Akhila, who works with brands such as Diageo, DFS, and Kellogg’s, tells us why she loves Cadbury ads, what can be done to improve diversity in data and analytics, and which book all marketers should read.

What’s the burning issue that brands are talking to you about?

TV advertising inflation is a big issue that brands are facing at the moment. Alongside wider macroeconomic and geopolitical trends, such as the cost-of-living crisis, rising interest rates, and the war in Ukraine, it is putting pressure on brands’ profit margins and their ability to increase marketing spend. Thanks to valuable lessons learned in the past, brands realize the importance of maintaining overall advertising investment. Using market mix modeling to enable robust, granular measurement is one way they can better quantify, forecast, and align on the optimal channel mix to achieve both short- and long-term growth targets.

What’s the best advert you’ve seen recently and why?

I have been a fan of Cadbury ads for several years. My recent favourite was the ‘Mum’s Birthday’ ad for their ‘There is a glass and a half in everyone’ campaign, which won the IPA Effectiveness Awards Grand Prix award 2022. Cadbury campaigns focus on the brand’s intrinsic purpose – the generosity within all of us – and this one clearly demonstrates how purpose that is connected to brand and product can drive affinity, equity, and sales. 

When it comes to data and analytics, the industry has a long way to go in terms of diversity. What can be done to improve this? 

An open culture in which everyone feels they belong, and visible, inspiring role models are key to making our industry more diverse. Providing relevant training, coaching, and formal mentorship programmes can also help to drive change, although this will take time. 

I think it’s vitally important that the executive leadership, diversity, equity, and inclusion (DEI) councils, and employee resource groups (ERGs) work in harmony. Executive leaders need to be the role models of the diverse and inclusive organisation they aspire to become – we must be aware, ready to learn or confront our own blind spots and our own habits. For their part, DEI councils need support from the leadership and have clear objectives, outcomes, and actions. Meanwhile, ERGs can provide a safe space and sense of community and belonging for members of underrepresented communities. Here at Gain Theory, we have two ERGs dedicated to diversity – Lumena inspires, empowers, and supports our women, while Roots champions greater ethnic and cultural inclusivity. 

Beyond this, change at the grass-roots level is needed too. More awareness around DEI at schools will encourage more applications in STEM subjects to universities from people with diverse backgrounds.

What book would you recommend to our readers and why?

How Brands Grow by Byron Sharp is a must-read for anyone interested in marketing or effectiveness. It would be fair to say this book had a disruptive effect among both marketers and academics. The author’s views about the importance of new customer acquisition driven by increased mental and physical availability, resulting in behaviourally loyal buyers, are seminal.

Contact Akhila to discuss any of the issues raised in this Q&A.

Photo by Nabil Saleh on Unsplash


Find out how we can help, demo our products and get more information about our global cross-sector experience.


If you think you have what it takes to make a difference at Gain Theory, please check out our careers page.

This article was originally published on The Drum website. Click here to view. You can listen to the related podcast ‘How are marketers balancing privacy and personalisation?’ on Spotify and Apple Podcasts.

Measuring the effectiveness of media campaigns over the short term continues to be a significant challenge for marketers. On the one hand, privacy regulation and changes to how tech companies manage data have rendered solutions such as multitouch attribution (MTA) obsolete. 

On the other, marketers are still under pressure to make sure every ad dollar works as hard as possible amid macroeconomic uncertainty and consumers moving into an expanding universe of digital platforms. 

The good news is there is an approach that can square this circle and deliver granular, privacy compliant insights when marketers need them.

An uneven global playing field

Although Google has again pushed back the date on which third-party cookies will be phased out in its Chrome browser, the direction of travel remains clear. Moves by the world’s largest technology companies to limit the extent to which online consumer behaviour can be tracked on their platforms demonstrate that the destination is a privacy-first world.

At the same time, regulators around the world continue to bear down on companies that do not comply with a smorgasbord of privacy legislation. The UK’s Information Commissioner’s Office fined a catalogue retailer £1.48 million for breaking data protection and electronic marketing laws in October, for example. 

The implications for MTA, which has promised to determine the value of each digital touchpoint an individual consumer makes on the way to a sale, are clear – it is no longer fit for purpose in a world where access to user-level data is no longer guaranteed.

But marketers remain unsure about the best way forward. While 92% of CMOs at multinational companies are prioritising an ethical approach to their use of data, half do not know what this means when it comes to the processes and practices they need to apply both internally and across their marketing supply chains, according to a World Federation of Advertisers report.

The power of granular geographic data

Access to privacy compliant data that enables all media to be measured and optimized on a timescale that gives marketers the insights to demonstrate the effectiveness of their campaigns is possible.

The glue that underpins this data is geography. Think about the delivery information that accompanies every ad placement. Whether it’s a 15-second TV commercial that runs on a particular network during a specific time slot, a digital display ad that features on a dedicated section of a website, or a paid social media campaign, each contains information about where the individual media impressions were served.

In the UK or US, for example, delivery metrics for every ad placement can be accessed at the postcode or Zip Code level. Crucially, this data does not contain personally identifiable information (PII). Unlike MTA, this means that it is compliant with data protection regulation by default. 

When you layer key KPIs, such as sales metrics, and other factors, such as local economic conditions, promotions, and weather on top, you can start to build up an accurate picture of how your campaigns are performing across a large, statistically significant number of micro-geographic areas.

What not everyone realizes is that there is relative homogeneity at a postcode or ZIP code level – in other words, there are striking similarities between people who live close to one another. This means that marketers can compare how ads resonate in micro-geographies based on key characteristics, such as age, gender, income, occupation, and more.

Want to compare how successful your paid social campaign was with high-income males in one city versus another? Geographic media delivery data combined with other privacy compliant information enables you to analyze its impact at this granular level.

Introducing foresight into your marketing strategy

But as well as providing you with hindsight about the past, these insights can also help you with present and future campaigns. Getting information in near time – via data automation and advanced modelling methods – allows us to understand performance across different geographies. This means you can optimize campaigns on the fly. If an advert is working in well in one area but not another, for example, or with one audience but not another, you could change the underperforming creative unit or pull the campaign entirely and reallocate the budget. 

By understanding audience performance across micro-geographies, you can improve who you target in the future and maximize performance. To enable this, it is important to focus on audience attributes that are buyable within programmatic media platforms. These attributes ensure that your measurement-to-insight-to-activation loop is a complete circle.

Crucially, all this can be done while using data that is compliant with a privacy-first approach to build trust with consumers. 

One global automotive manufacturer used Sensor, Gain Theory’s multichannel attribution solution, to measure and optimize across omnichannel campaigns that aimed to increase test drive numbers and, ultimately, sales of a specific vehicle. By identifying demographic attributes to target and household income filters to apply, Sensor was able to explain car-buying behaviour and media responsiveness. All told, test drives grew 18% and $56 million in incremental sales were generated.Measuring the effectiveness of media and advertising may feel like a complex and daunting prospect in the current data and privacy environment, but a tried and test approach, based on granular geographic data, can deliver improved results for advertisers and peace of mind to consumers.

Request a Sensor™ demo by contacting us here.

Photo by NASA on Unsplash.

Russell, who has a focus on new media measurement at Gain Theory, shares his thoughts on key issues in marketing effectiveness, his favorite current advert, and the commentator he leans on to get a fresh perspective.

What important marketing effectiveness question should readers be asking themselves today?   

How do I unify and centralize my approach to marketing measurement without compromising on timeliness and depth of insights? Centralizing measurement can enable coordination and accountability, but it can also slow things down and force data and insights to be aggregated. A thoughtful long-term data strategy and approach to measurement can help to solve these challenges. It can yield deeper insights that are timely, not just at the market and channel level, but at the individual ad placement and audience level too.  

What are you working on currently that excites you?  

The current release of Sensor Audience – a new audience insights capability within our Sensor attribution tool – has been super exciting and rewarding to work on. Increasingly, brands need to know how they can construct new audiences that are based on media responsiveness, as well as brand propensity, and then measure their in-market performance. Sensor Audience makes this possible, so being able to deliver this to our clients makes me feel like we’re arming them with a complete set of measurement solutions.      

What’s the best advert you’ve seen recently?  

I liked the Toyota “Keeping up with the Joneses” ad. It was a very clever and fun way to deliver a message that spanned generations. 

Which person is worth following to get a fresh industry perspective?   

I enjoy NYU Stern Professor of Marketing Scott Galloway. Much of his commentary covers the intersection between business and culture, and he doesn’t shy away from difficult topics. I agree with his view that the most important part of making a prediction is forcing yourself to think through the logic of why your prediction is what it is. This is critical in our business as we build our planning and forecasting tools to deliver insights for clients. 

Who inspires you and why? 

Our junior team members at Gain Theory. Their ability to adapt and grow during COVID-19 – a time when regrettably we couldn’t provide enough face-to-face mentorship to them – has been incredible to witness. Their enthusiasm for our product roadmap and our vision for what Gain Theory can become is also inspiring. A major challenge in analytics is analysts feeling like they do the same thing day in, day out. But the enthusiasm of our new recruits tells me that they see Gain Theory as a place where they will continue to learn for years to come. 

What’s the best piece of career advice you’ve received?  

“It’s a marathon not a sprint” is quote from every good boss I’ve ever had, but it doesn’t mean “slow and steady wins the race”. Marathons are highly strategic and feature moments that are extremely intense as well as moments that allow you to gather yourself and think. To win a marathon you must be highly conditioned, in both mind and body, and you must be a tactician. 

If you’d like to speak to Russell about new media measurement, please click here

Photo by Prateek Katyal on Unsplash.

This article was originally published on the Ad Age website. Click here to view.

I spend a lot of time talking to marketers about how to make their media investments work harder and reduce the pressure they’re under from other stakeholders. Given current events, it’s no surprise that the need to understand their audiences better is front and center of these conversations.

In a recent article, the International Monetary Fund noted that the global economy is facing “an increasingly gloomy and uncertain outlook” amid slower growth, higher-than-expected inflation and war in Europe. The repercussions of this backdrop for consumers are profound: Their purchasing power is reduced. Marketers who are experiencing media inflation and budget cuts themselves are facing implications that are just as clear. They need more reliable and timely insights about audiences’ changing behaviors to help their organizations through these testing times.

This uncertain picture is complicated further by ongoing changes to the data economy. As two MIT Media Lab professors wrote in this Harvard Business Review article, consumer mistrust, government action and market competition are converging to force businesses to make fundamental changes to how data is sourced, managed and used. 

Unlocking faster, smarter, deeper insights

Marketers know all too well the challenges surrounding this trend. For example, from a measurement perspective, multitouch attribution tools are no longer fit for purpose in a privacy-focused world. While marketing mix modeling remains a reliable way to measure the long-term, strategic impacts of media campaigns, what’s missing is a way to measure where, when and to what extent consumers are interacting with campaigns and delivering growth over the short term.

The good news is that there is a solution to help solve this conundrum. Here at Gain Theory, we’ve developed Sensor, an innovative tool that provides deep insights about audiences that marketers can use to optimize all their media on a weekly or monthly basis.

Sensor has been delivering real-world results for several years. One multinational beverage company, which measures the relative effectiveness of its media spend across digital platforms, credited Sensor with delivering a 30% year-on-year improvement in media ROI

In addition, a global automotive manufacturer, which used Sensor to measure and optimize an omnichannel marketing campaign, saw the campaign deliver $56 million in additional revenue and a 2.5% uplift in margin.

Crucially, companies like these are benefiting from being able to measure the impact their media investments are having on audiences with privacy-compliant data. 

Want to compare how successful your paid social campaign was with high-income males in Miami versus low-income males in San Francisco? Using granular geographic data, every ad placement on every channel can be analyzed on a daily basis. It can also be combined with other data such as local weather, economic indicators and promotions to create deeper insights.

Want to know how effective your radio advertising is at driving different audience groups to shop at your online store, compared with ads served on podcasts? Thanks to cross-channel analysis, this data can be displayed on a dashboard that enables marketers to make smarter investment decisions.

Want to see if millennial moms and Gen X moms respond differently to TV ads about the same product but with different creative on two different networks? You can measure the effectiveness of ad placements at a number of levels, including creative, network and partner, to understand how each one resonates with a specific audience group at a granular level.

Three key business benefits

Having a more in-depth and timely understanding of who is interacting with your media, as well as how and when, enables you to do several important things to drive growth.

First, you can act tactically to improve ROI—for example, determining if a campaign, channel, creative or network is resonating with one audience group but not another, so you can change the one that is underperforming or reallocate the budget.

Second, you can make investment decisions more quickly by gaining access to relevant data in a matter of weeks.

Third, you can build trust by using data that is compliant with a privacy-first approach to advertising. As every marketer knows, this is now a strategic necessity.

Request a Sensor™ demo by contacting us here.

Photo by Ryoji Iwata on Unsplash.

This article was originally published on the ANA website. Click here to view.

Marketing teams around the world are thinking long and hard about the “R” word. From New York to London, it’s hard to avoid stories that predict an upcoming recession. This report from the World Bank warning that the risk of a global recession in 2023 continues to rise is just one recent example.

Amid such forecasts, marketers are asking some fundamental questions as they fight to retain their budget. Conversations we’ve had with heads of growth and customer acquisition at e-commerce and direct-to-consumer brands have revealed they are under pressure from the C-suite to help them navigate the uncertain macroeconomic environment. In particular, they want to know: How will a recession impact my business? What can I do now to soften the blow?

Economic predictions, including those that aim to determine which indicators will anticipate how a specific brand is impacted by events such as a recession, are notoriously difficult to get right. But there are techniques that can be employed to help marketers prepare for what might be coming down the track.

How to estimate the impact of a downturn

For one client, Gain Theory used a “composite series” comprised of various time series selected from a wide range of economic indicators and sectors. Based on this, we were able to estimate the decline that an economic downturn between September 2022 and March 2023 would have on two KPIs – total order volume and the number of customers signing up for their service – at varying degrees of confidence.

Knowing the impact of events on key metrics is important, but it’s just as crucial to have a plan of action that enables you to mitigate any deviations from business goals and expectations from the C-suite.

Based on the modeling, we recommended our client increase their media budget in Q4 2022 and provided them with an appropriate mix of channels, partners and publishers they should target. Instead of purely focusing on growing the market, which is less realistic in the current circumstances, we put more emphasis on protecting and growing market share against targeted competitors.

The importance of using hindsight, insight and foresight

But we don’t just rely on foresight. Experience gained by working with brands during the last global recession in 2009 has informed many of Gain Theory’s recommendations to current clients. As well as knowing what the impact on a brand might be, for example, we know that shifts often happen on a consumer and segment level, which is why the right mix and distribution of the incremental budget is key.

History shows that customers who purchase less frequently and spend less on average can disappear, while others are likely to reduce either frequency or basket size. The flip side is that businesses can gain new customers who are trading down from more premium brands or are altering their behavior. Think about those consumers who, when inflation rises, move to a supermarket that offers more value brands, choose to order takeout instead of visiting a restaurant in person, or trade down from a premium spirit brand to a mid-range alternative.

To understand the impact of these different purchase behaviors, we can simulate potential scenarios. This helps to inform how to adjust media messaging, creative, promotions, and pricing strategies to maximize the potential of these changes in behavior. For example, messaging to existing customers can be tweaked to raise awareness of products that cost less or highlight promotions, free trials can be promoted to take advantage of potential new customers, and sign-up processes can be simplified to make switching quicker and easier.

Why “wait and see” is not a strategy

We understand that these are worrying times for marketers and the brands they work for. But waiting to see what will happen is not a strategy. The good news is that the proactive steps outlined above enable brands to ensure their business is in the best possible shape, whatever the economic picture looks like in the future.

Using data-informed analysis of how KPIs will be impacted and a practical plan of action to adjust marketing campaigns, ambitious brands can deliver growth even in a recession. Contact us to learn more.

A unified system fuses hindsight, insight and foresight to help manage short and long-term decision-making for an uncertain future

Planning for future uncertainty is a hot boardroom topic. It’s especially true now as we continue to deal with the fallouts from the COVID-19 pandemic: dramatic changes in customer behaviour, shifting media consumption, supply chain challenges and overall economic upheaval. 

History shows that, in times of disruption, organisational resilience depends on adaptability and decisiveness. Yet many organisations base decision making  on hindsight (understanding what happened) and insight (understanding why it happened).  

Living in a VUCA world

In a world now plagued with volatility, uncertainty, complexity, and ambiguity (aka a “VUCA” world) – hindsight and insight are no longer sufficient to accelerate sustainable growth and gain competitive advantage. 

To quote Professor Klaus Schwab, the founder and executive chairman of the World Economic Forum: “The pandemic represents a rare but narrow window of opportunity to reflect, reimagine, and reset our world.”

We now have an opportunity to reimagine how we manage uncertainty and navigate towards accelerated and sustainable growth. To do this, organisations need foresight. 

From uncertainty to manageable risk

Foresight takes uncertainty and turns it into manageable risk. It considers what is most and least likely to happen in the VUCA world and reveals the signposts and probabilities of them happening. 

Some fascinating examples of foresight already exist. Who would have foreseen, for example, that the Chinese e-commerce platform JD.com would outplay its competitor Alibaba by reliably delivering goods when lockdown hit (and increasing revenue by 21 per cent  in the first quarter of 2020) while Alibaba struggled to find couriers. Recruiting the right volume of couriers at the right time required foresight.

Whether it’s the shift to hybrid working, skyrocketing growth in e-commerce or chip shortages affecting the supply of new vehicles, VUCA events will continue to affect all organisations for the foreseeable future.

Scenario analysis by McKinsey shows that a single, prolonged supply-side shock would wipe out between 30 and 50 percent of one year’s earnings (EBITDA) for companies in most industries. The same study classified different types of shocks based on their impact, lead time, and frequency of occurrence, ranging from theft and common cyberattacks to pandemics and climatic events such as hurricanes. But knowing which possible events to focus on is not an easy task.  

The goal to accelerate digital transformation via online selling and marketing started in 2013 for L’Oréal. When the COVID-19 pandemic hit, the company was well positioned to shift ad spend online. In the first quarter to the end of March, ecommerce sales grew 53% in comparison to the previous year. L’Oréal had invested in this future lever of growth and it paid off handsomely at a time of unforeseen market upheaval. 

Diagnosis before cure

But how do you continue to make good decisions and successfully plan in a VUCA world? First, it’s important to understand what types of challenges your organisation is likely to face. Before you can cure, you must first diagnose. 

Are you facing volatility resulting from an unpredictable supply chain? Or uncertainty caused by new entrants competing with your products? Or complexity due to regulation changes? Or ambiguity raised by not knowing if the current vaccine will protect the population against new variants? Knowing the type of challenge, or combination of challenges, you face is an essential first step.

To make data-informed decisions that accelerate sustainable growth, organisations need faster, smarter, insights that are unified into a single version of the truth to understand what has happened, why it happened and what to do next.

Foresight is key to the unified version of the truth; it enables decision-makers to war-game and simulate; scenario plan and optimise; identify and track leading indicators and to understand the risk and impact of “wild card” events.

A single system for all-round future confidence

We have a unified decision-making system at Gain Theory – HiFusion – that fuses hindsight, insight, and foresight, across all sources and levers of growth, both known and unknown. It identifies long-term strategic opportunities and answers near-term tactical performance questions. 

In the HiFusion decision making system, there is a rich arsenal of foresight techniques to uncover future states of demand depending on the VUCA challenge faced: 

 Future Back Thinking: defining the desired future and then working backwards to identify the potential specific actions and signposts that connect the desired future to the present. For example, setting a target to be net zero carbon by 2030 might involve several potential actions such as planting more trees or swapping to solar power. However, the signposts, such as famine and food shortages, might indicate that planting trees is no longer an option and that we need to switch to solar power as land costs soar. 

• Delphi Method: developed by Project Rand in the 1950s, the method uses a group of experts who anonymously and repeatedly reply to questionnaires about the future, receiving feedback as a statistical representation of the group’s response. The aim is for the group response to converge after each iteration of the questionnaire, ideally resulting in consensus of expert opinion. While the overall accuracy is mixed, the method is used when looking at long-term trends in policy making and technology development.

• War Gaming Simulations: simulating different competitive settings and the impact on consumer demand. For example, the increase in demand for used vehicles during the chip shortage, depending on the duration of the shortage and the actions of competitors.

• Forecasting: projecting future events by taking signals from the past mixed with likely future states caused by internal and external factors and shocks, such as pandemics, international tension, climate change and competitor threats. For example, how will grocery evolve to meet the challenges of rising interest rates and potentially lower demand in 2022/23?

• Lead Indicators: measurable signals that provide an early warning system or prediction of what the situation could be in the future. For example, new car registrations used to be a reliable lead indicator of economic strength and consumer propensity to spend. But in the face of supply constraints, how reliable is this indicator today?

• Scenario Planning: making assumptions on what the future is likely to be and how your business performance might be affected, for the purpose of creating a more robust strategy. Often scenarios are run under “what if?” assumptions to work out the best possible path. For example, looking at the impact on both short and long-term sales and profit resulting from pulling all advertising in the final quarter of the year.

 Trend Analysis: collecting qualitative and quantitative information to spot patterns or new trends, and thus paint an indicative picture of what might happen. These trends can often be used later in scenarios. For example, Wunderman Thompson’s “The Future 100: 2022” which forecasts 100 trends to watch in the coming year or GroupM’s “This Year Next Year” forecast of ad spend, used to augment media response curves in future marketing campaigns.

By fusing hindsight, insight and foresight, HiFusion enables organisations to identify future sources of growth and give confidence to decision-making across all levers of growth. It enables organisations to invest in the right data sources, technology, and methods to confidently answer questions around: 

• the role of brand equity and customer experience in maintaining price resilience

• the short and long-term impact of investments

• drivers of customer choices and product trade-offs 

• identifying and tracking signals that provide an early warning system

• targeting new audiences based on their motivation

• acquiring new customers in a cookie-less world of walled gardens

• conducting complex multi-test-and-learn experiments at scale

• preventing customer churn

• growing spend from existing customers by making the right recommendations

• …and, ultimately, judging correctly what is likely to happen in the future.

HiFusion creates a holistic loop to support strategic decision-making by weaving foresight, insight and hindsight across time horizons and a range of stakeholders. It unifies tactical execution, plus annual and longer-term strategic planning into a single measurement system. 

In a VUCA world, navigating the range of future possibilities to accelerate sustainable growth requires an organisation to invest with confidence. Foresight gives us the ability to manage uncertainty and, ultimately, create and tap into future states of consumer demand. 

By knowing tomorrow, we can disrupt today. 

This article was originally published in Campaign – click here or download the PDF on this page.


  • We live in an uncertain world, where accurately predicting the future is nigh-on impossible.  
  • Companies that want to accelerate growth will use data and analytics to look at a range of likely outcomes to make data-informed decisions on where to place bets.  
  • Gain Theory has helped ambitious brands understand the likelihood of a range of outcomes, leading to more confident bets.  

Navigating Probability in a Time of Flux  

Uncertainty rules the roost. Covid, the state of the economy, demographics, trading patterns, logistics are all in a state of flux. To say nothing of environmental challenges likely to arise through changing consumption habits and government regulation. The slowest rate of change was last year and so using foresight capabilities to steer decision making has never been more important. 

With so much change, brands need to plan for many different contingencies. And they need to recognise which signposts are the most important in signalling that one outcome has suddenly become more likely than others. By explicitly modelling the probability of different contingencies we can help translate this uncertainty into quantifiable risk. 

The Direct and Indirect Impact of States of Flux  

To focus on just one example, consider the economic situation. We know that the state of the economy has a fundamental impact on the ability of all companies to grow their business sustainably and profitably. The economy can have a direct impact, the less confident people are, the less money they have in their wallet, the more they’re likely to seek value, and vice versa. But the economic backdrop can also have an indirect effect, changing the impact that each marketing $ invested has as the economic situation changes. 

We know that this indirect impact can be dramatic. In a recent analytics project, we found that the average uplift on sales of all marketing levers was worsened by almost 30% as key economic variables moved against them.  

Using probability theory to determine potential outcomes and manage risk  

Now, what does this mean going forward? Well, it depends on our expectations of where the economy will be over the next 12 months. Will it improve or worsen? If it improves, will it be a little or a lot? And what do we mean by “a lot”? 

hindsight vs foresight

This is where foresight comes into its own. Using an approach grounded in probability theory we can determine how likely each potential outcome is. We’ve shown this for youth unemployment in the chart – we have a central projection (solid line) and a range of feasible outcomes either side of this. 

The further we get from the central projection, the less likely the outcome. For example, looking at worsening unemployment we can see that while it is possible to jump to a rate of 15% by Q4 2022 it is not very likely. In fact, there is just a 2% chance that the outcome will be as bad as that. So, we shouldn’t spend too much time scenario planning around this.  

But there are a lot of credible scenarios where youth unemployment could move by 1 or 2 points – from roughly 10% today to say 12% by mid 2022. There is a smaller chance that it will stay at current levels. What plans can we put in place now that means we’ll be ready for each eventuality. Business plans that have a ‘business as usual’ approach baked in will almost always lead to poor performance. Alternatively, plans that have uncertainty baked in will generally perform much better and the executives using these plans will have more control over their future. 

Using these and similar techniques, Gain Theory is helping brands to navigate the future, using foresight to translate an uncertain world into more manageable contingencies.  

Exec summary:

Privacy updates made by Google and Apple will have implications on how media is delivered, what information can be captured, and how media performance can be measured.  These actions won’t kill major sectors of digital media and media measurement, but they will create significant challenges.  New processes will need to be built to account for these privacy restrictions.  These new processes will likely favor larger media partners which will continue to fuel media consolidation.  We expect that these updates will drive adoption of more privacy-compliant Multi Touch Attribution alternatives such as Gain Theory’s Sensor.

Key implications:

  • How legislation (GDPR, CCPA) and business decisions by the major internet browsers (Safari, Firefox, Chrome) have limited data capture, media delivery, and measurement over the past few years
  • How Google’s decision to block third-party cookies/tags will impact Chrome
  • How Apple’s release of iOS14 will impact user opt-in requirements on apps
  • The difference between “first-party” tags and “third-party” tags and the implications on consumer data capture and media delivery standpoint
  • The implications of third-party tag and cross-app restrictions on AdTech, i.e. the challenges of server-to-server integrations to audience pool sizes, retargeting, and implementation bottlenecks
  • The impact on Multi-Touch Attribution and potential alternatives such as Gain Theory’s Sensor
  • Media consolidation and how it plays to advertiser’s ability to create contextual audience targeting

Legislation and business decisions have limited data capture, media delivery, and measurement

Over the past few years, a combination of legislation (GDPR, CCPA) and business decisions by the major internet browsers (Safari, Firefox, Chrome) have made data capture on the open internet and via Apps more restrictive.  Legislation broadened (or formalized) the definition of PII (Personally Identifiable Information) for the digital age.  This broadened definition created a potential liability issue for non-compliant business (e.g. everything needs an “opt-in” now).  At the same time, the browsers and web operating systems have begun rolling out privacy features for several reasons: to protect consumers, to comply with legislation, and to avoid liability.  Beyond the basic implication that these events limit data capture, they also by proxy impact how media is delivered and measured.

Where we are right now?

Two major events are unfolding currently:

1. Google will soon be blocking third-party cookies/tags “by default”, joining Safari and Firefox, who are already doing so. 

Chrome represents a clear majority of U.S. web browser installs.  When Google activates this change, the impact will be huge.  Between the three browsers, 90+% of browser-based internet activity will have third-party tags blocked by default.  For definition, “by default” means the user of the browser would proactively have to opt-in for third-party tracking…but why would anyone do that? Once Google releases the Chrome update, consumers will begin updating their software and buying new devices.  As consumers start doing this, the % of third-party tags blocked by default will grow until it’s virtually the full Chrome user base.

2. The release of iOS14 Apple will require users to opt-in for IDFA (The Identifier for Advertiser) cross-app tracking, on an app by app basis. 

IDFA is Apple’s “Identifier for Advertisers”.  This will limit an app’s ability to see a user go to another app on the same device.  Cross-app tracking is similar in nature to third-party tagging (as described below) but one could argue that it’s easier for an app to make the case for how cross-app tracking can be a benefit to the consumer.  For instance, American Express could argue that allowing their app to be able to talk to the apps of their business partners could be a benefit to their cardmembers.  Since this iOS opt-in is at the app level, the consumer will be able to choose if they want to allow American Express cross-app tracking. As iOS14 is released, consumers and app owners will begin to face this choice very quickly.  And as consumers update their iOS and buy new devices with it pre-installed, it will become standard across the iOS ecosystem.

Why are third-party tags and cross-app tracking important?

Firstly, let’s look at the difference between “first-party” tags and “third-party” tags:

  • First-party tags: When a consumer is on Amazon.com, Amazon can track that visitor via first-party tags.  “First-party” in essence means the company doing the tracking owns the website the consumer is on.  This is akin to a shopper being in your brick and mortar store; they’re on your property therefore you have the right to observe where they go. 
  • Third-party tags: this is when the company who owns the “tag” is not the owner of the website the “tag” is placed on.  This allows the third-party to track consumers on sites they don’t own.  In the real world, this is akin to companies following people around all day and logging everywhere the people go and when.  This would obviously be a breach of privacy.  But on the internet, this is exactly what third-party tags had enabled companies to do unrestricted up until now. 

So, what were the implications? 

AdTech was basically built around third-party data capture.  Picture this: if a company places third-party tags across several hundred websites and paid media ads, capturing all the data of consumers hitting those media properties, that company could create a fairly accurate representation of internet behavior.  They could then mine their repository of cookies, all their associated behaviors, and leverage the data to deliver targeted media to their pool of cookies. 

Companies have been doing this for well over a decade.  And while cross-app tracking isn’t exactly the same as third-party tags it enables many of the same things: broad data capture, retargeting, audience/creative optimization, etc. 

If companies can’t use third-party tags and cross-app tracking, is AdTech dead?

No, AdTech isn’t dead. But it has to change.  There is a new approach to data capture that has come along to replace third-party tags and cross-app tracking: server-to-server integrations (S2S).  A server-to-server integration is when ABC.com matches its users to XYZ.com’s users so they can see the overlap in users on their sites.  Doing this is a bit more restrictive than what’s possible with third-party tags; ABC and XYZ need to have shared data elements they can match on.  They would typically match on an encrypted email address.  Net/net, they can only match consumers that have registered on their sites; they can’t match general visitors.  As a result:

  • Shareable audience pool size shrinks
  • Major implications on retargeting and, more generally, at any time digital audiences are created
  • Implementation bottleneck: for instance, Facebook can’t just do a server-to-server integration with every major media property and/or client overnight.  Right now, they have a waitlist that could be upwards of a year

Measurement implications: Is Multi-Touch Attribution dead?

No. However, everything above implies severe limitations as these changes take effect.  The challenge is that it will become increasingly difficult to create a “complete” user path of media touchpoints that led to a conversion.  There were always challenges at play, but they will only become worse.  Server-to-server integrations will become the major way raw media data gets stitched together across platforms by players such as LiveRamp, but they, too, face challenges due to the blocking of third-party tags (which LiveRamp uses heavily). Since server-to-server integrations only work for registered site visitors, not general visitors, the audience scale and matching accuracy of partners such as LiveRamp will suffer. However, with data integrators such as LiveRamp it is nearly impossible to actually QC their work; it’s all anonymized.  As a result, the true gaps created by the loss of third-party tagging and the switch to server-to-server integrations may never be known; this could sew distrust within the Multi-Touch Attribution (MTA) ecosystem, for example, raising questions like, “how accurate is this?”  Measurement partners will still deliver MTA solutions to clients but will the accuracy decline?  The marketplace doesn’t quite know yet.  If it does suffer, we can expect to see clients begin to reevaluate their measurement needs and become open to other near-term measurement solutions like Gain Theory’s Sensor

Media consolidation is at play as well

While this privacy battle has unfolded there has been significant consolidation in media.  We’ve seen the rise of major video platforms like Hulu, YouTube, etc.  These platforms have an endless supply of video content and consumers are typically logged in when they use them.  These two factors enable a massive amount of contextual understanding of the consumer. Given the scale of these platforms, and their ability to target, we’re seeing a massive shift towards contextually targeted media.  These platforms simply know so much about their massive audiences that they don’t need digital behavior data from outside their ecosystems to optimize media delivery effectively. 

Long story short, the large streaming video partners aren’t being hamstrung very much by the blocking of third-party tags and cross-app tracking.  And consumers’ attention and media consumption will continue to centralize around these platforms.  This means that with a handful of large-scale server-to-server integrations across these platforms, plus Google, Facebook, etc, we will be able to see a very comprehensive view of users’ digital media consumption. In the future, there will simply be fewer “critical” media partners, and all of them equipped with logged-in users.  This will make it easier to integrate a handful of major players than it was to track hundreds of third-party tags on mid-tier media partners.  Eventually, new measurement solutions or measurement workspaces like “white rooms” could become much more prevalent and holistic.

As originally published in The Drum, 13th May 2020

Although it’s a precarious time to make sweeping business changes, Gain Theory’s Russell Nuzzo makes the case for Near-Term Measurement in lieu of marketers’ favourite – Multi-Touch Attribution. 

Every organization is asking the same question: if the world economy is frozen, what can I cut that will still allow me to come out of the running blocks fast? Unfortunately for many marketers, the decisions around investment cuts are being made by finance, with only the savviest of chief marketing officers having a say. 

Right now, marketers need to justify every dollar by using the most accurate and relevant data possible to make fast decisions that will have a positive impact. In this climate, businesses that can, need to generate revenue now. 

If return on investment was important for advertisers before the Covid-19 outbreak, it’s absolutely mission-critical now. And for marketing measurement, if anything good can come of the current crisis, it’s that brands begin a renewed focus on mastering what we call ‘near-term performance measurement.’ This approach focuses on using immediate and precise data signals to make advertising decisions. 

Simply put – this crisis requires a strong, data-oriented approach for how we gauge the effectiveness of advertising. It’s time to zero-in on the data that is actually going to help you impact and protect your business immediately, rather than seek solutions through overly complex approaches with highly imperfect data. 

Understandably, the temptation will be to plug the near-term ‘water leak’ by using the multi-touch attribution (MTA) ‘hammer’ that sits in the marketing measurement toolbox. But the ‘hammer’ is not the only tool you could use. 

Historically, brands have tried to be as strategic as possible with their marketing spending using established approaches such as marketing mix modeling (MMM). These models help big brands figure out how best to allocate their budgets to maximize key measures such as reach and sales and evaluate tradeoffs. 

However, if we recognize that the next six-12 months will look nothing like the past two to three years, the old models may no longer apply. MMM should not be thrown out; it can provide a great framing exercise. But it needs to be coupled with a strong, near-term measurement solution that can read and react based on what’s happening right now. 

And if brands don’t get ‘right now’ right – their future isn’t necessarily guaranteed. 

MTA is not the Swiss army knife that marketers need 

Not long ago, MTA was once the darling of the marketing industry. Conceptually, it’s understandable. Who wouldn’t want to figure out exactly which ads drove which action? Even better, it promised marketers the ability to discern how different ad units across multiple media channels worked together to influence consumers and drive action. That’s essentially marketing’s holy grail: reach exactly who you want and don’t waste any money. 

On paper, this is exactly the kind of tool that would be valuable right now, as chief marketing officers grapple with maddening business questions. But MTA has consistently overpromised and underdelivered. 

A slew of less than satisfied customers can attest to this. I have heard of projects that take more than a year to get off the ground, significant data gaps and difficulties distilling tangible insights, not to mention the challenges of integrating MTA outputs back into clients’ adtech stacks. The continued deterioration of third-party cookies and IDs – the linchpin of MTA – will culminate in 2022 when Google, having followed Safari and Firefox, will begin blocking third-party cookies in Chrome. 

Now is the perfect time to re-evaluate the right tools for near-term measurement. 

The right tool for the right job 

While the climate will be challenging for the foreseeable future, creating a unified, near-term measurement solution coupled with robust scenario planning, fueled by the latest available data will be essential. 

There can be no garbage in and no garbage out. Not now. 

To fuel near-term measurement, a tried and tested data-led alternative is to lean into what we call ‘micro-geographies’ for targeting as opposed to cookies. It is much easier to determine where an ad impression was served than who it was served to. Plus, micro-geographies are far more stable from data validation and privacy standpoints. 

Additionally, most external impacts to consumer buying decisions happen geographically; local pricing, local distribution, local weather, local competition – even local Covid-19 infection patterns can affect the decision to purchase something. 

We have been able to build these local features into our near-term measurement solution sensor; a micro-geography-based alternative to MTA. 

For Covid-19 specifically, we’ve been able to build a set of indicators to quantify how the outbreak is positively and negatively impacting brand sales. Some of those trends are short term but others may have longer-term impacts on brand preferences and consumer habits. Some of the impacts will fundamentally change business models permanently. 

Pivot to near-term measurement 

Organizations need an understanding of short-, medium- and longer-term impacts of all their current investment decisions, not just marketing. Alongside longer-term scenario planning using simulation techniques such as agent-based modelling as well as established techniques such as MMM, marketers also need an immediate view on the here and now. 

We think near-term measurement is the ultimate replacement for user-level MTA, and the perfect tool in the box for marketers in the Covid-19 ‘new normal’ landscape. Using near-term data can also predict consumer patterns before they happen. 

Now more than ever, marketers need to make data-informed confident decisions at speed. 

Using the right tools for the job, rather than reaching for a hammer, will help articulate the impact of investments ensuring brands remain relevant and robust through turbulent times, and that much better situated to lead when things get back ‘normal.’ 

Right now, marketers are beset by reduced budgets, lack of actionable data about customers, and rapid demand & media habit changes. Data, assumptions, models, and plans older than 45 days are redundant. Now is the time to prepare for the ‘known unknowns’ and pivot rapidly.

We are living through a generational spike in uncertainty, volatility and complexity.  No one knows the full human toll and economic costs of SARS-CoV-2 and the disease it causes, COVID-19.  This demand collapse is deep, sudden, and touches so many industries a global recession seems inevitable.  Consumers are using products and services they haven’t previously, and necessity is driving them to new paths to purchase.  According to the science of pandemics applied to current data, infection rates will keep going up so assume significant impacts on the concerns, moods, and the economics of your customer base.  They will engage your brand differently, if at all. The ‘old’ norms around brand preferences need stats and price sensitivity requires a new lens.

We recommend you transition rapidly and thoughtfully into the ‘new normal’ using early data and existing insights.  Here’s how.

Assess your situation, rapidly and accurately

At the risk of being dramatic, this is a survival situation for many businesses. One of the first things to do in a survival challenge and preparing for the ‘known unknowns’ is to deepen your situational awareness.

  • Look for analogous situations that inspire your response options. For example, case studies from brand behaviors in the last recession or in the aftermath of 9/11.  An article in Harvard Business Review from 2008 argues for spending in recessions: “It is well documented that brands that increase advertising during a recession, when competitors are cutting back, can improve market share and return on investment at a lower cost than during good economic times’. Also, consider cases of ‘forced trials’ i.e. customers trying competitor brands out of need: will they bounce back to you if they have a positive brand experience elsewhere?

  • Consider digital business model acceleration, product, partnerships and distribution options: Go beyond communications and start asking yourself: how fast can you ramp up your e-commerce capabilities? What product/service modifications can you pivot to? Can you partner with ‘complimentary’ brands to literally deliver your value proposition differently and address current customer needs?

  • Revisit the basic heuristics of price sensitivity and segmentation.  Heighten your sensitivity to both. How much of your business was driven by consumers/segments that are quite suddenly trapped at home?  Unemployed?  How intense was price/value competition prior to the pandemic?  Does your current marketing spend keep you appropriately messaged and top of mind to prevent leakage to competitors for key consumers?  Broadly, maintaining investment in brand to support the value proposition can mitigate price sensitivity.

  • Assess whether you are using the right marketing channels to communicate.  Current media data is clear on short-term consumption shifts, your channel mix also needs to shift.  Millions of your customers are at home driving the 30%+ spike in television viewing globally.  Viewing has shifted to general entertainment, given the lack of sports programming.  Consumers are reporting much more streaming since the outbreak, films and TV shows are up 60%+.  

Act now, triangulating with your best judgement

Most businesses are acutely focused on cash flow.  Demand plummeted faster than many cost structures could withstand.  In this harsh new reality, how much of your marketing and media investments are being eyed by your CFO?    

If you can maintain marketing investment  we recommend:   

  • Ensuring that your in-flight messaging is appropriate.  Dig deep into every crevice of your messaging and review, review, review.  Is your ‘old’ message appropriate now? What should we be communicating? What impact will your messaging have on brand perception?

  • Optimizing the channels you use.  Not all marketing spend can pivot on a dime, but media consumption habits have been forced to shift.  For example, while TV dollars are likely pre-committed, Digital, Paid Social, Paid Search can be adjusted quickly.  This logic needs to be applied to your entire mix.  Before you shift or reduce however, you must consider if you can serve short-term demand spikes, like in-home delivery, or if you are effectively at a stand-still, example you are an airline.

  • Using the channels with demonstrated prior strength until the fog lifts.  Perfect is not the key objective right now, presence is.  Monitor closely any potential drops in CPM across your marketing mix because some of your competitors might be forced to pull back.  Marketers willing to spend more with their partners are likely to benefit relative to those who cut, just as they always are.  Although it seems a long way off, H2 spending also needs to be a consideration in your revised plans.  

Look forward to re-igniting demand.

 “This too shall pass.”  said Abraham Lincoln.

COVID-19 has disrupted demand as we knew it and will re-write business models.  It may change forever how we market to, engage with and deliver our value proposition to the customers we have and the customers we want.  Events like these have real impacts on consumer confidence and therefore, habits. To get ready for the coming new normal, you must immediately think about the long-term effects of your actions today on the brand. In fact, 93% of consumers believe that brands should “stand up and help.” If your firm can do something to help the current crisis like donations, make those socially important actions and values part of your narrative.  Consumers have a long memory of these things and almost all of them have more time on their hands to watch what you are doing. 

By adjusting and adapting now, you will be ready to capture demand when it comes back. Let’s hope it comes roaring back soon.

© Gain Theory 2022. All rights reserved.