Behind every great UA team is a great UA Analyst

A sneak peek at something we've been building: Angus, an AI analyst inside #Ktrl, because behind every great UA team is a great UA analyst.

Our mission at Kohort is a world with zero wasted UA dollars, which means putting predictive LTV in the hands of every UA team. Angus is the next, big, piece of that.

We've trained Angus on the knowledge built from predicting hundreds of games, billions in UA spend, and the real human feedback from both our team and UA teams across the globe.

Think a UA analyst who knows your data and predictions inside out, armed with more data signals than any one studio will ever see.

Some examples of what our clients have been using Angus for:

- "Which campaigns should I scale this week?" → a ranked list by confidence and spend - "How much further could we scale our spend?" → an objective analysis of LTV trends across all networks to validate scaling up - A board-ready UA summary, June vs May, with the risks and opportunities pulled out - Clear insight of an LTV dip that turned out to be a product retention issue, not a UA one

Bruno, our Senior UA Analyst at Kohort, shares a few examples.

---

Hi everyone. Our mission at Kohort is to build a world with zero wasted UA dollars. And to do that, we're making predictive LTV algorithms available for all UA teams. No matter the size of your team, no matter your UA budgets, or your access to data scientists.

Now, machine learning and AI have become an essential part of your day-to-day in your organizations. And now we're making sure that you can use our LTV predictions in whatever way you see fit. The first step of that was to create a set of APIs that allows you to export our predictions and our recommendations and plug them into your internal BI tools. Some Control users expressed the desire to use these predictions and these recommendations and plug them into AI language models to then query the results of our PLTV and allow them to make more informed decisions on UA.

So, now we've gone one step further and we're allowing you to run queries directly from Control through our new AI agent, Angus. Powered by Claude, Angus is like a new analyst that knows the Control tool like the back of their hand. And with a few simple clicks and some natural language, you can request customized reports or analytics deep dives based on the data available in Control.

Here are some examples. For UA managers, we've asked Angus which campaigns are good candidates to scale this week. And so Angus has delivered a list of campaigns prioritized by confidence level and by spend that have room for scale based on recent performance. You can then save even more time when looking at our UA recommendations.

For CMOs preparing for a board meeting, we've asked Angus to prepare a monthly summary of UA performance in June versus May so that they can share that in board meetings with callouts to current campaigns, risks, and opportunities. And this is repeatable so that you don't need to go through this process every month.

In one of these instances, we realized that there was a decline in LTV performance. So we asked Angus to look at the trends of retention and monetization between paid and organic channels. This revealed that there was a decline in both, which pointed towards an issue with the product performance to be investigated with the product team.

Or, if you have an upcoming strategy meeting with your finance team, you can also ask Angus to explore different views of ROAS between paid and incremental blended, or between gross and net. This helps to identify if there's opportunity for further scale if you change your ROAS targets, potentially even leading to an extension of your payback period.

As AI becomes a smarter, stronger ally to UA managers everywhere, we want to make sure that you can still use Control as your source of truth and the starting point to getting all of these answers. So that you can now rely on Angus as your new junior UA analyst to do all of this repetitive reporting work that stops you from doing more important things and really strong campaign execution.