When measurement gets tough, the tough get crafty

When measurement gets tough, the tough get crafty

It’s no secret that here at Booster Box we love raising the innovation bar by delivering new projects & tools. And it’s also no secret that we’re the best* at it, according to our moms.

Moms are always right, but we want to prove our awesomeness with more scientific evidence. That’s why we are thrilled to announce we have a brand new shiny tool in our work belt…

 

 

But first, let’s start with what was broken: measurement

Before finding a remedy, we needed to understand the scope of the damage. Measurement is getting complicated, especially when we talk about multi-touch attribution. Several trends are pushing towards further limitations for MTA: the beginning of the Cookieless era marked by the deprecation of 3rd party cookies in the name of consumer privacy; worry over whether brands have over-invested in performance marketing at the expense of Brand building, thus, moving away from Bottom of the funnel initiatives; Platforms being walled gardens and User Journeys being more of a “messy middle” instead of a “Funnel”. Life is complicated and the more complicated it gets, the more platforms or channels it takes place on and the more difficult it is for us to measure. And…”If you cannot measure it, it does not exist!”.

But…

Dear negativity, please close the door on your way out. 

All these trends are challenges but they also mean we have the opportunity to do a better job in understanding the contribution of each channel to the overall marketing mix. 

 

 

So, here comes the fixing tool

You know that feeling when you make planning smoother? Well, that’s exactly how you’ll end up feeling! Any business worth its salt should start its planning with budget optimization and all marketers wanting to make their lives easier should optimise future spends to maximise effectiveness. But how can one plan future success, you may ask. Well, with Marketing Mix Modeling (a.k.a. MMM) by your side, you have the perfect assistant to help you with this task. 

Let’s take a trip down the statistics hole

The MMM approach is based on statistical analysis using advanced techniques to work with top-down, macro-level information allowing marketers to measure past performance and chart a path for future success. By quantifying the impact of various marketing channels as well as offline activities on sales or other specific conversions, it helps brands understand and assess the true value of their investments and then it can forecast the impact of those very activities on future sets of tactics.

 

 

Basically, the purpose of using MMM is to establish which mediums are working better than others, to understand how much each marketing input contributes to sales, and to measure how much room for growth there is in each marketing channel. Practically speaking, by measuring brand sales contribution and ROI for each element of a marketing mix and estimating diminishing return curves, you can easily allocate your budget by shifting money from low ROI mediums to high ROI mediums thus maximising sales while keeping the same budget.

The Booster Box approach

We wanted to give a taste of the medicine so we built a simplified version of the model with a tool enabling clients to create their very first Marketing Mix Model. From this basic version, we can then evaluate if the data is sufficient to use more advanced models. To obtain an accurate forecast, we compare different models, trying to understand which one has the best outcomes for each and every single client. Our approach aims to reduce human influence by using automated algorithms and allows actionable decision making, providing a budget optimiser and easy comparability between models. 

The complete approach to MMM is much more detailed than our first mini tool – which is still useful for taking a quick look at the overall picture -. 

 

The full analysis has higher complexity with the final outcome including:

– Modelling of saturation functions, in order to account for diminishing returns;

– Estimates of the delayed effects of media spends;

– Optimal budget allocation across Advertising Platforms;

– Continuous recalibration of the model

To 🌯 it up, thanks to our entire procedure, you will be provided with priceless insights for actionable decision making. And we promise your future will be so bright it will burn our eyes!

 

The New Frontiers of Privacy: Austria fires shots at Google

The New Frontiers of Privacy: Austria fires shots at Google

Google Analytics better run, better run

Faster than the Austrian Data Protection Authority’s bullet…?

The new year started with a bang, the DPA pulled the trigger, but why? An Austrian website has been using Google Analytics without properly implementing IP address “anonymization”.

The Datenschutzbehörde (could it sound any more authoritative?) has stated that IP addresses are personal data because they can identify visitors if combined with other digital information. In this particular case, the visitors’ data stored by Google Analytics could have been accessed by the US government, thus violating the European GDPR laws and regulations.

The prosecution

Now that we all know the case, let’s move on to the heart of the story.

The Verdict raised several issues with many nuances but the major points we want to highlight are:

  1. A wrong configuration of Google Analytics could give Google LLC access to personal data without EU user consent;
  2. Despite all GA settings controls, Google cannot prevent personal data being accessed by the US government.

In simple words, a wrong implementation of Google Analytics by a local web publisher could provide inadequate levels of protection. Moreover, quite a lot of other online tools have their base in the US or have the possibility to access EU databases without consent. Therefore, if the Verdict were applied to all, we could  see the fall of many publishers and small businesses who use the web.

The defense

What does Google have to say in its defense? Here you will find their responses to all questions asked by the Austrian DPA. The document sheds some light on the debate and includes interesting info about what GA does and what it does not do, its use of data and the controls that are in place to limit data access. Long story short, it is a reassuring document 🙂

And then, come on, let’s face it:

“In 15 years of offering Analytics services, Google has never received the type of demand…speculated about”, Kent Walker.

 

The Booster Box bulletproof vest

 

 

Ok everybody, let’s put down our paper bags!

After having listened to both the prosecution and the defense, our main advice is: don’t panic! We have a whole list of solutions:

  1. First, breathe and wait. Will the USA take notice and amend their data privacy and surveillance laws? We honestly don’t see that coming.
  2. Review your site to check its compliance with GDPR regulations and its alignment with Google best practices.
  3. Remember that you do have control over the data that you collect using Google Analytics. For instance, you can ensure that:
    • full IP addresses are never further processed or logged;
    • data collection is partially or completely disabled;
    • time limits are set for user-level and event-level data before automatic deletion;
    • data is deleted from Analytics servers;
  4. Implement a Server Side solution, ensuring even better control over the data flow and preventing GA from ever receiving certain PII, e.g. the user’s IP address. This will give your data another layer of protection as you will be the one deciding what info is sent to the Server.
  5. Make sure you select the European hosting region when you set up a server Google Tag Manager for the EU website.

Finally, “don’t panic” doesn’t mean you should forget about it. If you have a EU business presence, we strongly suggest you do not miss any further developments in the coming months.

And okay…if you say please, we vow to monitor the situation for you and share major updates. Just promise you’ll stay tuned!