UA Automation, Defining & aligning LTV predictions with live performance signals
As part of our Bubbles of Wisdom Series of UA expert interviews, we were delighted to be joined by Tom Nguyen, Founder at Tucmedia. Tom has had an illustrious career within user acquisition, previously working for Goodgame Studios and Zalando.
Over the next week, we will be releasing 10 videos of Tom’s expert insights over 3 posts. You will also be able to listen to the interview in full via podcast here.
The first part covers Tom’s thoughts over the current and future uses of automation within user acquisition and the accuracy of different strategies for attracting valuable users. He also provides insights into his top 3 plots to monitor performance and inform growth. Tom will finally cover how an effective and accurate LTV prediction model can be harnessed to achieve goals within performance marketing and track daily achievements to refine and redefine ad strategy.
Uses of Automation for User Acquisition
The extent to which you can use automation as a marketer for a mobile app publisher depends on the company, monetization, genre, and even specific game. Automation can provide major advantages right now and this will evolve even further in the future, the main areas which can prosper from automation are day-to-day tasks and prediction models involving mass data. However, there will always be large input necessary from humans and your team will need to strategically decide where and how to implement these emerging technologies for beneficial increases in both efficiency and performance.
Top 3 Plots to Monitor UA
This is very individual to each application and dependent on monetization, objectives, and LTV. UA increases in complexity if users are only starting to be ROI positive after longer durations of utilizing the application, where D1 ROAS is negligible the trackable metrics to ascertain the value of newly attracted users changes inherently. However, other applications with much quicker monetization can also be tricky to map when predicting D30 ROAS from D1 or D7, given that other conditions can affect user behavior and these can be unpredictable. Achieving a balanced, accurate, and consistent methodology of utilizing metrics to visualize predicted campaign performance is hugely beneficial.
You have to be able to track metrics early on in your campaign which help indicate correlations & relations within attraction, behavior, and how this equates to your end goal. This will never be an exact science but should be something you can constantly track and refine with data to increase accuracy and foresight when planning and running campaigns.
Designing an LTV curve and implementing the teachings within UA Strategy
IDFA depreciation has added to an already complex issue, within gaming it is slightly easier to reverse engineer how to get to break even and then move to profitability. You can work from the ideal point which you know a user will cross over into becoming ROI positive and examine user actions at various times during their user lifetime to achieve a plot of preferable behavior throughout their journey within the game. Whale-driven games are more difficult to find a median pathway of behaviors to anticipate a valuable user, this involves metrics but a smaller data set and a harder LTV pattern to audit which can throw up many anomalies. With renewing users on a subscription monetization model, you are relying on a metric that can take 365 days to reveal success. This can be incredibly tricky to predict and will rely on a vast amount of user data throughout all their interactions within the app.
Keeping an LTV prediction should be a must within day-to-day performance marketing optimization, however, you should be aware that this LTV curve isn’t 100% accurate and should only be used as a guide and a working document where data is constantly input back into the model to understand current trends.
Be sure to check back next week for the second & third part of Tom’s interview and further insights into user acquisition.
Tom Nguyen is the Founder of Tucmedia, after previously managing UA at Zalando & Goodgame studios. If you would like to know more about Tom & Tucmedia, please visit his UA Wavemaker profile page.