Skip to main content
xee.ai/solutions

Whoever runs your app, the agent works for them

One person wearing every hat, or a team that splits the work. Pick who you are to see what Xee does for you, what you decide and what it costs.

Your agent writes to you every Monday

app.xee.ai/apps/candlenest/to-decide
To decide 1 idea

Your first sandbox purchase works. Next I would test a shorter headline on your paywall. Want me to set it up?

See it as

Indie developer

Ship subscriptions in an afternoon, then let the agent grow them

You write the app. Your coding agent adds Xee with the skills, and the dashboard agent suggests what to test next. No backend to build, no percentage of your revenue.

Your first week

  1. Add Xee with your coding agentOne prompt adds the SDK and activates it with your app’s key.
  2. Make a sandbox purchaseBuy with a sandbox account and check that premium turns on.
  3. Approve your first testThe agent suggests it with a reason. You start it.

The agent does

  • Adds the SDK and pushes your paywall from your code
  • Suggests what to test next
  • Checks in every week on what moved

You decide

  • Which tests start and which ship
  • What goes live in your paywall
  • Nothing changes without your yes on Free

Use cases for indie developer

When trials start, by day since install
RevenueCat, 2025
Conversion to paid by day 35, median
Hard paywall12.1%
Freemium2.2%
RevenueCat, 2025
app.xee.ai/apps/trade-ideas/paywalls
The paywall editor in Xee: a phone preview of the paywall next to its testable values
  1. Nearly every trial starts on day one, so where the paywall appears in onboarding matters more than anything later.

    Put the paywall in your onboarding and test where it shows, as a view test. The skill moves its copy and values into Xee so you can change them without a release.

  2. You can’t tell whether to lock the app behind a paywall or let people in for free.

    Test both. Judge the test on revenue, and watch refunds next to it: hard paywalls convert more, and refund more too, 5.8% against 3.4%.

  3. You set your price by guessing what similar apps charge.

    Make the price a value on your paywall and test it. Across the market, higher-priced apps earn a median $55 per payer in the first year, against $8 for low-priced ones.

  4. You’re one person and have no time to read dashboards.

    The agent checks in each week with what moved and the next test it suggests. You say yes or not now.

Product and growth

More tests, fewer meetings about tests

The agent reads your numbers and proposes tests with a reason, a variant B in every language and a goal. You approve, watch the odds and ship the winner without an app release.

Your first week

  1. Make your paywall testablePush its copy and values as variables the agent can change.
  2. Run an A/B testPick a value, write B, choose a goal and read who is winning.
  3. Test onboarding tooScreens that sell nothing can be tested as views.

The agent does

  • Proposes tests, each with a reason
  • Writes variant B in every language
  • Shows the chance B is better

You decide

  • Which tests start, and the goal that judges them
  • When to ship the winner, and when to undo it
  • Whether the agent may run tests itself, on Scale

Use cases for product and growth

Trial-to-paid by trial length, median
17 to 32 days45.7%
3 to 7 days26.8%
RevenueCat, 2025
Subscribers still paying after a year, by plan
Yearly44.1%
Monthly17.0%
Weekly3.4%
RevenueCat, 2025
app.xee.ai/apps/trade-ideas/tests
A/B tests in Xee: two running tests with their variants and the chance each is winning
  1. Your trial is 3 days because everyone else’s is, but longer trials convert better across the market.

    Test a short trial against a longer one with trial-to-paid as the goal, and let the agent tell you the chance the longer one wins.

  2. Yearly subscribers stay far longer, yet your paywall leads with monthly.

    Test showing the yearly plan first, judged on revenue per viewer rather than trial starts, so a cheaper-looking plan can’t win by accident.

  3. People drop out of onboarding before they ever see the paywall.

    Test the onboarding screens as a view, with the steps you send through Xee.track as the goal, and keep the version that gets more people through.

  4. A test looks like a winner, but you’ve been burned by early results before.

    Read the chance B is better before you ship. Shipping changes the value in Xee, so you can undo it if renewals drop later.

Marketing and UA

Spend on the keywords that bring paying users

Apple Ads spend sits next to the revenue each install brought. Send subscription events to your MMP, and let the agent move budget away from what costs more than it earns.

Your first week

  1. Connect Apple AdsLink your account so the agent can read your campaigns.
  2. Get App Store keyword ideasFind keywords with room to climb and fit them into your listing.
  3. Send events to your MMPSubscription events go to AppsFlyer, Adjust, Meta and more.

The agent does

  • Sets each campaign’s cost per install against its return
  • Suggests bid and budget moves
  • Finds keywords that climb

You decide

  • The limits the agent stays inside
  • Each change before it goes live
  • Which storefronts to start with

Use cases for marketing and UA

App Store downloads that come right after a search
Apple, 2022 data
US median cost per tap, by campaign type
Competitor$3.26
Generic$2.75
Brand$2.05
Discovery$1.16
AppTweak, 2025
app.xee.ai/apps/trade-ideas/analytics
Analytics in Xee: revenue over time with the metrics list on the left
  1. Most installs start with a search, and you pay for words you could rank for yourself.

    Use the agent’s keyword ideas to climb in organic search, then buy only the terms where ranking isn’t enough.

  2. Competitor keywords cost almost three times what discovery does, and blended numbers hide it.

    Keep brand, competitor and discovery in separate campaigns, so each is judged on its own return.

  3. Search ads turn more than 60% of taps into installs, so install counts look good whatever the keyword.

    Judge each keyword on what its installs go on to pay, not on how many there are.

  4. Taps are cheap, but you can’t see which installs ever paid.

    Xee sets each campaign’s spend against the subscription revenue its installs brought, and the agent proposes where to move budget. You approve each move.

App owner

Know what earns, without opening five tools

Revenue, installs and trials on one page, set against last month. Ask the chat why something moved, and approve the agent’s changes from your phone.

Your first week

  1. Open your home pageSee revenue, installs and trials for every app, against last month.
  2. Ask why something movedConnect Claude and ask about your numbers in plain words.
  3. Turn on refund saverAnswer Apple’s refund requests with how the app was used.

The agent does

  • Sets this month against last month
  • Explains what moved and why
  • Answers refund requests with real usage

You decide

  • Which changes go live, from your phone
  • Which apps sit in your account
  • When to move to a bigger band

Use cases for app owner

Payers who ask for a refund, by app category
Education4.9%
Health and fitness4.7%
Travel1.5%
RevenueCat, 2025
Subscribers lost at the first renewal
Weekly plans30 to 50%
Monthly plans15 to 40%
RevenueCat, 2025
app.xee.ai/apps/trade-ideas/home
The Xee home page: revenue against last month, with the chat box at the top
  1. Refunds take money you already counted, and Apple decides them without hearing from you.

    Turn on refund saver. It answers Apple’s refund requests with how the app was actually used.

  2. The first renewal is where most monthly subscribers leave, and the revenue chart doesn’t show it.

    Read renewals in cohorts, then ask the agent for a test that moves people toward the plan that lasts.

  3. Tools that take a percentage cost more every month you grow.

    Xee charges a flat price for your revenue band, so the bill only changes when you move to the next band.

  4. Revenue moved this week and nobody on the team can say why.

    Ask the chat. It answers from your own numbers, and from Claude if you connect it.

Or start from where your app is today

Sources

What each role runs into, what the industry data says about it, and what you do about it in Xee.

  1. RevenueCat, State of Subscription Apps 2025. Medians across apps on RevenueCat.
  2. Apple Ads. App Store data, worldwide, 2022. The over-60% install rate is Apple’s figure for November 2024 to October 2025.
  3. AppTweak, Apple Ads benchmarks. About 2,800 US advertisers in 2025.

Numbers in these pictures are examples. The dashboard pictures show Trade Ideas, our demo app.

Ready to grow for free

Every agent comes standard, and it is free until your app makes $5K a month.