An Analytics Carol

[2023 Update] 2022 is drawing to a close and we can say that it has been a year of changes in the world of data and analytics.

Surely something has been lost compared to the past, there have been changes in the present and many expectations for the future. Almost like the three ghosts of Christmas. From this point of view, we interviewed some of the beautiful analytics-data minded professionals by asking questions about past, present and future analytics, creating the Analytics Carol!

Among the interviewees: Ana Paula Maluf, Jim Sterne, Mikko Piippo, Steen Rasmussen and Tim Wilson! [2023 Update] now we have onboard Brian Clifton, Camille Chaudet and April Siviglia

• Ghost of the past: Is there anything you miss or hate in the past years in the world of analytics and data analysis?

A.P.M.: Miss it being easier… There is more and more concerns and set ups regarding consent and data privacy regulations to specific regions now.

J.S.: I really miss the ability to meet in person in safety! I really hate the current “tri-demic”!

M.P.: I miss the past years when everything was simpler… Of course, the complexity of measurement, technologies and legal environment can be very profitable for consultants. There simply is a lot to do. At the same time, it is difficult to stay up to date on everything one needs to know.

S.R.: Both a hate and a miss… All the data that people gathered which actually was never used. So on one hand the tendency to want to track everything, but use nothing and getting stuck with maintaining everything versus all the lost potential from the data of Christmas past.

T.W.: I don’t miss how hard it often was to actually get data from key systems: putting in a request to a data warehouse team, fighting to get it prioritized, waiting for the development to happen to actually bring the data in, and then finding out that it had some issues that required repeating the process. It’s not that this has been perfectly solved now, but the increased proliferation of APIs that provide direct access to data in different systems, as well as the evolution of the data stack, has often made it much easier to get low-latency, granular data from multiple independent platforms.

B.C.: I miss the simplicity of contextual advertising. Although technically this was hard to do well, it made obvious sense to me. Essentially Google (the leader in this field circa 2005), placed an ad on pages based on the context of the actual content and the search terms/referral info that led the visitor to the page. In other words, I miss the ad ecosystem *not* being the digital surveillance of individuals that it is now.

C.C.: HR : difficulties in recruitment compared to the market needs, GDPR and legal evolution

A.S.: I was not fond of all the custom code and workarounds we had to develop to capture essential click elements on a page.

• Ghost of the present: This year has been quite daring in terms of data analysis: for example we have come out of the pandemic “bubble” and in Europe GA Universal has been declared “Illegal” as the data transfer from EU-US. What considerations can you give us about this year based on your experience? Improvements? Worsening? Other?

A.P.M.: GA4 is still under construction… but will get better soon… lots of improvements ahead

J.S.: 2022 has raised the awareness of analytics much higher than in previous years due to all of the attention it’s getting. Data science is no longer sexy but necessary. There are finally enough case studies – proof of value – that the use of analytics is a matter of corporate culture rather than an ROI calculation. If your company is not all-in on analytics, it’s time to find a new place to work.

M.P.: For me, 2022 was finally the year when people understood there are alternatives to GA. Previously, only regulated industries like healthcare, finance and insurance seemed to consider any alternatives to the big players (Google and Adobe). Marketers, of course, mostly want to use GA but nowadays also they are willing to have a look at alternatives.

S.R.: Well the frustrations that everyone who wants to talk about this talk about Google Analytics, my dream would actually be that people learned that there is not one, but 3 different versions of the platform, if we include Firebase. Also the level of panic some people try to spark… “are you ready exorbitant fines?” Well you shouldn’t be… is Google Analytics a radical breach of personal data… not really compared to people loosing bank accounts, social security numbers and medical data by the buss load. For a lot of companies it’s like freaking out over the windows are not clean, when the house is on fire.

T.W: I absolutely love that cookies and pixels are rapidly losing their place as a dominant mechanism in the data collection ecosystem for two reasons. First, they’ve always been creepy and have consistently overstepped what consumers were actually comfortable with. Second, they promoted the fallacy that, with “complete” user- and interaction-level data, “actionable insights” would be easy to come by. The latter, in particular, drove the conflation of “attribution” (an exercise in choosing how to assign value to activities or interactions) with “measurement” (of incremental impact; “incremental” is the key here). I’m excited that the entire marketing ecosystem is going to have to fundamentally shift how they think about data and analytics!

B.C.: The issue of Schrems II has been waiting to bubble to the surface ever since GDPR launched in 2018. It’s awkward and annoying, but ultimately the wild west of data collection is coming to an end. At least it is the beginning of the end – and that is a good thing. Expect a lot more of this going forward as other parts of the free world realise citizen privacy is a valuable thing that needs protecting. Just like democracy, it cannot be taken for granted.

C.C.: I feel concerned about Data Ethics so GDPR was a good thing for me personnaly. It enables me to be aligned with my ethics standards, as users are now aware of data collection and can monitor their consent. On the other hand, I feel concerned about the analysis we provide, as we just analyze a sample of the users. Are the insights and conclusions still relevant and actual?

A.S.: We will see more robust data transfer laws being enforced faster globally. Find one standard solution you control and apply it to all sites globally, allowing marketers to adjust to the new data metric points and patterns while strengthening your implementation formula before it is needed.

• Ghost of the Future: Simple but not trivial question: how do you see the future of Analytics in 2023 and beyond? Apocalypse AI like in blade runner or the beginning of a new era?

A.P.M.: Beginning of a new era that will require more skills and motivation to keep studying

J.S.: New Era! Combine thousands of highly-skilled technical experts, laid off from their jobs with lots of time on their hands, with ChatGPT and whatever comes next as a platform for a Precambrian explosion of innovation. This will naturally spill over into analytics and create all sorts of new tools and processes. It’s going to be a wild ride!!

M.P.: I don’t think there will be any huge changes in 2023. Server-side tag management, data warehousing will slowly become the standard. But it will take time until SMBs understand how limited the free tools are – and analysts will struggle in proving the paid tools can provide enough value for SMBs.

S.R.: It’s always the beginning of a new era… if it is not a legal problem, then an ethical one or a new platform. Analytics is the art of extrapolating insights from imperfect data and I don’t see that changing anything soon… AI or not.

T.W.: I think I fall in the “new era” camp. It’s not that AI is going to make things any easier and allow analysts to just push some simple buttons and have insights and recommendations magically emerge, but there’s no way to NOT get excited by the promise of evolving analytics tech. And, who knows? Maybe ChatGPT will help me find the perfect set of analogies to help business stakeholders internalize what they really need to understand when it comes to causal inference!

B.C.: The death of visitor profiling… However Visitor Profiling != Web Analytics. That is, just because privacy is important and needs protecting, it does not mean all data is bad. There is HUGE value is collecting benign, aggregate data – for businesses, schools, hospitals, governments… we all need data to make informed decisions. For example: Knowing how many visitors came; How long did they stay; What pages/products did they look at – was it just a glance or did they “engage”; What campaigns were running or did they interact with to arrive on the site or visit your store etc. These benign data points apply to both online and offline worlds and are benign because they are aggregate by design. Grouping data in this way means it is not possible, or required, to profile an individual, let alone try to identify them. I also think contextual advertising will make a come back. It has to… what choice do the Adtech players have?

C.C.: I wonder if Digital Analytics will still exist within a few years, with GDPR, GA4 and all the evolutions that lead us to collecting less and less data. I also see a huge needs from companies to lead their business data driven, especially the ecological and ethical transformation. As Digital Analyst, we might have to switch to other data bases. With this vision, we shall focus on analysis and other strategic and consulting skills, and not on a specific product like GA.

A.S.: It is the beginning of a new era. Buckle up. Enjoy the ride.

Thanks to these great data-geek for taking time answering to the Analytics Carol. What do you think? Which are your analytics ghosts? Let me know in the comments!