I’ve said this many times, but here we go again: AI has flooded the stores with vibe-coded apps as the barrier to building an app has been drastically lowered.

There are more and more solopreneurs trying to make a living from their apps, but the more I talk to them about User Acquisition, the more I see the same pattern: they try to replicate what they see in success stories from 5 or 10-person teams. The result? Everything gets done halfway.

And the solution isn’t working more hours. It’s designing a smaller but complete system that covers all the basic aspects you need to scale UA until you get to the point where you can afford to bring someone else onto the team.

Warning before starting: AI is needed

Although this point can be unnecessary, I have to say it before I start elaborating the whole article: You will need to rely on AI for multiple parts of the system because the day only has 24 hours and you also have to get some time to eat, sleep and be a human being.

I assume that for most solopreneurs developing apps, AI is already like a third arm for them and they probably use it better than I do but if not, stop reading now and start learning how to use Claude or any other LLM and then come back here to build your UA system.

Besides that, I also want to clarify that I don’t have all the answers. I am just explaining what I have done for small projects and how you can apply what I have learned after running UA strategies for more than 125 apps in the last 7 years. This means that you may agree or disagree with some points or maybe you even do some processes faster or more efficiently already. That is completely fine. I am just elaborating on this system to help all those people who don’t know where to start.

The four pillars you can’t skip in UA even when you’re flying solo

Running UA efficiently is not that complicated when you’re trying to go 0 to 1. Things get more messy as you grow but for this system we’ll focus on having something that works well for apps going 0 to 1.

In this regard, there are four pillars that need to be addressed or controlled if you really want to achieve decent success with your paid campaigns:

  • Channels
  • Creative Production
  • Tracking
  • Schedule

If you manage these four variables more or less efficiently with the tricks and strategies that I’ll explain in the corresponding sections, you won’t need to spend money hiring someone until reaching a decent level of growth. Moreover, you will also learn how UA works along the way, so when the time to hire comes, you will at least understand the foundations and principles of UA. 

With that being said, it’s time to lay out the plan. 

Pillar 1: Channels – Where should you promote your app?

The biggest mistake solopreneurs make is launching on three channels (Meta, TikTok, ASA) at once. As Lucas Moscon explains in his piece on ad channel diversification, launching more channels does not necessarily mean lower risk. Quoting the article as it explains it perfectly in its first paragraph:

“Ad channel diversification can reduce performance when budgets, audiences, or team capacity are too limited to support multiple learning phases, creative pipelines, or attribution workflows.”

I know that when you’re starting you feel the FOMO for not being running campaigns in all the possible channels and figuring out which one brings the best performance for you but that feeling is actually counter-productive. Based on my experience, if your product has PMF and the main metrics are within the benchmarks of the category (discover your benchmarks with RevenueCat’s calculator), the difference that you will observe across channels is simply not determining your success or failure when running UA.

You can get 15-20% better CPAs just by changing the channel but that potential improvement comes with a high cost for a one-person team: new creative formats, new attribution logic, new account management. In the best-case scenario, you get an improvement that is not going to determine your viability to grow but in the worst-case scenario, you will burn out in less than a month for not seeing results in any of the channels after investing an unnecessary amount of time.

So the first learning is simple: Pick one channel and make it work at a decent level of spend. Once you start to struggle with scalability and your creative production system can double the production with no issues, then you can start considering other channels.

Since I’m assuming that you’re solo, I will briefly talk about the main channels as these must be the ones to consider when you want to start UA for the first time:

  • Meta Ads → It fits well for most app categories and gives you 5 different networks at the same time to deliver your ads (Facebook, Instagram, WhatsApp, Threads and Audience Network). This gives more than enough to find your ICP and prepare ads specifically for them. This is my pick in +90% of the subscription apps I work with.
  • TikTok → If your ICP is below 40 years old, TikTok is definitely a channel to consider although it demands a higher velocity with the creative production as the algorithm tends to burn creatives out much faster than any other channel. Sometimes you can get better CPAs here but you will need to rotate twice as many creatives easily compared to Meta.
  • Google → They launched OCM+IDM this year to offer real-time signals and more accurate reporting for app campaigns but it’s still too inconsistent. If you want to go with Google, you must go with web-to-app campaigns. DemandGen and Search campaigns are in this case the first ones to test.
  • Apple Ads → If your category has a clear search intent you can try this channel but right now, Apple Ads is very competitive and if you target the US, you won’t likely win many bid auctions at the price that you need to be profitable.

Except for Apple Ads, all the other channels allow you to optimize your campaigns either for installs, events or value. My recommendation is to focus directly on events if you really want to see how the algorithm looks for potential subscribers for you.

The quality that install campaigns bring is just too low to be sustainable and I only run these campaigns in very specific cases for creative testing but never as a growth pillar in my UA campaigns.

Regarding VO campaigns, these are a next step further as they rely on the value that you send to the ad network and if your subscription value happens on day 3 or day 7 because you offer a free trial, that is too late to get a proper learning so you’d need to work with proxy values and that is definitely a topic that you don’t need to know in the beginning (watch the Signal Engineering webinar with Thomas Petit if you want to learn more about this topic).

Pillar 2: Creative Production – How many creatives do you need? And how should you structure them?

It doesn’t matter how good you are with AI: You will never outproduce a 5-person team. So don’t try it.

I started with that sentence because after creating an X account this year, I see an endless amount of posts every day where solopreneurs brag about the complex systems they have for their paid campaigns. While they make less than $10k MRR.

Let me tell you something: Having successful paid campaigns is not that hard if you apply common sense. You just need to steer the algorithms to deliver your creatives towards the audiences that you consider valuable for your business.

I’ve already written about how I approach creative testing for each stage, but this time I will elaborate a bit on the logic behind the flowcharts rather than the rotation process, which the article already covers in detail.

When you’re starting from scratch, you have an idea of what audiences can work best but you don’t have the data to confirm it. Therefore you must create different creatives that attack all the pain points your potential customers may face and show how your app solves those problems.

Applying the first structure from that article ($0 to $500), your first task is to find which angles resonate best for your target event. Start your campaign with a single ad group and put different creatives that attack different angles and audiences. Check the metrics and then start to pick winners and losers every week while you keep refreshing the campaign according to performance.

This first stage won’t give you deep granularity on the performance per audience or angle of creative but it will help you to confirm the main hypotheses you can test in this first step:

  • What is the average CPI and CPA you can expect from a CPA campaign and whether that aligns with the LTV that you have projected
  • What conversion rates you get from paid campaigns vs organic at each step of the funnel and if these metrics are within benchmarks for your category
  • What angles get the best performance
  • How the algorithm distributes the spend towards the different angles
  • What age groups and placements absorb more spend for each angle

I have also written about creative fatigue and how to detect when it’s time to rotate creatives in your campaigns, so if you wonder what metrics you should look at, go and check that article.

Once you have found the answer to all these questions and while you have scaled to over $500/day, it is time to move to the second structure where you will have much more granularity and space to test more and also push the audiences that work better.

In that second scenario, you will likely be at a point where you need someone else helping you to produce creatives as you will have multiple ad groups that need to be fed with different creatives, every single one at least once a week (but likely twice according to my experience).

Pillar 3: Tracking – What tools do you need to track the performance and how you should track it in the beginning

Let’s be clear: You don’t need a $30k/year MMP setup to start any campaign and see if you can really scale UA. In fact, you don’t need to spend anything on tracking tools for your campaigns as all the main networks (Meta, TikTok and Google) offer free solutions to track the campaigns.

  • Facebook and TikTok have their own SDKs
  • Google has Firebase (which demands another SDK)

So the only tools that you need to measure your campaign performance are:

  • RevenueCat for LTV, Cohorts and subscription data (non-negotiable)
  • The SDK of the ad network that you start with
  • App Store Connect
  • A spreadsheet

In today’s market, you will likely get close to zero pure organic traffic from the store, but if you are running organic videos on Facebook, Instagram and TikTok, you may get some installs and in-app purchases from there. So when you start your paid campaigns, you won’t really know what can be attributed to the campaign by just looking at App Store Connect.

I think this is the most likely scenario for all solopreneurs trying to scale their apps and that is why I will show what I recommend in this stage.

Once you set up the ad network SDK and start running your first paid campaigns, you will start seeing conversions attributed to your campaigns. The first thing is to realize how many real conversions are happening, as the ad networks normally use probabilistic attribution for iOS campaigns and that can result in campaigns under- or over-reporting (in Android, you can fully rely on the number that you see in the ad network).

The simplest way to measure that is by checking the difference in the baseline before and after running the campaign across all the data points that you may have in place and attribute the uplift in the baseline to the paid campaigns.

Obviously, you need a stable period to establish your baseline across your data points (RevenueCat, ASC and your HDYHAU survey in case you have it in the onboarding). Then, you will clearly see how that baseline jumps the moment that you start the paid campaigns, so attributing that difference will tell you how much you are basically paying per trial or direct subscription.

Moreover, by creating different baselines across different data layers (RC, ASC, HDYHAU), you will start to see what the usual discrepancy is between platforms, so you will start to learn what kind of margin you have when you make decisions when looking at your campaign’s numbers.

After getting that CPA, you should go to your RC chart section and pull the LTV data for paying users so you can see if that CPA is low enough to cover:

  • Your LTV in the timeframe that you need (in case it’s a direct subscription)
  • Your trial start and trial conversion rate (in case you optimize for trials).

In parallel to this rudimentary (but reliable) attribution for your only active channel, you must create a blended report that gives you a high-level overview of your business: How much you spend and generate per day across all channels, how many installs you get, how many trials and how many subscriptions based on your average trial conversion rate. You must also create LTVs for different periods of time and then project the revenue on that period of time so you can see where things stand today (ROAS) and what the expected profitability is for the trials that you acquired today across all your channels.

I have a real example of one of these reports that I helped create last month:

As you can see, it is pretty straightforward and gives you a really clear vision of how your business is doing today and how much margin you have in the future based on your investment and conversion rates. This is also great because it quickly tells you if you are suffering any drop-off on the product side or if you are losing efficiency at any step of the funnel.

Pillar 4: Schedule – How often should you execute?

The main advice here is pretty clear: Consistency will always beat intensity, so drill this into your head: Checking campaigns 5 times a day and making reactive changes will hurt your paid performance more than help it.

If you are running a campaign with less than $500/day, you will need to execute changes twice a week at most. The reason is that your budget is simply not high enough to feed your creatives enough to make more changes, and since in the first stage you must simply have one ad group with different angles, the only iterations you must do is to pause the creatives that don’t perform and upload others to test.

In order to give you a more detailed plan rather than rough recommendations, I took the liberty of creating a plan specifically for UA so you can use the rest of the time for other tasks that are also super important for your business (product improvements, A/B testing, organic distribution, finance, etc):

  • Monday (1-2 hours): Refresh your campaign report and blended report and analyze how the week closed compared to the previous one (creative distribution, placements, age, engagement metrics, etc). Decide if any creative needs to be paused and analyze the distribution and metrics of the winning ones so you can think of more ideas related to that winning angle. Sketch out a potential A/B test you would like to test in the paywall and plan a 1-week test for it.
  • Tuesday (3-4 hours): Plan the new batch of 3-5 assets you want to produce this week after you gather all the insights from the reports that you refreshed yesterday, and try to produce them within the same day if possible.
  • Wednesday (1 hour): Check the campaign and see how the creatives you uploaded on Monday are performing. Pause older creatives that didn’t get spend or didn’t perform well and upload 1 or 2 from the batch that you created.
  • Thursday (20 minutes): Refresh the reports and make sure everything goes right. Pause bad-performing assets if and when they have enough impressions (10k+).
  • Friday (1 hour): Refresh the campaign again with the remaining assets that you have from the batch from Tuesday if it’s necessary. If performance is good, save them and simply decide if you want to scale or decrease the budget by looking at the performance of the week.

Besides these tasks, you must spend 15 minutes extra every day. During these 15 minutes, only look at two things:

  • Is the pacing correct?
  • Are there any major anomalies that need to be investigated right away?

You can also automate some of the decisions with automated rules as both Meta and TikTok let you set rules like “pause this ad group if CPA exceeds $X over the last 3 days.” Set them up once and let the platform handle the routine decisions. This is the one place where you let the algorithm work for you instead of against you.

When and how to hire your first external help

If you start scaling, there will be a point of course when you will be overwhelmed by any of these pillars and that is totally fine. You will simply start hitting the ceiling of what one person can manage, and that’s not because the system is broken, but because the opportunity will outgrow the system. That’s when you start thinking about your first hire or freelancer to delegate: a creative producer, a data analyst or a part-time UA manager. These are the main roles that would free up the time you’d otherwise spend on UA.

The best process to see which role suits your company the best is to analyze your own bottlenecks:

  • If creatives are slowing you down → freelance creative on a per-project basis
  • If analytics is overwhelming → data analyst by the hour
  • If campaigns are eating all your time → consider a part-time media buyer

The rule here is to outsource what is repetitive and executable, not what is strategic.

I hope this guide helps you to thrive in this competitive market, whether you’re starting UA from scratch or your current system isn’t working.

I will happily help anyone with further questions on my LinkedIn or X.