The Tenjin team had a great time attending Mobile Growth Summit 2018.
It was a fantastic opportunity to connect with the growth marketing community as a whole and exchange frontline knowledge with some of the sharpest folks in the industry. I even had the pleasure of sharing a few insights gleaned from our own experience as the industry’s only comprehensive growth data infrastructure. We ran a hands-on workshop that explored the ideal application of granular marketing data when optimizing for LTV across mobile app portfolios. We had a fantastic turnout that sparked some compelling discussions, but I found the most compelling take away to be this:
If you want to be a successful digital marketer, basic coding skills are no longer optional.
Back in 2008, mobile marketers relied heavily on what’s called “summary data”. These were multiple dashboards with 10's of rows that didn’t much resemble what we consider to be actionable data by today’s standards. The infrastructure simply wasn’t there. During this time, app developers spent millions buying installs based on little more than CPI and app store chart ranks. It was a time of indelicate solutions, plagued by costly business decisions that were often based on intuition rather than fact.
Thankfully, as the market matured, data became more accessible. The level of granularity increased and while technology still hadn’t yet been built up around it, firms were starting to make better use of the data they had. Up until about 2015, most mobile marketers relied on what’s called "processed pata", which included 1000s of rows and made heavy use of Excel-based manipulation and analysis techniques. This made decision making for campaigns actionable, but also tedious and error prone.
Here’s an example of an Excel spreadsheet that shows how to use processed data to build a predictive LTV model for a single segment of customers that come from a specific ad network.
The issue that marketers run into when using Excel is the sheer volume of data that gets introduced as the number of customer segments increases. When marketers only ran a few campaigns on a few ad networks, the number of Excel calculations remained proportionate to the campaigns that were running. However, as technology allowed marketers to get more specific with their campaigns, things started looking more like this:
Campaign count = Ad Networks (10) x Countries (100) x Sites/Keywords (1000) x Demo targets (100) x ...
This would require nearly 100M computations every day with new data. In other words, if you wanted to get the most accurate predictive LTV calculations using the kind of Excel sheet linked above, you'd need to do 100M calculations every day to get every permutation of all user campaigns! Not the most feasible of options.
Finally, as recently as 2017, we’ve started to see a shift. The industry’s most successful marketers are writing basic SQL/Python/R to automate these calculations. This is affording them the accuracy and efficiency they need to properly understand the users they’re acquiring. It’s allowing for infinitely greater insight into the strategies that are most effective not only for acquiring users, but for retaining and monetizing them as well.
In 2018 and beyond, we predict that marketers will start learning the technical skills required to automate their data processes and reach actionable insights faster. As they do, they’re going to need tools that enable and optimize these new kinds of programmatic workflows. We’ve built the Tenjin growth data infrastructure with tomorrow’s marketer in mind. Our goal has always been to empower technically minded marketers to achieve their goals in order to usher in a new standard of efficiency and profitability across their mobile portfolios.
The age of the non-technical marketer is over, and we couldn’t be happier. If you’d like to learn more about how Tenjin can help, we’d love to show you! Request a demo today.