Analytic Edge’s Garth Viegas on the new wave of measurement and the Marketing Mix Modeling

As a part of the WARC X Amazon report, Garth Viegas, Marketing Director, North America, at Analytic Edge, speaks about the biggest shifts in the market and how strategic market mix modeling is key to understanding what drives the business.

Could you tell us a little bit about your role and its relationship to measurement?

I manage the Analytic Edge business in the Americas. Analytic Edge focuses on marketing analytics, especially in measuring marketing attribution. We work with six of the top 100 consumer-facing companies in the world, and we've got offices in about ten markets.

Over the past 20 years, I have been passionately involved in analytics with multiple blue-chip companies worldwide. Throughout the journey, I noticed that successful companies do two things well: 1. Understand what the consumer wants. 2. Understand how money is made. I believe that measurement is the red thread that links these two capabilities.

We’ve seen some big shifts taking place in the market – the pandemic accelerating omnichannel shopping to media landscape fragmentation, and from the arrival of ATT to the death of the cookie. Which shifts are having the biggest impact right now?

That's right – all these things are challenging our notions of marketing. From a measurement perspective, I see three significant changes happening right now.

The first one is that we need new measurement frameworks as the existing ones don't reflect the new world that we live in. For example, the resurgence of MMM with the loss of signal due to cookie depreciation.

The second one is speed to action, which is quite interesting. Marketers would rather be 80% correct and have the answers today than wait for three months and be 100% correct because the world is moving so fast.

The third significant change is rapid experimentation, and it will shape the way we think and measure marketing. We've used experimentation such designs as A/B tests. There is a direct relationship – the closer the match (the difference between A and B is very small), the smaller the marketing investment you can measure. Today, new technology creates synthetic controls that enable you to create a perfect control scenario. Consequently, you can measure tiny consumer changes with a high degree of accuracy.

But within all these shifts, how can marketers evaluate what is effective?

As a marketing community, we've gone through this journey in three phases. In phase one, we got good at understanding what's working by channel. For example, if a company invests X dollars in TV, we can establish the ROI accurately.

Then, the marketing community evolved to a second stage, where we could measure the impact of multiple media working concurrently. For example, how do TV and social media work together and improve the campaign's effectiveness.

We're now on the cusp of the third wave, which is the people component. We've done a lot of work for quick service restaurants (QSR), and interestingly, we found that one of the key drivers in our model was the effectiveness of the restaurant manager. Everything else being equal [among like establishments], the store manager's performance was vital in driving sales. Now, this is intuitive, but we are measuring this factor accurately. Companies need to quantify all consumer touchpoints. This is the challenge for the business in the future.

You mentioned marketers are turning to Marketing Mix Modeling – why are we seeing that rise in relevance?

We must take a step back and understand what the role of a marketer is to put Marketing Mix Modeling into perspective. There are multiple definitions of a marketer, but there's one I'm slightly partial to that can be attributed to James Culliton. Culliton was an American academician and professor at Harvard, who in 1948, first came up with the idea of mixing ingredients to drive sales – ingredients such as media, price, distribution, etc.

This fundamental definition hasn't changed over the years – marketers have many more ingredients due to channel and media fragmentation. While the number of ingredients has increased, you must still get the mix right, which is challenging. As we move into a privacy-first world, this becomes even more challenging!

Market mix modeling (MMM) is resurging as it can address the challenges mentioned above. I think marketing mix has gained relevance today because it helps marketers understand the marketing drivers in a privacy-first world.

Should marketers add MMM to their toolkit? Who within an organization can benefit from those insights?

Yes, marketing should add MMM to their toolkit. I believe that the entire organization could benefit from market mix modeling as it helps the organization understand what drives the business – external factors such as weather, COVID, competition, and internal factors such as media, pricing, etc.

To ensure that MMMs are well embedded and leveraged, we need to reframe the MMM. The first M in a MMM is marketing, and hence, consequently, people think MMM is related to the marketing department exclusively. In actuality, the MMM helps the business understand the impact of marketing, sales, and external factors (e.g. economic conditions, competitive activity), that come together to drive the business. The MMM framework enables a dialogue among departments. We have seen clients replace the M (marketing) with a C (Commercial) to rename the MMM as a CMM to ensure widespread usage within an organization.

How can marketers ensure they are getting it right? What are the success factors that go into implementing and using MMM?

I believe there are three key factors to implementing and using a MMM.

Firstly, set up a cross-functional with executive sponsorship to ensure organizational buy-in.

Secondly, update the MMM regularly – for example, from annual to quarterly, and by increasing cadence, you get better usage and insights.

Thirdly, organizations should look beyond the data to incorporate tacit knowledge into the model. Bayesian priors enable organizations to capture tacit knowledge.

What’s the primary challenge marketers face today in the complex measurement landscape, and how can marketers move forward?

Marketers need to operate in a privacy-first world, both in terms of execution and measurement.

We've known this, it's been coming for some time, and we also need to learn how to operate in this privacy-first world quickly. It changes the KPIs, it might even change the analytical framework.

At the same time, we are in the age of information obesity, where we get tons of data daily. Data availability will only increase with media and digital fragmentation. The ability to separate the wheat from the chaff will be critical for marketers. In our experience, 20% of data gives you 80% of the answers are a question of realizing what 20% of the data contains and ensuring that those KPIs link to your ultimate business drivers and sales.

An MMM addresses the two issues by enabling the business to understand the key drivers. Those within your control (e.g., media) and those outside your control (e.g., weather.).