The future of marketing measurement: How to win in a world of data with Mikael Thuneberg
In this episode of the Marketing Analytics Show, Mikael Thuneberg, CEO of Supermetrics, will discuss the challenges of marketing measurement, and more importantly, what you can do to overcome them.
Here's what you'll learn
Subscribe to the Marketing Analytics Show newsletter
Be the first to know when a new episode drops.
Subscribe to the Marketing Analytics Show podcast
Learn everything you need to know about marketing analytics.
Today, we’re going to talk about the future of marketing measurement and how to win in the world of data. And we have a very special guest, Mikael Thuneberg, founder and CEO of Supermetrics. Mikael, welcome to the show.
Thanks, Anna. Great to be here.
It’s awesome to have you here. And you’ve been working in marketing for quite a while now, and you’ve mentioned that, since 2012, marketing has gotten a lot more challenging. Can you please explain why?
Sure. So, first of all, as I’ve mentioned, the rise in competition. So pretty much everyone is doing online marketing nowadays. And the average consumer sees thousands of ads online per day. So, it’s very difficult to stand out, much more so than a decade ago. This also means that prices are increasing all the time. You need to bid more; the CPMs are going up. So, it’s more difficult actually to make a good result out of your marketing.
On top of that, it’s much more difficult to actually track whether you’re making good results because you used to be able to track the full journey from the marketing activities into the conversions. And that often is no longer the case due to the decline in third-party cookies and iOS 14 and so on. So you really need to invest in a proper data infrastructure to see how your marketing spend produces results.
I could also mention the explosion of data. So, there are so many different ad networks and marketing technology tools that are on the market right now. I’m sure you are all familiar with the Martech 5,000 infographic, where you have all of these tools listed, and that infographic is just growing all the time. So there are thousands and thousands of different tools there. And even if you’re only using a fraction of this, it still means that you have a lot of different tools in use, and all your marketing data is fragmented in these different silos. And it’s very difficult to get the big picture of all of how your marketing is working.
That sounds very, very interesting. Now let’s go deeper into this answer. So, you’ve mentioned that there are several challenges, but how can marketers, analysts, and data teams overcome them?
Sure. So, first of all, you need to invest in your data infrastructure. So, I think the foundation for that would be to have a CRM in place to have your customer data collected there. Of course, you need to invest in quality customer data so it’s not full of garbage.
And then, when you have that in place, the next step would be to have a central repository for all of your marketing data. So you would bring the CRM data, the customer data, and from there, you would bring the marketing data from all of these marketing technology tools and ad networks, plus your web analytics data and other data sources that you may have, together into one place, where you can really get the big picture of your marketing activities.
And that can initially be a spreadsheet, Google sheet, or an Excel file, where you collect this data. And many companies are doing it like that. And that may be enough for you if you are a smaller company and you don’t have huge data volumes. But if you do more data, you may want to think about moving into a data warehouse where you collect that data. So Google Big Query, or Snowflake, for instance, can be a good data warehouse to operate.
Yeah. So you mention that marketers should try to use data warehouses or move into data warehousing. Can you please talk more about the marketing data stack evolution? So what can the teams do to future proof their marketing analytics data stack?
Sure. So, a lot of marketers are using spreadsheets for their data, and that’s fine. And I think spreadsheets have their use, but we would also suggest that you think about data warehousing. Maybe you can still keep the spreadsheet, but you add the data warehouse also there. And there, in data warehousing, we see three different tiers that I could go through here.
So tier one would be that you have the Google Big Query or another data warehouse in place. And then, you start having the data flow into that tool. You can use a tool like Supermetrics to get your marketing data into that data warehouse. And then, you have a reporting tool on top of the data warehouse. So you can have Google Data Studio or Looker, or Power BI, or another reporting tool that you use for reporting and analyzing that data.
So in tier two, the key distinction between tier one and tier two is having a separation between the raw data and the reporting data in the data warehouse. So, if you are doing reports directly on top of the raw data, you’re missing out on a lot of opportunities there. So we recommend that you have a separation between these two. And that means you can do cleanups on the data. You can do transformations. You can enrich that data before it’s being fed up to the reporting side. And on tier two, you would typically use something like [inaudible 00:05:19] running every night to process the data from the raw tables into the reporting tables. And that’s a good solution. You get a lot of the advantages of the data out of that tier.
The disadvantage here is that typically it’s on rather shaky ground. So, the pipeline complexity can be quite a lot in this model, if you add a lot of different data sources, and if you need to do a lot of these [inaudible 00:05:51] and processing. And it’s also typically quite dependent on individual people in the company, which is always quite risky.
So what we see the most mature companies moving into or having is tier three. And there, we use specialized tools for all parts of this process. So, first of all, testing. It’s super important. Like what’s the point of all of this great data infrastructure if you can’t trust the end results? If you’re always questioning whether this data is correct or not? You need to have a robust testing framework in place. And you can use a tool like DBD or Data Form to achieve that.
Second, I would talk about orchestration. So these [inaudible 00:06:35] chops, they can work for you, but they also can become quite complex. So, you could move into tools like Airflow or Google Cloud Composer to handle all of this data processing in the data warehouse.
The third point would be version control. Of course, in software development, as we do here at Supermetrics, that’s a critical part of the process. And that’s becoming more so also on the data infrastructure side. There, that allows you to, if you see any problems in the data, possible problems, you can see exactly what’s being changed in the data processing. You can identify possible reasons for these issues in the data. You can see who’s done them and when. And if you need to, you can also roll back these changes easily to get to a previous working state, if needed.
All right. These three tiers sound amazing. Thank you for such an elaborate answer. And now that marketers have, for example, invested in taking it all the way up to tier three, how can they use these tiers? So what can… Results can they achieve using these tiers?
Yeah, there’s a lot of benefits to putting this data stack together. So obviously, you can do analysis. You can analyze the data. You can do reports on the data for your organization. That would be kind of the basic answer, but you can also do much more. So, especially now in this [inaudible 00:08:04] world, where multi-attribution is no longer that effective, marketing mix modeling is making a big comeback. So that means that you can measure the effectiveness of your marketing channels, even if you don’t have full tracking of the user journey. But you can do that on an aggregate level and get a pretty good estimate of how different marketing activities are paying off.
Once you have all of this data in a centralized data warehouse, you have that cleaned up, and good to use, then running marketing mix modeling on top of that is a good next step.
I could also mention feeding this data back into the marketing platforms. That’s something that’s still quite rare. So, I would say it’s just the top 1% of marketers who are doing that. But that gives a lot of benefits. So you can actually directly optimize your activities on the marketing side based on the data that you have collected in the data warehouse. And you can gain a lot of benefits from data. So you can then optimize all of your marketing activities for the individual consumer groups.
This was a very interesting chat, Mikael. Thank you so much for coming to the show.
Thank you, Anna.
Try Supermetrics for free
Get full access to Supermetrics with a 14-day free trial.
No credit card required.