Polaris Turns One!

Exactly one year ago, we publicly launched Polaris on VentureBeat (thanks again, Dean Takahashi). At the time, it was the first of its kind: a measurement product pairing media mix modeling (MMM) and geo lift experiments to provide incrementality performance metrics in the same form factor as an MMP. Actually, it remains the only product of its kind, one year later.

While we didn’t fully open up the floodgates until Q3 2021, the past year has been a wild ride. The response from the mobile industry was far greater than I could’ve imagined. Our partnership with Meta and collaboration with their world-class Marketing Science and MMM teams have helped us to continually improve Polaris and the science behind it.

Our growth over the last 2 quarters since we went fully live is nothing short of incredible. We’re adding new people to our amazing team and expanding to new regions across the globe. We’ve onboarded so many notable customers over these months it’s been an absolute blur. None of this would have been possible without my teammates at MetricWorks. I’m humbled to be a part of this journey with all of you.

To learn more about Polaris, submit a demo request or visit our Polaris product page.

Video: How To Apply MMM To Mobile Apps

While there seems to be considerable confusion out there about MMM (Media Mix Modelling), it is actually very relevant and applicable to mobile app marketing.  To hear more about how to apply MMM to mobile apps, and why MMM can be more easily applied to mobile apps than to CPG companies watch this conversation between Shamanth M. Rao and Brian Krebs (CEO of MetricWorks). Full video recording available below.

Thanks again to Shamanth Rao (CEO of Rocketship HQ) for hosting MetricWorks’ CEO, Brian Krebs on The Mobile User Acquisition Show.  This is an informative episode that explains why Last Touch/SKAdNetwork is NOT enough for successful post-IDFA measurement, how MetricWorks has seen first-hand the success of media mix modelling for mobile apps, especially at levels of scale far smaller than that of most CPG brands, in which space these models have traditionally been used.  In this episode, Brian describes some of the ways in which these models are specifically applicable to mobile marketing campaigns and argues that these measurement tools are not only useful post-IDFA, but also provide distinct advantages over other measurement tools used by mobile marketers. 

MetricWorks’ own measurement solution, Polaris also makes it easy for you to enjoy these distinct advantages.   It requires really low effort on your part and we can get you started in 24 hours. No need for device IDs, migrations, extra SDKs, or heavy implementation lift. Keep your existing KPIs and marketing processes. Polaris can even do a side-by-side comparison of your incrementality data versus your last touch data. It integrates directly with your last touch MMP to pull the necessary data into a single repository along with your incrementality data. Email us today for a demo

Measurement granularity is a major gap within SKAdNetwork and we’re glad to see The Mobile User Acquisition Show draw attention to this major issue. Our incrementality MMP, Polaris which can co-exist with SKAdNetwork, offers a comprehensive, granular and accurate alternative that addresses the following glaring issues with SKAdNetwork:

  1. SKAdNetwork leaves marketers with only 100 campaign IDs of granularity
  2. It does NOT offer a way to cohort KPIs by install date
  3. Critical post-install performance KPIs that are foundational to your mobile growth strategy today like retention, revenue, LTV, and ROAS, have disappeared and been replaced by a single conversion value with its own significant limitations.

Does your measurement plan address the above gaps?  Contact us today to discuss how our measurement solution, Polaris can boost the effectiveness of your marketing and help you stay ahead of your competition.

You can also listen to the full podcast on iTunes or view the episode notes at The Mobile User Acquisition Show.  Please submit your questions below.  We look forward to answering them and making the future of your measurement brighter!

Guest Blog: Thoughts on the shift in measurement from a practising UA professional (By Claire Rozain)

This past year was challenging on so many levels and filled with constant doubt that each of us felt at both a personal and professional level. Apple certainly didn’t make things easy on us. Indeed, this was the year that Apple decided to challenge all of us in the mobile advertising space and by leading a significant change in the attribution landscape with the release of SKAdNetwork attribution.

Still today, I do not have all the answers on how this change is going to be standardized in the future and the true impact it will have on developers from the smallest to the largest. However, to me it really is an eye opener and appears as a unique opportunity to track and attribute advertising data via incrementality to arrive at a holistic view. In this article, I share my own experience with iOS14 and how excited I am to shift from last touch to incrementality.

SKAN’s challenges including dealing with only 100 campaign IDs

When Apple announced the 100 campaign limitation, it made me anxious. It meant a limitation in deterministic data but also in the number of campaigns a marketer is able to run. So what was I going to do? No more experimentation? My first thought was “what is my job going to look like and how will I still enjoy doing user acquisition? 

To me, the restriction in the number of campaigns was really perceived long term as a threat to my UA manager position. Indeed, fewer campaigns meant less fun. It also meant fewer ways to experiment and challenge my thinking, fewer ways to find new pockets of growth, fewer ways to extend my media mix, fewer ways to easily optimise on the next crazy event designed by data science, and fewer ways to get direct feedback from the market without a timer. 

This left me with the strange and disconcerting feeling that the advertising platform did not need UA managers anymore in order to set up a AAA campaign without targeting or anything similar as everything was now heavily simplified and restricted.
 

SKAN does NOT offer a way to cohort KPIs by install date

However, what I got from this situation was that SKAN attribution was the new rule and I was interested in the new areas that Apple wanted to innovate in. I was really keen to have that information, but for a year, it was really difficult to get a clear and comprehensive explanation from Apple. Apple made several announcements during the WWDC and each time I remember that my UA manager friends and I  were really nervous about these because we did not know how things would pan out. After a year, we finally saw the first shape of SKAN and ran the first test on it. 

For now, the SKAN solution has  not met all of my expectations. Apple did a lot of back and forth that led in a year to the delivery of a disappointing SKAN attribution model, not at the level of the privacy promised but also disappointing from a market overview. A lot of simple things like offering daily KPIS by install date is currently really complicated with SKAN. All data is aggregated, the timer is super long and the sample size in order to receive installs postback data is still unclear to me. 

Funny enough, it made me happy that it was not perfect because it created opportunities for other smart companies to fill these gaps and they certainly did! I was happy that this complicated attribution was the real opportunity I was waiting for to push mobile companies to consider holistic marketing via incrementality and finally empower creativity in advertising!

 

Critical KPIs have been replaced by a single conversion value with its own significant limitation

Because advertisers were unhappy about the iOS14 change and all the limitations that it imposed by following only a single conversion value, I saw advertisers start to be resilient, to evolve and to innovate. People finally stepped outside of their comfort zones and challenged the status quo. The forgotten topics of creative and incrementality became the new topics everyone is now excited about to refashion their growth! 

I saw incredible growth masterminds and leaders that I deeply respect lead conferences about this topic, share resources to the entire market and open the doors to understanding incrementality. All of these people truly inspired me as a young UA manager. I could only imagine the struggle they endured everyday to explain to their high level management the change that they were implementing. At the same time, they also inspired other companies to help lead the change.

SKAN implications for budgeting, forecasting and strategy

SKAN budgeting is complicated at the company level and I think all leaders struggled and are maybe still struggling on how to figure it out. Investors are still unsure about the negative or positive externalities of spending heavily on iOS right now. Does it send a good signal to investors to spend heavily on iOS right now when we belong to SKAN? Or is it better to spend on Android? 

At the company level, Apple was NOT very vocal about the changes they were making and this created a lot of uncertainty. Almost every week, Apple was updating yet another release note with new information.  This made budgeting highly complicated for marketers as the day to day approach was to wait for an update prior to doing any forecast. Growth is not just a matter of days and months, it is a matter of years that involves you working with many people to be successful. Each decision can have a wide-ranging impact.
 

Last touch is the source of the above challenges

Last touch attribution on Apple was not an option anymore unless you had an ATT prompt where the user clicked YES. Since I worked in CRM before, I knew the optin rate would differ from one app to another and the deterministic data we get would certainly not represent 100% of the cohort. However, I was happy about it because last touch attribution has always been something I was really a fan of.

Indeed, to me it was always a way to favor the SANs, a way to say brandformance does not exist as it did when I worked in the past with an offline user acquisition channel. I also knew the last view or click was imperfect or non-ideal data as it also attributed to paid traffic what might have been organic, and was sometimes even fraudulent. So overall, I was really happy and excited about all the sparks that came out about this topic and the innovation in incrementality and holistic marketing views in organizations.

No need to lose hope as incrementality is real, usable and effective in overcoming the limitations of last touch

Don’t lose faith and adopt change! That’s really what all the leaders I watched on all the webinars made me think. They really made me happy to be part of this industry! As challenging as it was in the past to pitch for incrementality ; iOS14 was now an eye opener for many to lead efforts on sustainable growth and promote incrementality! iOS14 has shown us how painful it can be to get data and that deterministic numbers are the best answer to lead investment. 

Many companies now finally consider getting better and better knowledge about holistic views and are pursuing incrementality as a new way to increase and maximize their real profitability. I think this change was also driven by the fact that mobile generated more revenue during the pandemic. Advertisers discovered new channels already heavily used on the web (TV, influencer etc.) and expanded their media mix. Those channels also needed to be assessed in performance and iOS14 was only another reason to go deeper into measuring those incrementalities with a proper measurement approach.

Benefits of incrementality

For a long time, we called user acquisition growth hacking with the famous white and black hats. In the past, we had no rules as everything was new and the legal landscape did not exist. Now it does, and I am really happy about it. I am happy that it is not legal to send IDFA by CSV file anymore. I am also happy that companies have to ask me for my consent. 

As both a user and an advertiser, incrementality is the best solution to respect my privacy and also provide the necessary and accurate information that advertisers need to ensure sustainable growth. Privacy-compliance is at the heart of incrementality as it requires no device IDs. It is simply a clean scientific means of measurement.

Incrementality does not have artificially created attribution windows and prevents the over-attribution of SANs like Facebook and Google. This makes me really proud of our industry – we now can provide all channels the same chance of success. Once again, this is going to challenge how we think and what we think we know. Incrementality measurement is the only approach that now is able to show you the real impact of each channel based on fair and accurate data. I really hope this will reshape our industry and challenge the famous duopoly.

The other thing that I’m really excited about is that cannibalization will now be taken into consideration. For a long time, marketers ignored cannibalization in order not to suffer a decreasing budget. I hope that cannibalization of paid and organic is going to be reduced and that incrementality’s holistic view will provide a better understanding of the true paid impact on organic. Indeed, it can sometimes have a negative impact but there can also be a positive impact on organic uplift for instance, provided by an app burst that increases your ranking. What about your TV campaigns? Now everything will finally be measurable to understand the true effectiveness of all of your marketing efforts.

Working in mobile means that learning is a day to day activity. As challenging as that is, everyday is about making our space a more advanced space. I am really proud to be part of the mobile space. Incrementality is one of the best results of iOS 14 and without it we would perhaps never have this opportunity to make measurement better for the advertiser and more importantly for the user. We can finally heave a sigh of relief because we survived! 

I look forward to your questions. Thanks.
Claire

Claire Rozain is passionate about growth strategy. She is currently a Senior UA Manager in casual games for Gameloft. Prior to that, she was in social casino for Product Madness, in dating for Match EMEA, and in mobility for Mappy. She had opportunities to explore different verticals in the UA and CRM world and is passionate about top of the funnel paid and organic strategies to get the right users. You can follow her YouTube channel here. 

Why do you need an incrementality MMP?

By now, you would have certainly encountered the deficiencies and frustration of last touch, and SKAdNetwork in particular.  While the scars from those encounters make for great conversation starters, they are also important reminders that it is time to embrace change and re-examine your idea of the ideal measurement stack.

You see, as marketers, for the longest time, we’ve become too comfortable with the simplicity and convenience of last touch.  In an attempt to cling to that comfort, we’ve even accepted a very impaired remnant of last touch, SKAdNetwork and the ecosystem that supports it.  In doing so, we have forsaken any attempt at realizing true marketing visibility and effectiveness.  As a reminder, and please don’t flinch, here are the glaring issues with SKAdNetwork:  

  1. SKAdNetwork leaves marketers with only 100 campaign IDs of granularity.
  2. It does NOT offer a way to cohort KPIs by install date.
  3. Critical post-install performance KPIs that are foundational to your mobile growth strategy (like retention, revenue, LTV, and ROAS) have disappeared and have been replaced by a single conversion value with its own significant limitations.

Now, fortunately in order to just survive and function in a “SKAdNetwork world”, you have last touch MMPs that try to piece the fragments of this last touch signal together.  While keeping your last touch MMP makes sense, it still does NOT address the glaring gap that decimates your marketing strategy.  That is where an incrementality MMP comes in due to the following reasons:

  1. An incrementality MMP is fully compliant with Apple’s privacy policy because it doesn’t depend on device IDs (IDFA or GAID).
  2. Does NOT require any changes to your marketing processes, KPIs or methods of use.
  3. Provides you the very same KPIs (cohorted by install date) that you have grown to cherish.
  4. The data that Polaris needs will always be around and doesn’t require any form of attribution.
  5. Can provide side-by-side comparisons of last touch vs incrementality data.

     

    Which is why we built Polaris, a privacy-centric incrementality MMP that does NOT need IDFAs or GAIDs.  It offers the “MMP experience”, but built for incrementality, in order to overcome the deficiencies of last touch MMPs.  More info about Polaris can be found here.

    We know that change is difficult, but fear not, because we have built Polaris to be quick and easy for you to adopt. So easy in fact, that it is a turnkey drop-in replacement for last touch. No migrations, no extra SDKs, and no heavy implementation lift. Polaris even integrates directly with your last touch MMP to get the data it needs.  

    After having helped many leading mobile gaming companies, we are confident that we can get you started with Polaris in 24 hours. We are here for you. If you are trying to close the gaps in your measurement stack, we are offering a chance to run a POC to compare incrementality measurement to last touch.  You will be astounded at how much better Polaris performs.  Contact us today to get started.

Submit a demo request to hear about the powerful features of Polaris incrementality MMP.  We look forward to helping you.

The MetricWorks Team