
Growth
AI Marketing Analytics for Google Analytics 4
Answers marketing performance questions across your analytics and ad platforms, connected natively to Google Analytics 4 so it reads sessions, conversions, channel…
Growth
AI Growth Intelligence
Brings acquisition, activation and retention signals into one weekly growth briefing.