Most people get Google App Campaigns completely wrong. They treat it like some magic box where you chuck in a few pictures, a bit of text, set a budget, and then pray for a flood of high-quality users. Tbh, that's a surefire way to burn through your cash with very little to show for it. The truth is, while the system is heavily automated, it's not intelligent in the way you think. It's a powerful engine, but you are the one who has to give it the right fuel and a destination. Without your strategic input, it'll just drive in circles, finding you the cheapest, lowest-quality users imaginable.
I've seen inside dozens of app ad accounts, from early-stage startups to established players. The pattern is nearly always the same: a single campaign, a handful of tired old creatives, and an optimisation goal of "installs". This setup is doomed from the start. You're basically telling Google, "find me anyone who will download this thing, I don't care if they open it once and then delete it." The algorithm is ruthlessly efficient at doing exactly what you ask of it. If you ask for cheap installs, you'll get them. But you won't get a business out of it.
So what's the first step I should take?
Before you even think about campaign settings or ad copy, you need to answer one question: what is a good user actually worth to you? Without knowing this, you're flying blind. You can't know if a £2 Cost Per Install (CPI) is good or bad if you don't know what that user will generate in revenue over their lifetime. This is where calculating your Lifetime Value (LTV) comes in. It's the most critical number for any app marketer, and it dictates your entire paid acquisition strategy.
Forget vanity metrics like download numbers. You need to focus on what really moves the needle: revenue and profit. The LTV calculation tells you exactly how much you can afford to spend to acquire a customer while remaining profitable. It turns advertising from a guessing game into a mathematical equation. We've worked on campaigns for apps where we had to deliver tens of thousands of signups, and the only way we could do it profitably was by having a crystal clear understanding of the LTV, allowing us to set a realistic target Cost Per Acquisition (CPA). For a comprehensive guide on this, it's worth understanding how to balance CPI against LTV to ensure you're not just acquiring users, but profitable ones.
How should I structure my campaigns for growth?
Okay, once you know your numbers, you need to give the algorithm a clear path to follow. This means moving away from the "one campaign to rule them all" model. A structured, phased approach is far more effective. You need to guide the machine from broad data gathering to highly specific, value-driven optimisation. I typically structure this in three distinct phases or campaigns.
Phase 1: Volume & Data (Optimise for Installs - tCPI). The goal here is simple: get data into the system, fast. You're teaching the algorithm about the *type* of people who are generally interested in your app. Your target CPI should be realistic. You're not trying to make a profit here, you're buying data. This phase is about volume and learning. Don't expect a positive ROAS.
Phase 2: Quality & Engagement (Optimise for Actions - tCPA). Once you have a decent volume of installs and data on a key in-app event (like 'tutorial_complete', 'registration_complete', or 'first_level_beaten'), you launch a new campaign. This one is optimised for that specific action. You're now telling Google, "Don't just find me installers. Find me installers who are likely to do *this specific thing*." Your CPA will be higher than your CPI, but the quality of user should be dramatically better. This is where you start focusing on acquiring engaged users, which is the foundation for scaling. A lot of founders struggle with this transition, but this is often where you can start seeing a clear path to scaling your user acquisition profitably.
Phase 3: Profit & Scale (Optimise for Value - tROAS). This is the final stage, reserved for apps with in-app purchases or subscription revenue. Once you have enough purchase data flowing through, you can launch a campaign that optimises for a target Return On Ad Spend (tROAS). You are now telling Google the ultimate instruction: "Go find me users who will spend money, and I want to make at least £X back for every £1 I spend." This is the most powerful optimisation setting, but it requires clean, consistent revenue data to work.
Moving through these stages is a methodical process. Rushing to Phase 3 without enough data will just lead to the campaign failing to spend. You have to earn the right to optimise for value by first feeding the system with volume and then quality signals. A key part of this is also ensuring your users stick around, which is why having a strong app retention strategy is just as important as your acquisition campaigns.
Goal: Gather initial user data and teach the algorithm. Focus on install volume, not immediate profit.
Optimise: tCPIGoal: Acquire users who perform key in-app actions. Shift focus from quantity to engagement.
Optimise: tCPAGoal: Acquire high-value users who generate revenue. Optimise directly for profitability and scale.
Optimise: tROASAre my creatives really that important?
Yes. In fact, they are the single most important element of your entire campaign. Let me be blunt: your creatives *are* your targeting. In a world without keyword lists or audience selectors, the images, videos, and text you provide are the primary signals you give Google's algorithm to find your ideal user. If you use generic, corporate-looking stock photos, you'll attract generic, low-intent users. If you show intense, fast-paced gameplay from your mobile game, you'll attract hardcore gamers. The algorithm analyses every asset and matches it to user profiles.
This means you need a *lot* of creative assets. Not 5 images and 2 videos. You should be aiming for the maximum allowed: 20 images, 20 videos, and 5 text assets for both headlines and descriptions. And you can't just set and forget them. You need a system for constantly testing and refreshing your creatives. I remember one client, a medical job matching SaaS, who came to us with a £100 CPA. The account was stagnant with old creatives. By implementing a rigorous creative testing framework and shifting their optimisation strategy, we brought their CPA down to just £7. That's not a typo. The change was almost entirely driven by giving the algorithm better, more relevant fuel to work with. It's also why having a robust creative strategy for your app ads is not optional, it's fundamental.
- Video: This is your most powerful asset. You need variety. A 15-second portrait video for Shorts and App placements. A 30-second landscape video for YouTube. A simple square video showing UI for Display. Test UGC-style testimonials, animated explainers, and direct screen recordings of the app in action. Show, don't just tell.
- Images: Go beyond simple screenshots. Show your app's UI in a real-world context on a phone screen. Use lifestyle images that evoke the feeling or benefit of using your app. Test images with bold text overlays calling out a key value proposition. Make them thumb-stopping.
- Text: Your headlines and descriptions need to work together. Use the Problem-Agitate-Solve formula. Headline: "Struggling to track your finances?" Description: "Bills piling up and you have no idea where your money is going. Get a clear view of your spending in minutes and start saving." Be direct and focus on the user's pain point.
Your job is to provide a diverse portfolio of assets so the algorithm can mix and match to find the perfect combination for different placements and user segments. A lazy creative strategy is the number one reason app campaigns fail. A wierd thing is how many people overlook this.
What about tracking? What events actually matter?
"Installs" is not an event you should care about beyond the initial data-gathering phase. It's a vanity metric. You need to track what happens *after* the install. Your tracking setup should mirror your user journey, creating a funnel of micro-conversions that lead to your ultimate goal. This gives the algorithm valuable signals to optimise towards.
Think about the steps a perfect user would take in your app:
- Install: The starting point.
- Onboarding Complete: They've gone through your initial setup. This is a great early signal of an engaged user.
- Registration Complete: They've committed by creating an account. This is a much stronger signal than just an install.
- First Key Action: This is specific to your app. It could be 'level_5_reached' for a game, 'first_song_played' for a music app, or 'first_item_added_to_cart' for an ecommerce app. This is often the best event to optimise for in Phase 2 (tCPA).
- Purchase / Subscription: The ultimate goal. This is what you optimise for in Phase 3 (tROAS).
By setting up and passing these events back to Google Ads, you're creating a roadmap for the algorithm. It learns that users who complete registration are more valuable than users who just install, and users who make a purchase are the most valuable of all. Without this roadmap, the algorithm is stuck at the starting line, only able to optimise for the least valuable action. This detailed tracking is also vital when you start exploring other platforms, as the principles of what makes for a good user are often universal, even if the mechanics of something like getting app downloads from Google Ads is different from other channels. And sometimes you have to compare, for example, between Apple Search Ads and Google App Ads to see what works best for your specific audience.
How do I set my bids and budgets?
This is where a lot of people get nervous and either bid too low, starving the campaign, or set the budget too high and get scared by the initial spend. You need to be methodical. The first 5-7 days of any new campaign is the "learning phase". The algorithm is exploring and gathering data. During this time, you MUST NOT make any significant changes to bids, budgets, or creatives. Let it learn. Any changes will reset the learning phase and you'll be stuck in a permanent state of inefficient learning.
Setting Initial Bids: Don't just invent a number. Look at industry benchmarks if you have them, but it's better to be guided by your LTV calculation. If you know a user who completes registration is worth £15, you can confidently set a target CPA of, say, £5-£7 for that action. For an initial tCPI campaign, start with a bid that feels reasonable for your industry, perhaps £1-£2 in a developed market, and be prepared to adjust it up if you're not getting impressions.
Setting Budgets: Your daily budget should ideally be at least 10x your target CPA. So if your tCPA is £7, you should aim for a budget of at least £70/day. This gives the algorithm enough room to get the 50 conversions it typically needs to exit the learning phase within a week. If that budget is too high, you can start lower, but be aware that the learning phase will take longer.
Scaling: Once a campaign is performing well and has exited the learning phase, you can start to scale the budget. The rule here is slow and steady. Increase the budget by no more than 20% every 2-3 days. A sudden large increase can shock the system and push it back into the learning phase, often wrecking performance. Gradual increases allow the algorithm to adapt and find more users at a stable cost. We've used this exact method to scale app campaigns to thousands in daily spend while maintaining a target CPA. It requires patience, which many founders lack. For those in the UK market, there are some specific nuances to consider which are covered in our growth framework for Google App Ads in the UK.
This is the main advice I have for you:
To pull all of this together, here’s a summary table of the best practices we’ve discussed. Think of this as your checklist for setting up and running Google App Campaigns that actually work. Following this structure is what seperates the amateurs from the pros.
| Step | Best Practice | Why It Matters | Expert Tip |
|---|---|---|---|
| 1. Strategy | Calculate User LTV first. | Dictates your entire bidding strategy and defines what a "good" CPA is. Without it, you're guessing. | Your LTV to CAC (Customer Acquisition Cost) ratio should be at least 3:1 for a healthy business model. |
| 2. Tracking | Implement a full funnel of in-app conversion events. | Gives the algorithm high-quality signals to optimise towards, moving beyond cheap installs. | Pick a "First Key Action" that strongly correlates with a user becoming a paying customer. Optimise for this in your Phase 2 campaign. |
| 3. Structure | Use a phased campaign structure (tCPI -> tCPA -> tROAS). | Methodically guides the algorithm from learning to optimising for quality, and finally for profit. | Don't move to the next phase until the previous one has enough data (e.g., 1000+ installs for Phase 1, 100+ key actions for Phase 2). |
| 4. Creatives | Upload the maximum number of diverse assets (20 videos, 20 images, 5x text). | Creatives are your targeting. More assets give the algorithm more options to find your ideal user. | Refresh your creative assets at least every 4-6 weeks to combat ad fatigue. Always be testing new angles. |
| 5. Optimisation | Respect the learning phase. Scale budgets slowly (max 20% every few days). | Avoids shocking the system, which can reset learning and destroy campaign performance. | If a campaign's performance tanks, check your app store rating and reviews. A sudden drop can definately impact ad performance. |
Why might I need expert help with this?
As you can probably tell, running Google App Campaigns effectively is not a simple task. It’s not about finding a secret "hack" or a magic button. It's a rigorous, data-driven process that requires a deep understanding of marketing strategy, a constant pipeline of high-quality creative production, and the discipline to analyse performance and make methodical changes.
Many app founders are experts in building a product, but not necessarily in the complex, ever-changing world of paid user acquisition. You might not have the time to produce 20 different video ads, the expertise to correctly set up a full-funnel event tracking system, or the experience to interpret the data and know when to scale a campaign and when to kill it. That's where specialist experience becomes invaluable. An expert can help you avoid the costly mistakes many make in the beginning, build a scalable acquisition engine from day one, and ultimately help you grow your app faster and more profitably.
If you're looking to scale your app and want a second pair of eyes on your strategy, we offer a free, no-obligation consultation. We can review your current campaigns, discuss your goals, and give you some actionable advice on how to improve your results.
