POST-PANDEMIC CHANGES FOR DIGITAL MARKETERS
White Paper | Authors: Vikram Bansal, Meta, Chandrahas Shetty, Oracle Advertising, Gitali Halder, EssenceMediacom, Ben Poole, fifty-five & Camila Martins, TikTok; IAB SEA+India 2022 Data & Attribution Council’s Measurement subgroup
After three years of unprecedented disruption, budget cuts, short-termism and flexing of business models, what are the lessons learnt for marketers and what is the new normal for marketing measurement?
What’s changed for digital marketers post-pandemic?
Behavioural shifts caused by lockdowns saw the share of digital ad spending accelerate during the pandemic.
However, with increasing spends, the focus on the true business value of digital marketing came under scrutiny. Marketing attribution is the analytical science of determining which marketing tactics are contributing to sales or conversions. Today, marketers deploy multi-channel strategies to carry out advertising campaigns, both online and offline.
Based on the consumer journey, a proportionate weight is given to each media touch point and these insights help marketers customise, and optimise, their media strategy. Historically, attribution methodologies were traditionally last-click, what we call the Click Through Attribution (CTA) model which gives 100% conversion credit to the last platform or medium from which the user converted.
However, behavioural and technological factors continue to evolve. We'll go on to discuss two key phenomena that may impact you over the next 4-6 quarters.
BEHAVIOURAL CHANGES
Ad Blocking - what you need to know
Ad blocking continued unabated with more than 40% of the global Internet population using an ad blocker. Most significantly, Indonesia, Malaysia and The Philippines ranked respectively as the 1st, 4th and 7th ad-blocked countries worldwide. Meanwhile ecommerce saw seismic shifts, with Southeast Asia having the fastest digital sales growth globally, almost doubling from $59bn in 2020 to $102bn in 2023.
The new frontier of video attribution : Measuring the Value of a View
It should come as no surprise that videos are consumers' favourite type of content – based on the latest online video consumption statistics, 92.6 percent of global internet users watch videos every week. In fact, a recent study by Deloitte, showed that 6 in 10 Gen Zs and Millennials spend more time watching user-generated video content than they do video streaming services.
Reels, Meta’s short-form video format, is seeing significant growth. There are now more than 140 billion Reels played across Facebook and Instagram each day. That's a 50% increase from six months ago. On Instagram alone, people already reshare Reels nearly 1 billion times daily through DMs, a clear indication that video will stay at the forefront of consumer consumption habits for the foreseeable future. And it's not just consumers. Businesses should also take heed. InMobi reports that video advertising represents 36% of all programmatic ad spending in the region – with short-form video making up the majority of the inventory for a platform such as TikTok.
A TikTok-commissioned survey by Toluna showed that 9 in 10 users take action after viewing ads on the platform. Based on these statistics, marketing in 2023 demands tailored strategies that include video in a meaningful way. With that in mind, CTA models alone might not paint an accurate picture for categories that traditionally have longer attribution windows.
The customer decision cycle on a video-first platform is even more dynamic with 58% of users delaying their visits to brand websites or apps instead of disrupting their content browsing experience. Some platforms are showing a highly engaged audience with average session durations of 389 seconds.
In reality, this multi-screening behaviour makes user journeys increasingly complex as they consume content and conduct research for brands/products across a wide variety of platforms. This is where marketers should consider a hybrid approach to better understand a campaign's performance.
Here's where View-Through Attribution (VTA) comes in. The VTA model measures the sales driven within an attribution window, which on average in Southeast Asia is approximately 24-hours, but can go as high as 48-hours. Moreover, the study from Appsflyer found that SEA Markets have 24 to 48-hour VTA windows with up to 83.8% conversion rates.
By leveraging VTA along with CTA models, marketers are able to optimise their spending for effectiveness/efficiency and allow campaigns to pass through the learning phases quicker too. Based on internal platform studies, marketers who enabled VTA + CTA models achieve a significant improvement in their campaign performance – showing 113% faster learning period exit, a 79% higher conversion rate, and 10% higher auction stability.
TECHNOLOGICAL CHANGES
The coming transformation of Attribution: Privacy
For the last few years, cookie deprecation and privacy regulations have been the talk of the town. In addition to web browser technology changing, restricting third-party cookies, Apple’s app tracking transparency feature, implemented in 2020, changed the game for platform access to consumer information.
Google also announced new privacy restrictions that will reduce cross-app tracking on Android devices. Google’s extending the timeline of phasing out third-party cookies has raised questions on this move. There is a popular belief that it’s hard for data partners to strike a balance between privacy, personalisation and profits.
Along with audience targeting, measurement and attribution also pose unique challenges in a cookieless ecosystem, particularly since most incumbent forms of attribution are heavily dependent on third-party cookies. Privacy-centric measures tend to make attribution even more challenging.
In a similar vein, Apple and Android have both recently provided users with more transparency on how apps will handle permissions and data. This forces advertisers to rethink marketing campaigns as certain attribution touch points have been limited by these privacy-centric measures.
For the last few years, cookie deprecation and privacy regulations have been the talk of the town. Now, with Google extending the timeline of phasing out third-party cookies it has raised a lot of questions on this move and there is a popular belief that it’s hard for data partners to strike a balance between privacy, personalization and profits.
Apple officially launched a range of new privacy control updates with iOS 15 that allow users to access an easy-to-read report showing which apps have been accessing which permissions, as well as revealing the apps that are tracking your iPhone.
Similar to Apple’s iOS Privacy Labels, Android recently added a “Data Safety” field in Google Play to provide users with transparency on how apps will handle their data. Android 13 imposes more limitations and user controls. Apps now need to request permission to access audio files, image files and video files separately as part of an effort to limit access. Meta has introduced a number of measures to provide users transparency and control including investments in a portfolio of Privacy Enhancing Technologies (PETS) in collaboration with the industry.
With more user controls on tracking and blocking, advertisers now need to rethink marketing campaigns on Apple and Android devices as targeting and measurement can both pose real compliance challenges.
How can digital marketers be ready for the new normal?
As brands are now re-emerging from this crisis and are back in growth mode, what are the considerations for readiness in the next normal measurement?
With a maturing digital ecosystem, attribution models will require integration with traditional techniques like Marketing Mix Model. Often this will mean providing data that is less specific—for example, aggregating interactions and conversions as opposed to providing user-specific ad engagement & conversion.
THE NEW NORMAL FOR MEASUREMENT: BEYOND THE LAST CLICK & IMPRESSION
The impact on attribution and measurement for digital marketing
The implementation of privacy-centric measures in a post-pandemic world, means that advertisers must diversify their approach through experimentation, and use existing solutions in order to thrive within these constraints. It is time to replace last-click measures and focus on maximising the LTV of consumers by connecting conversions to media touch points.
Every advertiser, at the end of every campaign, must still answer the questions: "Did it work?” and “How can I make it work better?” These questions are becoming more difficult to answer. The right measurement strategy helps businesses understand what is working and what’s not, and identifies opportunities for increased investment and better performance. We recommend the following approach:
Question and review your campaign, and creative and measurement frameworks to Plan for privacy via flexible measurement tools and techniques.
Evaluate data access to connect digital marketing to business value and outcomes and Anchor on true value by optimising decision-making with the most incremental methods and metrics available.
Choose the right approach, answering key questions with solutions that accommodate business needs and constraints and create an agile and experimentation-enabled culture in your teams.
Question and review your campaign, creative and measurement frameworks to plan for privacy
Increasingly we see that marketing leaders establish credibility in austere business environments based on how well digital marketing can articulate and attribute value using data in real-time to plan for the capabilities and channels with flexibility.
Take stock of how relevant digital measurement operations are relevant to respond, factor in, or shift to dynamic industry pressures marketers’ face.
How you measure is critical as comparative measurement norms may no longer apply and are becoming unpredictable eg, a global pandemic.
Building for future-orientation, means ensuring post-pandemic comparisons are done recognising changing consumer trends (and not compared with baseline pre-pandemic).
Varying market pandemic-related cycle of opening economy and immobile lock-downs may mean re-evaluating location based incrementality assessments in the short run.
It is difficult to say for certain what the impact of privacy changes on specific measurement techniques will be, but businesses must prepare for a more privacy-conscious future, anticipating that some measurement approaches will be more affected than others, and that this may change yet again over time.
Measurement techniques which will be most impacted are those which are reliant on granular user level data. Primary amongst them is multi-touch attribution (MTA) to understand fractional contributions across devices and channels. At the other end of the spectrum are aggregated and modelled techniques, such as MMM (media market modelling), which are now seeing a resurgence in their application and scaling. Advances and innovations in experiment-based techniques, coupled with access to opted-in conversion data, continue to play a critical role in improving the accuracy and reliability of MTA and MMM.
The key is to understand the potential impact of signal changes on measurement strategy.
Staying informed. Understand types of changes which could impact your business most and track them closely.
Reassess impact. New solutions to mitigate the impacts of ecosystem changes will continue to arise. In some cases, adoption will require evolving business processes.
Actively engage with the industry. Advocate for critical measurement applications with industry partners who are working to shape the advertising ecosystem.
Considering additional measurement techniques. Adding privacy-conscious techniques to measurement strategy, such as those that rely on aggregated data, like marketing mix models (MMM), or data that your customers share directly (like brand lift surveys).
Anchor on true value
Next is to anchor measurement strategy on incremental business value for more precise insights leading to better marketing decisions. Incrementality measurement helps determine how and where marketing campaigns are contributing to bottom-line success. Incrementality is defined as "the degree to which a measurement method estimates the true causal effect of an isolated marketing activity."
As digital marketers will have to make do with less data available, the ability to deliver clearer links between post-pandemic ROI and digital marketing channels is key.
● Marketing leaders with an intentional plan on consolidating digital and offline data will be well-poised to answer questions on business value.
● Connecting data sets across various parts of the consumer journey is required to fill in the gaps between offline and online experiences coming out of lockdown. For this, brands need secure, privacy-first omnichannel commerce partnerships across tech partners, publishers, retailers and service layer to complement in-house marketing data.
● As customers return to greater economic mobility, identifying channels that bring robust ROI across digital, combined with offline marketing mix will be largely dependent on both access to and connectivity of the data available - and measurement champions across the funnel willing to learn new toolkits. Loyalty programs and digital coupon redemption can cultivate ongoing one-to-one connections with customers, bridging both digital and offline data sets. This will continue to update and enhance customer-centric journey maps of data as economies continue to be in flux.
● Channels that connect customers across their shopping journey both online or offline will become more fragmented - so the ones that can link channel association to business metrics such as sales, profits, growth, will be able to demonstrate the value of their digital marketing investment.
According to Meta’s research, almost 25% of the time non-incremental and incremental measurement disagree on a winning marketing tactic. When marketers choose the wrong tactic as a result, businesses lose out on an average 64% average improvement in cost-per-action (CPA).
Incrementality uses experimental methods to help you identify the advertising that drives conversions by comparing the behaviours of people who see or don’t see your ads. Incrementality is more accurate than approaches like last-touch attribution because it doesn’t just assign credit to a conversion.
But it’s not only the method that matters, it’s also the metric. To inform business decisions, as far as possible measure true business outcomes (ie, sales), as they reflect a business’s bottom line, rather than, say, clicks.
Now let’s put this all together. Select the best available incremental measurement by evaluating the incrementality of both method and the measurement metric.
Take MMM, for example, which is a technique that is resilient to change. Calibrating MMM models with experiments provides a relatively high level of incrementality, moving it to the more incremental side of the spectrum. This indicates that calibrated MMM comes closer to evaluating true causal effect, improving the confidence in using it as a decision-making tool.
With a focus on incrementality as the foundation of measurement strategy, choose the tools (or combination of tools) that can best address marketing goals.
Choosing the right approach
With the ongoing changes in the signal landscape having a far-reaching impact on the advertising and marketing measurement ecosystem, a common question is where and how to start on this journey? Here is a suggested framework that promises to remain very resilient, even as the signal and measurement landscape continues to evolve.
● Identify and align on business goals: Start with the decisions which need to be made
● Consider constraints: For each type of decision, identify the business and data constraints
● Select and test an approach: determine the measurement tool or tools that best meets the identified business goals
● Assess and adjust: continually assess if the chosen approach is working well in decision making, and adapt when necessary.
Organisational leaders who have advanced in recovery mode have notably made choices in adopting an agile, experimentation mindset of flexibility instead of reverting to business-as-usual, organisational silos of measurement teams and marketing channel leaders. Measurement is not a role relegated to the back office marketing function; rather an increasingly important function acting as a central ‘spine' of decision-making in marketing.
The most agile leaders early on in the pandemic customised approaches by tactically identifying the cross-channel and in-channel optimal mix required to stay connected to shifting consumer purchasing journeys in post-pandemic times - and were able to show why, in increasingly challenging business environments, marketing brings the sales to support bottom lines.
Through partnership inside an organisation’s marketing functions, the real insights come from identifying patterns of value. These distinguish consumers within connected brand and performance marketing data sets to further segment consumers to understand post-pandemic demand to identify opportunities to to meet consumer needs. This will be key to ensure marketing has a direct link to business outcomes (beyond media metrics alone).
Ultimately upskilling digital marketers to increasingly understand how to combine long-run and short-run methodologies to build a post-pandemic understanding of the new baselines through statistical modelling for digital marketing or modelling with expanded marketing data and sales sources offline will increasingly be valuable to the business ROI conversation.
CONCLUSION
Whether it’s the changing consumption habits that suggest consumers increasingly prefer video, or the recent privacy-centric measures, organisations will need to create a signal-resilient measurement approach to move forward with a KPI-driven digital marketing agenda. Marketers must constantly evolve to navigate signal changes, and a measurement strategy will continue to be a critical tool in their arsenal along with a flexible, agile, decision-making framework.
