2020 changed things. Some changes were disruptive: businesses closed (some temporarily, many permanently); working from home (and just being home all the time) dislodged familiar consumption habits; purchase decisions took on new importance. These changes gave rise to digital innovations for select companies fortunate enough to align with this shift, ranging from new online learning platforms to contact-tracing solutions.
Yet not all pre-pandemic inertia was disrupted in 2020—some change, already set in motion, survived. Within digital analytics specifically, the march towards data privacy continued, machine learning kept evolving, and analytics and media platforms upgraded. These trends continue into 2021.
Businesses not so favorably positioned now face the question of where to focus attention in order to regain orientation and stability in 2021. With fixed, even fewer resources than those enjoyed in 2019, what should they invest in?
Is the answer to spend their way out of a dismal 2020 and chase shifting eyeballs with revised media programs? Do they play the long game and focus on data privacy? When would the right time be, and what are the risks? Do we throw out our anomalous data 2020 and rely on 2019 instead?
With so many dizzying questions, we present our four must-dos (and don’ts) to deal with the fallout from that once-a-century kind of a year, 2020.
Learn beyond the margins
Whether you’ve observed a boom in conversions or struggled through a challenging year, the data that you’ve earned in 2020 extends outside the margins of typical “business as usual” operations. Digital campaigns were completely paused or switched off entirely. Competitors pivoted. Mobile devices took breaks, along with work IP addresses, while home offices sprouted all around. Our optimized machine learning models were evicted from their comfortable local minima or maxima and thrown into a world of uncertainty.
This relocation offers the potential for finding a new or global point of optimization, but also demands a thorough understanding of the mechanics of 2020 for your organization.
While it is very tempting to dial up the resolution on granular first- and second-party data, understanding 2020 is impossible without widening our peripheral vision. And while it may be an oversimplification to blame the fact that performance fell in early 2021 exclusively on the pandemic, to not consider it—and the concomitant effect on global macro-economics—may lead to attempts to recover by setting unreachable targets, or by overhauling a media strategy that, under other circumstances, may have been effective.
We can have a habit of only considering internal data, even though disruption to the market and to society in general are most certainly the largest drivers. We like to focus on what we can control, but adjustment in tactics—like content or bid adjustments—and refinement of levers may be tantamount to fiddling with micro-controls on an airplane in the middle of a hurricane.
Augmenting your organization’s data with external sources to put your KPIs in a holistic context is critical to understand 2020 performance. Use publicly available COVID count data and find proxy data sources that may act as canaries in coalmines for your industry, including:
- Consumer household spending surveys (where are the biggest shifts in finite budgets?)
- Interest and exchange rates
- Media consumption habits (surveys, research, Google Trends)
Data in 2020 was generated in a flurry of rapidly changing events. Existing models may have been asked to perform outside the parameter ranges they were built for, and decisions, often emotion-fueled, had to be made on real-time, short-term datasets. However, as 2020 pushed consumer environments to the extremes of their range, it provides a rich source to develop your models and strategies to respond to extreme events—pandemic or otherwise—that may arise in the future.
Capture golden opportunities
Accounting for other variables when reporting against historical time periods is already a complicated process; with everything that happened in 2020, this process became a lot more involved. Against a changing landscape of lockdowns, working from home, and near-complete closures of certain business verticals, what can you infer from saying you’re up 21% in Q2 2021 against the previous year? Might we have to introduce a metric like YoYBL (Year-over-year-before-last) to cope?
Despite the continued shift to trackable digital media, the John Wanamaker problem remains as relevant to marketers as ever. We all want to know which half of our budget is being wasted but the absence of a comfortable comparison period—paired with incredible new technology—presents the perfect condition for substantial change. If ever there was a time to reform the way success is measured (e.g. moving away from a last-click attribution model), 2021 is the year to strike.
Although the Adobe Summit pivoted to virtual in 2020, it reinforced the immutable fact that AI’s place in marketing technology will continue to surge. For instance, Adobe Attribution AI is there to give you some additional tools to let you uncover the value of each customer touchpoint. In addition to the standard heuristic attribution models we know and love, Attribution AI offers additional algorithmic models that aim to provide you with additional insight into your marketing performance and make better informed decisions around budget allocation and campaign optimization.
The very human reflex to explain the shifts in 2020 performance data also invites customer intelligence into the foreground of organizational discourse. Often there is fragmented, non-uniform changes in customer bases, where certain groups react radically different than others. Adobe Customer AI is designed “to generate customer predictions at the individual level with explanations”. By calculating conversion or churn propensity scores, Adobe Customer AI could provide an additional, behavioral layer to your segmentation. A large dispersion in 2020 results makes the conversation about investing in customer segmentation much smoother in 2021.
Don’t worry about privacy…
At the beginning of 2020, Google confirmed the existence of a ticking clock set counting down the two years it will take for third-party cookies to all but disappear. Indeed, 2020 featured many subsequent innovations and features supporting their commitment towards building a more private web. The technology giant’s plans to usher in an ad-friendly, user-centric, secure internet proved impervious to any shocks brought on by the pandemic and any related emergent browsing behaviors. By year’s end, Chrome cemented its position as the world’s default browser, closing in on two-thirds of the world’s market share.
Privacy projects are slated to continue in 2021. In its TURTLEDOVE proposal, Google discussed the possibility of browsers holding customer information, with ad platforms and websites not having visibility of these data. Criteo put forward SPARROW (Secure Private Advertising Run Remotely On Webserver), which uses third-party “Gatekeeper” to serve interest-based ads. Google Ads’ DOVEKEY develops SPARROW, NextRoll’s TERN builds on TURTLEDOVE, and Magnite’s PARROT proposes an improvement to TURTLEDOVE. It’s clear that there is a lot of research going on in this area so, if you had been worrying about the post-cookie world, you might be pheasantly surprised. Embrace the hobby of casual birdwatching and leave the heavy lifting to the ornithologists.
…worry about your privates instead
With Chrome moving towards the obsolescence of third-party cookies (we assume all other browsers will follow), first-party data steps up to the plate. Digital campaign design will demand media and strategy teams working closely with web analytics and CRM specialists.
Web and app properties are heading towards being the new normal when it comes to your Google Analytics deployment. Google Analytics 4 is now the default version of the world’s most popular web analytics tool. With its move to an event stream-based model, there’s a lot more flexibility in how you use events and parameters to capture your data. However, losing the standard category, action, and label structure may put more importance on basing GA4 implementations on a robust solution design reference to ensure your website is collecting structured, governed data.
Keep calm and understand 2020
After such an unpredictable year, the key is to not only focus on what you can control but monitor and factor in what you can’t control too. If your industry has boomed, now is the time to grab market share—but, in contrast, if you lost business in 2020, it may well be down to extrinsic factors, not intrinsic. There’s no sense pursuing business that simply isn’t there anymore, like planting twice as many seeds on the same sized plot of land after a flooding season ruined your harvest.
If 2021 sees your vertical revert to some kind of familiar normality, your 2019 strategies could still serve you well. Just remind yourself to acknowledge the context of external forces, remain calm, and try not to be overly reactive to things that are almost always beyond our control. For everything else, we’re ready to drive your business forward in 2021. Let’s talk.