The Product

Each of our toolsets answers a fundamental question about your app marketing:

  • AIM - What marketing works?
  • VERIFY - Do people see your ads?
  • PROTECT - Is each acquisition valid?

AIM - Acquisition Incremental Measurement

What marketing works?

AIM is a data-driven attribution product created to give our clients a different view on their app marketing results and provides a way to maximise ROAS.

How? AIM uses machine learning to build a unique client model. It’s intelligent, so it knows which sales have come from past marketing vs today’s efforts. Then it accurately calculates future results.

AIM works with three core factors:

Brand Demand

Sales that would happen without today’s marketing.

Incremental Drivers

Sales that happen because of today’s marketing.


Sales that will happen in the future.

Acquisition Incremental Measurement tools

Measure the conversions driven by each supply partner

Incremental lift means the conversions you get from a supplier that wouldn’t have happened otherwise. We measure this for each supplier, every day, so you can compare different channels. For example, you can compare Facebook and TV ads, like for like.

Our always-on incremental measurement is unique because:

  • It’s always on, so your results are always available and up-to-date without scheduling tests
  • It uses the marketing you’re already doing, so you don’t need control and exposed groups
  • It works on every channel – online and offline - and you’re not restricted to measuring one channel at a time

Discover the supplier’s point of diminishing returns

All your suppliers have a limited capacity – even Google. It’s the point of diminishing returns where each conversion gets much more expensive.

But there’s also a sweet spot where you’re getting the highest return on investment. That’s what network optimum spend helps you find.

Knowing this means you can optimise your spend, and tells you when to stop buying from one supplier and increase your budget with another.

Accurately forecast future results

Our models accurately forecast short, medium and long term marketing results.

They consider the thousands of dynamic variables in your brand profile – and they’re constantly learning to become even more accurate.

Every day you can access updated data to make informed strategic decisions. How much should you spend with each supply partner? When? And what will the results be?



Are ads being seen by real people?

Our Verify tools verify ads from programmatic suppliers.

Whether you pay by click, by impression or by acquisition, they improve performance by avoiding placements which go unseen.

Verify tools

Make sure people are seeing ads

For an ad to be viewable, at least half of it needs to be in view for a second (according to the IAB and MRC). And – of course – it needs to be seen by a real person, not a bot.

If users miss an ad because they scroll down too fast or it doesn’t have time to load, you need to know. This is what our ad-viewability tool checks.
It verifies views and keeps track of publishers’ viewability – which help improve programmatic results.

And if you’re using view-through attribution there’s another big benefit. Attribution platforms don’t measure viewability, so they still attribute unviewable impressions to installs. Our data shows if ads go unseen, so you can dispute these misattributed installs and save budget.

Only bid on quality inventory

This tool helps you buy quality inventory by checking it before you bid.

If you’re running a self-serve DSP, it checks inventory against real-time campaign data before you bid – including ad viewability and our enriched database.

It builds campaign-specific profiles, meaning you can limit bids to the best-performing inventory.



Is each install valid?

It’s always been a tough question to answer – there are plenty of ways to manipulate last touch attribution.

This has meant bad suppliers took credit for conversions they didn’t generate and installs with no value.

So we built our Protect tools. They’re a suite of tools which stop more fraud than any other service – and saves a lot of money for app marketers.

Protect finds fraud in 3 areas:


Prevent fraudulent clicks pre-install


Detect invalid installs to dispute


Identify suspicious user behaviour

Protect tools

Find fake clicks

We find clicks that come from a human, on the correct device, in the right location. If a click is anything else, it’s fraudulent – and we’ll protect our clients from it.

Removing fake clicks will improve any campaign’s performance. It cuts misattribution, protects your organic installs, and lets you concentrate on the honest suppliers.

Protect yourself from fraudulent installs

We start with campaign data and enhance it with our own datasets to show whether an install is genuine or fake.

We do this by looking at what real people do when they install an app. This behaviour shows whether it’s genuine – and reveals an average of six to seven times more fraud than other systems.

Uncover fraudulent app use

Fraud doesn’t stop when an app’s installed. With in-app protect, we look closely at how users interact with your app to find fraud flags.

So, we’ll analyse how people use your app, where they are, and much more to detect fraudsters trying to game the system.


App marketing is fast and ever-changing, there’s a lot for marketers to consider. We appreciate it’s impossible to know it all.

We provide clients with all the support they need from ad-hoc technical requests to fully managed services and everything in-between.

Talk to Machine in confidence about a life without fraudulent installs