Facebook and Instagram ads

Facebook app install ads

App install ads on Facebook and Instagram for UK businesses with a live iPhone or Android app. I set up the tracking, build the campaigns in your own ad account and optimise them towards sign-ups, trials and purchases, so you pay for users rather than downloads that are never opened.

Facebook and Instagram app install ads put your iPhone or Android app in front of people who are likely to download it and, more to the point, use it. I plan and run them for UK businesses with a live app, using Meta’s app promotion objective, and I judge them on what happens after the download: an account created, a trial started, a first order placed, a second visit a week later. An install is where the measurement starts, not where it ends.

What this fixes, and what changes afterwards

App install campaigns usually go wrong in one of three ways. The campaigns are set to buy installs, so Meta finds the people most likely to tap “Get” and least likely to open the app again. The numbers in Ads Manager, the measurement tool and App Store Connect disagree, and nobody can say which figure to believe. Or the iOS results look so thin that the owner suspects the ads are not working on iPhones at all.

Once the work is done, campaigns optimise towards the in-app action that pays your bills, the tracking behind them is tested, and the gap between Meta’s reporting and your own back-end figures is explained rather than argued about. You still will not get a perfect count on iOS; nobody can since Apple’s privacy changes. What you get is a set of numbers you understand well enough to make budget decisions with.

Who it suits, and who it does not

  • Subscription apps (fitness, language learning, wellbeing, productivity) where a free trial or paid plan is the action worth paying for.
  • Retail, food ordering and booking apps where the app is a main sales channel and first orders can be tracked.
  • Money and finance apps, with the extra rules that brings; my page on Facebook and Instagram ads for fintech covers the FCA side.
  • Games and entertainment apps with in-app purchases or a clear engagement event to optimise for.

It is the wrong moment if the app is still in testing, crashes on common devices or has a thin store listing with few ratings; paid traffic will expose those problems at your expense. It is also hard to make work on a very small budget, because Meta’s delivery system needs a steady flow of the event you ask it to find. If you are not yet producing that flow, I would start by optimising for a cheaper, earlier event, or look first at Google app install campaigns, which reach people already searching the app stores.

What the work covers

App registration and event set-up

The app needs registering with Meta and connecting to your business portfolio and ad account, and it needs to report installs and in-app events, either through Meta’s own SDK or through a mobile measurement partner such as AppsFlyer, Adjust, Branch or Singular. I write the event specification: which actions matter, what each is called, which values it carries and how each maps to one of Meta’s standard events. Your developer adds the code; I test that every event arrives once, with the right value, from both platforms.

iOS measurement after App Tracking Transparency

Since iOS 14.5, an app has to ask permission through Apple’s App Tracking Transparency prompt before it can track someone across other companies’ apps, and many people say no. For those users, Meta gets aggregated, delayed and partly anonymised results through Apple’s SKAdNetwork framework and its successor, AdAttributionKit, combined with Meta’s own Aggregated Event Measurement. In practice this means iOS conversions arrive late, cannot be tied to a person and are partly modelled.

The detail you get back depends on a conversion value scheme that someone has to design: a short list of values, each standing for an action or a revenue band, usually set up in your measurement partner. Left on its defaults, it often reports little more than “installed”. I design it around the actions you care about, and I explain on paper which iOS figures are measured and which are estimates. On Android there is no equivalent of the tracking prompt, so reporting is usually fuller, which is one reason I often run and judge the two platforms separately.

Campaign structure and the optimisation event

Meta’s app promotion objective can optimise for installs, for a chosen in-app event or for purchase value. The choice of optimisation event matters more than any targeting setting. I pick the deepest action the account can produce often enough each week for Meta’s delivery system to get through its learning phase; an event that fires only a handful of times a week leaves the campaign guessing. Depending on volume, that might be a manual campaign or one of Meta’s automated Advantage+ campaigns, which hand more of the targeting and placement decisions to the algorithm.

Creative made for app ads

App ads work best when they show the app doing its job: a screen recording of the first useful minute, a person using it at the moment the problem occurs, a before-and-after the viewer recognises. Much of the delivery on app campaigns goes to vertical placements, so I brief and script for Reels, Stories and in-feed video first, with several distinct ideas rather than one idea in five colours. New concepts go in on a schedule, because creative fatigue sets in quickly when the same few videos carry most of the spend.

The store listing and UK rules

Every ad sends people to an App Store or Google Play page, and if that page does not match the promise in the ad, the install rate drops and you pay for it. I check the first screenshots, the opening lines and the ratings against what the ads say. On the legal side, PECR covers SDKs that store or read data on a phone just as it covers cookies on a website, so the Meta SDK and any measurement partner should respect the user’s consent choice in your app, separately from Apple’s prompt. App ads also fall under the CAP Code, so an app that is free to download but useful only with a subscription must not be advertised as simply free.

How I run it

  1. Measurement check. I install a test build or use test devices, trigger each event and confirm what Meta and your measurement partner record. Nothing launches until this is clean.
  2. Target cost. We work back from what a paying user is worth to you, which the LTV and CAC calculator helps with, to an affordable cost per trial, sign-up or first order.
  3. Build. Campaigns, audiences and creative go live in your own ad account, usually split by platform so iOS and Android are each judged on their own data.
  4. Learning and testing. I leave campaigns alone through the learning phase, then test creative concepts and the optimisation event in turn, one change at a time.
  5. Review against your figures. Each month I compare Meta’s reporting with your back end: installs against registrations, trials against paid conversions, and how many users are still active after a week and a month.

Problems I see most often in UK app accounts

  • Campaigns still optimising for installs a year after launch, with cost per install falling and paying users falling with it.
  • Events counted twice, because the Meta SDK and a measurement partner both send the same purchase to Meta.
  • An iOS conversion value scheme left on defaults, so iOS campaigns are judged on installs alone.
  • One creative concept running for months, with spend sliding onto the same few tired videos.
  • The SDK starting before the user has made a consent choice in the app.
  • Results judged only on the measurement partner’s last-touch report, with no test of how many installs would have happened anyway.

What you receive

  • A written event and conversion value specification your developer can work from.
  • A tested campaign build in an ad account you own and keep.
  • Creative briefs and scripts for each concept, with the placements each is made for.
  • A short weekly note and a monthly report set against your own registration, trial and revenue figures, with what I plan to change next.

Next step

Send me your store link, tell me which in-app action matters most to you, and say what you spend now. On a free 30-minute call I will tell you whether your tracking is ready for Meta app campaigns and what I would set up first. App campaigns can be run as a one-off build or as part of monthly Facebook ads management; I quote once I know the scope. To arrange the call, get in touch through my contact page.

Frequently asked questions

How much should I spend on Facebook app install ads?

I would not give a figure before knowing what a paying user is worth to you and what each sign-up or trial currently costs. The useful test is whether the budget lets the campaign record a steady weekly flow of the event it is optimising for. If it cannot, optimise for an earlier event or put the money into fewer campaigns rather than spreading it thinly.

Is Meta's SDK enough, or do I need a mobile measurement partner?

If Meta is your only paid channel, the Meta SDK is often enough. Once you also run Google, TikTok or Apple Search Ads, a measurement partner such as AppsFlyer or Adjust earns its fee, because it counts installs across every channel by one set of rules and helps manage iOS conversion values. Either way, only one route should send each event to Meta.

Why does Ads Manager show fewer iOS installs than App Store Connect?

App Store Connect counts every download, including people who found the app on their own. Meta only claims installs it can attribute to an ad, and on iOS most of those come back in aggregated, delayed form because of Apple's tracking rules. Some gap is normal; a gap that suddenly widens usually points to a tracking fault worth investigating.

Should I run Meta and Google app campaigns at the same time?

Often, yes, because they reach people at different moments. Google app campaigns catch people searching the app stores and Google, while Meta creates demand among people who were not looking. Run them side by side with one measurement partner deciding attribution, so the same install is not claimed twice. My page on Google app install campaigns explains that side.

Can you make the ad videos?

I write the concepts, scripts and shot lists, and review cuts before they go live. Filming and editing usually sit with your designer, an in-house team or a creator, and app screen recordings can often be captured by whoever builds the app. Good creative for app ads is rarely expensive; it is clear, quick and shows the app working.

Can I use app install ads to promote a web app?

No. Meta's app promotion objective needs an app listed on the App Store or Google Play. A web app or progressive web app is promoted with a sales or leads campaign that sends people to your site, measured with the Meta Pixel and Conversions API rather than an SDK.

Ready to talk about your project?

A straight answer about what would move the numbers, and a written proposal if we are a fit.