Turning MMM data into insights with Juha Nuutinen

Key takeaways

MMM is more accessible than ever

What once took months can now be implemented in days. With improved automation and integrations, marketers can quickly access insights without heavy lifting.

Privacy-driven shift toward MMM

As third-party cookies fade and digital tracking becomes more challenging, MMM provides a future-proof method for measuring marketing effectiveness across all channels.

MMM is not just about data, but actionable insights

Having data is not enough—true value comes from interpreting it effectively. MMM helps marketers allocate budgets wisely, identify hidden opportunities, and scale winning strategies.

Combining MMM with incrementality testing yields more accurate results

Experimentation is key. By using lift studies and geo-testing to calibrate MMM models, businesses can refine their understanding of each channel’s true impact.

Balancing short-term and long-term goals is crucial

Over-prioritizing short-term ROI can weaken brand equity. Successful brands use MMM to strike the right balance between performance-driven marketing and brand-building efforts.

Scenario planning is a must for marketing leaders

Instead of reacting to changes, marketers should proactively test different budget scenarios. MMM-driven optimization enables smarter decisions on where to invest and scale.

Get started, even If it’s small

Waiting for a perfect setup is a mistake. Even a basic MMM implementation with digital channels can provide game-changing insights that improve marketing efficiency.

Final thought: As Juha puts it, “Don’t try to eat the elephant in one bite.” Marketers should start small, refine their models, and continuously iterate to improve accuracy and effectiveness. If you're looking for a way to gain control over your marketing budget and prove ROI, now is the time to embrace MMM.

Bonus: If you’re a Supermetrics user, check out the free pilot program with Sellforte to start your MMM journey. Visit Sellforte’s landing page to learn more.

Related resources:

• Listen to Olivia Korey’s episode on Decoding Incrementality Testing to understand how geo-testing and experimentation enhance MMM accuracy.
• Watch Outi’s SuperSummit session on Modern marketing measurement with Supermetrics and Google Cloud for additional insight
• Read our latest blog on MMM data: Building a solid data foundation for your marketing mix model
• Read our latest blog on Custom Data Import: Get data from any sources to get a holistic view of your marketing performance

Transcaript

Outi Karppanen (00:00):

Hi, and welcome to another episode of the Marketing Intelligence Show by Supermetrics. I am your host today, Outi Karppanen, and I'm the lead marketing industry strategist at Supermetrics. And with me today is Juha Nuutinen, who's the CEO of cell, and we will be talking a lot about MMM and going beyond the basics of actually how to turn that insight into action.

(00:25):

So welcome Juha, please introduce yourself and Ste,

Juha Nuutinen (00:29):

Thank you so much. So my name is Juha Nuuteinen, CEO, and one of the co-founders of Sellforte Startup. Or I would like to think scale up focusing on a marketing mix modeling. So providing marketing mix modeling as a SaaS and what makes us different compared to anybody else is we care more about our customers than anybody else.

Outi Karppanen (00:55):

That is perfect. And you were founded 2017?

Juha Nuutinen (01:00):

That is correct, yeah,

Outi Karppanen (01:01):

I would say you're beyond startup now.

Juha Nuutinen (01:03):

Okay, okay. Thank you. In my box at least.

Outi Karppanen (01:06):

Great. So as we said, we're going to be talking about MMM, and I've had the pleasure of working with you guys when I was working at Dentsu as a media strategist, so I think we can both bring interesting viewpoints to the conversation. Very nice. So let's set the stage. MMM has been around for a while and it's been gaining more interest now and it has evolved to some extent.

Juha Nuutinen (01:31):

Yes.

Outi Karppanen (01:31):

How do you think it has evolved and why is it more important now than ever?

Juha Nuutinen (01:37):

So where do I compare? I guess I could compare to 2017 the company was founded. Okay. Back then the world, back then, we actually come to a customer's office and we bring a USB stick.

(01:49):

We

(01:50):

Actually bring a USB stick and in this USB stick we basically, it's encrypted and all so very, very safe, but all of the customer sales data is then transferred through this USB stick to Sellforte the laptops where we are locally doing kind of the analytics. So this is how it all started, but that was, I guess, hey McKay was only starting Supermetrics back then, so we didn't have the great data pipeline tools that we have today. So today it's completely different. So all of the customers, we can basically plug in all of the digital data sources in, they take half an hour meeting with the customer. We typically spend 15 minutes, customer is clicking just the data sources, connecting their data to software. So super easy, we can do it. We have a very nice cooperation also with Supermetrics to make it very, very easy for all of the customers to connect their data to MMM. And then there's the sales data that's now living instead of the USB stick and taking export from SAP. So that's now also living in customers BigQuery, AWSm S3, it's Azure, Snowflake. You can easily integrate to all of these data sources and automate these daily updates. So life is very easy in terms of data integrations nowadays.

Outi Karppanen (03:25):

So you would say it has been solved of getting the data and no more Excels manual?

Juha Nuutinen (03:30):

Yeah.

Outi Karppanen (03:31):

Yes,

Juha Nuutinen (03:31):

Exactly. No more Excel. Well still sometimes the most comfortable way for the customer in collecting and sharing their offline media investment data is building a shared Google sheet that is automatedly read every time the model is run.

Speaker 3 (03:49):

So

Juha Nuutinen (03:50):

Sometimes that is still used, but let's just say that that's also completely machine readable and very easy to maintain for all the parties.

Outi Karppanen (03:59):

What else would you say? Has there been any? I think it's just overall the support and it's just easier.

Juha Nuutinen (04:06):

Yeah, yeah. I think when we started there was no reason absolutely whatsoever for an e-commerce company to do MMM.

(04:16):

They were thinking that hey, we are already capturing, we have perfect information from all of our customers. We are reading all of the touch points. When the customer sees the ad, clicks the ad, and we do this great multi touch attribution model, everything is traceable. But I think today, okay, so back then it actually wasn't traceable. So that's one thing. So the things haven't changed. Meta ads, Facebook ads has never looked great on your GA four report or GA report, UA report. That has always been the store. So it wasn't traceable actually then. But now because of the privacy issues and you are actually not getting 100% of your sales traced in your web analytics tool GA four, you are actually getting 75% 70. So in some markets, 60% of your sales are traced in your GA4. So our customers are like, Hey, I don't want to understand how my marketing works like 60% to 60% of my sales. I want to understand how it works to 100% of our sales. And many also have their marketplace sales in Amazon, they have their own retail stores, so they want to understand how marketing affects also these channels where they don't get the attribution.

(05:40):

So MMM is becoming a mainstream,

Outi Karppanen (05:42):

The market has, the marketing or media landscape has become very, very fragmented from the age of only having TV and maybe print. And also I would say that we have gone from the ages of digital coming up and gaining attraction and then everyone would be like, oh, the amount of data and this is the truth that we have to realizing actually it hasn't probably been at any point a hundred percent accurate. And also with the data privacy, now those who are used to e-commerce as being, we would call them data rich, they're also now becoming more data poor, how I would say like FMCs or retail clients have been all along. So I think everyone's coming to a more plain field of playing ground of

Juha Nuutinen (06:30):

If

Outi Karppanen (06:31):

You want the answers to your marketing.

Juha Nuutinen (06:32):

And also I think a big change in the market is that now the ad platforms are starting to speak about MMM. So we are getting a great support from meta, from Google's YouTube team, from TikTok, even like Pinterest, these ad platforms that are speaking that hey, MMM might be the way to actually get a more accurate read on how our products, our channels are performing for the customers.

Outi Karppanen (06:59):

Exactly. How do you see, because evolved to this point, I think in the industry there's been a talk about calibrating your MMS through incrementality and attribution. How do you see that from, are you guys doing it and what do you think of it?

Juha Nuutinen (07:15):

So this is part of a great topic and if you want to understand more about dos and don'ts about incrementality testing, I just listened to create Supermetrics podcast episode with Olivia Corey.

Outi Karppanen (07:30):

Yes.

Juha Nuutinen (07:30):

Yeah, yeah, yeah, that was a brilliant one. So I would recommend to go there, but also I think when it comes to connections to marketing mix ing and attributes and then brilliant resource is Google's modern measurement playbooks. I think I heard in a pre discussion that you have been also talking about that.

Outi Karppanen (07:50):

Yes, I've read it through, it's a long read for everyone to know. But I would say especially at the beginning, even if you're not technical, I recommend it if you work in marketing or our CMOI highly recommend reading it or checking out our super summit recording where I go over it. But then it gets more technical, it gives really good instructions on how to do it. And I would say that's probably the end goal where people should be going. But how many do you think at this point are actually doing that and calibrating? Is it gaining popularity?

Juha Nuutinen (08:23):

So more and more

Outi Karppanen (08:24):

For

Juha Nuutinen (08:25):

Some less technically oriented customers, they need more like help we point them to resources, how to set up a meta conversion lift test or how to set up a geo lift test or how to do a sat down test in one of your markets. So of course give information to our customers, encourage to use because I think it's worth also discussing the basic principles of what they say in the Google modern measurement notebook.

(08:57):

So I think the key idea is that the marketing mix modeling is holistic methodology to basically capture how are the channels really performing, especially for the biggest channels and for those channels where there is more uncertainty in the marketing mix modeling what the right answer is. If the ROI posterior estimates are very wide, what is the right ROI level, then we encourage to do test then doing this tests nowadays is pretty simple, two to six weeks and platforms are giving great tools, especially meta is giving super easy to use tools for doing these tests. So we encourage and then when you do the test, you can calibrate your marketing mix modeling. So hey, at this point of time, okay, before we use this experiment, we were telling that market mix modeling was finding out that the ROI for meta advantage plus campaigns for example of us four, but then the test is telling you, hey, actually during this test period it was like 3.2 plus minus 0.3.

Outi Karppanen (10:09):

Yeah.

Juha Nuutinen (10:09):

Okay. So then you actually put this 3.2 plus minus 0.3 as the prior to your model and then refit the model. So then you actually might end up with 3.5 as the ROI, and that's better than not doing any calibration at all.

Outi Karppanen (10:28):

And it's probably what do you recommend is how often should they be calibrating the MMM?

Juha Nuutinen (10:34):

Well, here again, I'm actually promoting a lot of these research papers that we are reading very carefully at Sellforte here. I would refer the readers to take out this Harvard Business Review article, a new gold standard for digital measurement. So that's like a brilliant article. The summary is that they are arguing that marketing mix modeling combined with this experimental testing calibration is the new gold standard for your digital a measurement. And what the team there is saying is that, okay, so the bigger your spend is, the more often you should do this incrementally testing. So if you are running like let's say $100,000 per month on a channel, it might be enough. You basically calibrate it once a year.

(11:32):

But if you are actually spending 1 million per month on the channel, it makes sense to do quarterly incrementally the test because actually your channel might be performing completely different way during your peak season, like pre-Christmas compared to January, February, March. So it makes actually sense to do it on a more recurring basis. But of course there's a work involved in doing these tests. So realistically, the more channels you are, the more rarely used then calibrate and incrementally test every channel. But we are definitely thumbs up for this approach. But I would like to also add that marketing mix modeling does not require any test to

(12:18):

Work.

(12:19):

So you actually get a sensible read on what is the ROI of your channels, even if you haven't done any test. Because also the marketing mix modeling how we are doing the model calibration, we are not doing it in isolation. For me it's very important that our model gives accurate recommendation to our customers 100 times out of 100.

(12:44):

So we don't leave it to chance if the model is getting good results and we cannot allow the model to fit to one poor data point. And then you are getting completely wrong recommendations to customers. So what we are using is rich data from we are pulling the ad platform attribution data, so conversions, conversion values reported by the ad platforms with all of the attribution settings you are using with the harmonized attribution settings, we are taking all of the Google analytics for attribution data. We are matching that Google analytics for attribution data with your sales data using this unique key transaction id, we might be using your Shopify analytics reporting. Of course we are using all of your experiments, we are analyzing independently, all of your shutdown tests swings up and down in all of your media channels. So this is the information we are basing the MMM result. And actually when you use fully use all of that information in your marketing mix modeling, this is not coin flip. You are getting the same results repeatedly every single time.

Outi Karppanen (14:00):

So it's very robust.

Juha Nuutinen (14:01):

Very robust.

Outi Karppanen (14:02):

And the more you put in good quality data, the more robust it becomes and it becomes reliable.

Juha Nuutinen (14:09):

Exactly.

Outi Karppanen (14:09):

And I think that's great and it's remembering it's not about just getting one answer once it's constantly improving because you're doing, you're most likely in your marketing campaigns doing different things and as you said, depending on the budget or depending on how often do you change your creatives or change your tactics, I think that then also influences how often you should probably calibrate or do incrementality tests and seasonality. Because if you're in e-commerce, peak season, black Fridays Christmases or whatever happens to be like if you sell ice cream, most likely summer will be your peak season, especially in Finland where we are. So I think that then all you to consider all of that Super good.

Juha Nuutinen (14:54):

I agree.

Outi Karppanen (14:55):

So now we've gotten a little bit on how to go about the data and how MMM has evolved. Let's now pretend we're marketers and we now have that data in our hands, the robust data in our hands. What can marketers actually then do with it Now we have a result, most likely a dashboard on your case or however you have mm MM, what can we do with it? What can they do to make better decisions based on MMM or is that the holy grail? You now have the data go and run?

Juha Nuutinen (15:28):

On a very fundamental level, if the customer does not do any changes in their marketing based on our tool,

(15:38):

Then we have failed because customer we love impact. And if the customer doesn't do any changes in their marketing, there cannot be any impact. The customer is just doing everything that they used to do before. Our goal is to provide accurate data for customers decision making and help them make better decisions and then track and follow what was the effect of those decisions in the real life to their sales and to their margins and to their overall performance. So that's what we do and we have different tools to help customers then make those decisions. I think one of the tools is of course, like you mentioned dashboard, right?

Speaker 3 (16:29):

Yes.

Juha Nuutinen (16:30):

So okay, so you see all of your historical investments, you basically update this dashboard which includes all of your ad platforms, all the metrics, investments, impressions, clicks, conversions, conversion values from all of your ad platforms, Google ads, meta ads, TikTok ads, all of that platforms you are getting your data on daily level. But then you are also getting on daily level your own sales data, which is actually then linked with the modeling to the marketing data. And you basically have what is GA four saying? You basically get the sale from the same dashboard. This information, they are basically getting what does the ad platform say? What does GA four say? What does Shopify analytics say? What does our internal whatever we ask the customer on the counter, where did they come from? So we basically have all the data including the marketing mix modeling in this one dashboard. So I call it Palantir because I'm a lot of the rings fan because it shows you basically everything, kind of all the information in one place you can just, so that's of course one, but I think this is still table stakes because even if you have all of this information, very few marketers are like, okay, let's drill down with the data now, right?

Outi Karppanen (17:50):

And we've done actually research and it's going to come out soon about people feel like they're drowning in the data now and they're looking for solutions or tools to help them gather that inside. Mari, not a lot of CMOs have time to dig in there and dig into the dashboard. I've used your dashboards and common clients we've had and I've had the time and it's been my role to go in deep there and just like, oh, what if I look at that and what if I look at pre peak season or during peak season and what about this media that? And I love the granularity that you can get in there, but I think everyone's now looking for something to help them get more pre chewed

Juha Nuutinen (18:35):

Insights spot on. So realistically, people responsible marketing planning, they don't have the luxury of spending many hours in the tool. So what we want to do is we want to make making decisions easy

(18:50):

For the customer. So that's why we have for example, this campaign dashboard. You can look at the past 30 days, past 60 days, past 14 days, whatever time period, all of your campaigns that are currently live or were done during that period. And you basically get, hey, what is the ROI based on ad platform based on GA4? And if you are using, many of our customers are using the ROI from MMM as the single source of truth as the common currency to compare against different tactics like Google search brand compared to Google search, generic and pmax, but also different ad platforms. So meta advantage plus compared to TikTok awareness compared to Google shopping. So you can do all of these comparisons and marketing mix modeling, ROI is the kind of common currency. So then we basically get recommendations on every campaign level. So hey, this campaign is a winner. The target draw us of that campaign is five use actually kind of invest 20% more on that. So hey, change your target ROAS to 4.5 so that the ad platform can spend more.

Outi Karppanen (20:08):

Yeah, and I think then you guys and whoever partners they may be using our media and agency or who's the CMO, they then can you guys are providing also the insight and support on that? And do you have tools like scenario planners so that then if you want to start future planning, okay,

(20:27):

Because

(20:27):

Now we're getting the where are we now, that's the basic of MMM. The first time you do, you're like, here we are. But then you want to be like, where do we want to go? What's possible if we do a little optimizing and what's probably the optimal media mix?

Juha Nuutinen (20:42):

MMM is also like a strategic tool. So we have been sometimes comparing it flight control tower where you actually see where all of the planes are flying and you can basically get really the big picture and I think it's very important is that, okay, so we have some planning events that we want to support. One important is annual planning. So you need to create the annual annual budget and facing to different seasons, facing to different quarters, how much are we going to invest in different channels TV compared to paid search, compared to paid social performance, and then how to allocate that investment throughout the year. So this is something that our tool provides. So you can basically have multiple different scenarios for your next year and you also I think very much use this, okay, so hey, our budget is like 10 million, but hey, if we get pressure, this is how much sales we lose. If we need to go to 9 million or 8 million next year, and this is the business case for going to 11, so this is how we would spend it and this is where the channels where we would increase it, the seasons, the campaigns where we would invest it, and this is how much incremental sales or profit we would get. So the controlling is always very interested in those scenarios and we are making it for our users, very easy to play with different scenarios.

Outi Karppanen (22:16):

And I think that's great because I think the biggest question marketers are getting asked, especially by CEOs or CFOs is, okay, we're putting this huge amount of money into marketing, what are we getting in return? And that's basically what MMM gives answers. And then if you can do those scenarios and say, well actually if we would put 1 million more, we could get 3 million gen who's going to say no to 3 million extra sales? No one. So it could be that that is an easy way to justify why invest more.

Juha Nuutinen (22:49):

Yes, yes. And maybe one more thing I could add, our customers are also using this scenario or we call it optimizer or media optimizer, so they're using it also for planning next month or next quarters media plan.

Outi Karppanen (23:11):

And I think that's the good thing about MMM as you can look at it as a strategic tool and look at it on a yearly scale and look at the, as you said between different medias and that's exactly what I used to do with it. Or then you can go deeper if you're a digital planner or if you're focused on certain campaigns or you are in charge of tactical campaigns and someone else's more on brand campaigns, you can focus on your KPIs and your different measurements. So I think that gave an overall, I think just to summarize, MMM gives you understanding of where you are at this moment and where you're going to be going and with what media mix. Moving on to, I think these are the ideal things of what we could do. And I think solving the problem with too much data and giving insight already is great for all marketers, but what do you think, what are the most common mistakes marketers make when they interpret the data or what are they the common pitfalls or what aren't they considering when they're looking at the data?

Juha Nuutinen (24:10):

I think many things have already went really well if we are at this point that the customer is looking at the data, because I would say I want to answer maybe a bit different because I think the biggest mistake is that you actually never get started or you get started and you never finish this test or pilot that you wanted to do. So I think that's the biggest mistake. I have so many, I hear from so many places that okay, if you have just one ad platform and you are investing, I don't know, less than 1 million per month on paid media using into mm m, and I'm like, that's not right. I violently disagree with that statement. So I think definitely if you are actually investing even more than if you're investing 50,000 US dollars per month or more, it actually makes total total business sense to invest test with MMM. And even if you are only doing Google ads or even if you are only doing meta ads, I would say it makes total sense. It actually makes super much sense to do MMM because MMM also one of the kind of things that we also want to provide to our customer, but it's the business case then of adding meta.

(25:39):

If you're only doing Google ads at the moment

(25:42):

Or like, Hey, actually in Google ads you might actually have 10 different tactics. You might be doing p max, you might be doing Google search, generic Google search brand, you have demands and campaigns, you have your YouTube, you have your display advertising campaigns. So you have so many tactics within Google ads. You have so many countries which are in different phase of growth, completely different incremental factors. You might be very mature already one market and the second market is growing like race. You want to actually get a tool to manage and optimize all of these palettes. So it makes total sense. So I couldn't disagree more when people say that it doesn't make sense for most advertisers to do it.

Outi Karppanen (26:33):

And I totally agree with that because the reality is, and I think one pitfall I've seen in history is not then doing variety in your data or you just think like, oh, this is the one truth I now have and I'll just continue optimizing, optimizing, optimizing, but forgetting to add on every now and then a new media into the mix or testing this, and even though you tested it, let's say two years ago in another market, does not mean that it wouldn't work now on this market or even that market. So it's just continuously bringing up new tests to it, seeing what the data is giving you answers. So I would say that's one other that I've seen from my aspect is focusing only on short-term ROI, so only wanting, I want to maximize my short-term return on

(27:27):

Investment

(27:28):

And forgetting the brands.

Juha Nuutinen (27:30):

Yes, I can relate to that. I can relate to that. Yeah,

Outi Karppanen (27:34):

I've seen an example of where I, not naming any names of what companies, but they were investing more constantly on, it was looking great, it was giving ROI and immediate sales, but then when they had done MMM and they had seen in the past three years of data, they could see that the baseline sales was decreasing and the effect of marketing saw how much marketing is affecting your sales and bringing more sales was increasing. So total sales was pretty much the same, but I could take from that data is that their brand was slowly dying, like the brand value of it. So meaning they were focusing too much on the short term. And I come from a background of doing a lot of brand marketers, so retailers, FMCG. So I'm used to preaching about the role of brand and how that helps in the short term

(28:27):

ROI.

Juha Nuutinen (28:27):

Yes. And marketing mix modeling is also great tool for analyzing these long term effects. Customers who have long consideration period or are not constantly at the market like, okay, I have a mobile phone and I don't need a new iPhone at the moment, but when I'm ing the phone I'm very suspective, we'll do marketing. So in these cases, majority of the marketing impact can come from the long term. So what is happening beyond the first two weeks after the advertising? Of course none of the attributes and solutions have a good answer on that, but in marketing mix modeling, you can basically add this second component to your media features that is the long-term part of the media with a retention rate of 0.99 and uplift, type of uplift, small duplicative shape and all of this that our data scientist can tell you much more about than me. But yeah, you definitely get that out from the MMM as well,

Outi Karppanen (29:32):

And I think that is one of the misconceptions of MMM, that it is very short term, but in reality nowadays, especially I think that's the evolvement of it as well, is you can look at and maybe brands effects and well, I'm not going to say creativity is the limit, but in a way creativity is the limit of what you can come up with on measure as long as you have the data for

Juha Nuutinen (29:54):

It. Yeah, yeah. We are on campaign level and we are on ad group level, but we haven't yet gone to the kind of creative or app level in our solution as well. So you could say that creativity is the level, but we are slowly approaching

Outi Karppanen (30:09):

And that's a whole lot of conversation of what creativity has in it. I think one other, a lot of misconception, but pitfall is maybe not understanding the underlying conditions of each media, even though data can tell you anything. Let's say it tells you, oh, you could be increasing your radio investment because the ROI is super great, but in reality it could be that you're already reaching 80% of the audience. So not looking at that or not understanding realities of it can I think

Juha Nuutinen (30:46):

Be a problem. So where are you at the marketing response curve basically? Are you fully saturated already or closed to fully saturating the channel or are you just on the very beginning of the marketing response curve where you can still double and triple the investment and still get very high returns for the marginal investment? So I think that's very valuable information. For example, in our Google branch church might actually look fantastic not only in GA four and not only in the ad platform reporting. It can actually look fantastic. Also, when you are taking a look at the marketing dashboard, what is my average ROI for Google search brand, it might be the highest ROI channel, but it doesn't mean that the next US dollar you are investing in marketing give us the best return in that channel. You might be actually getting 95% of impression share on your branded search givers already. Marketing mix modeling is the tool that captures that, and that's basically then built in do the optimizer recommendations. So even if one channel looks fantastic, we are actually now optimizing for the marginal ROI of the last dollar invested in

Outi Karppanen (32:03):

Marketing. And I think that's super great because the reality is is that it is complex. Marketing is very messy, yet it's very complex. And I still think my personal view on anything basically automated or giving you data or not even talking about AI is we humans are still needed for our understanding of the data. We can, yes, data can process more, but we can process differently and understand also what is your brand like and where does it fit? And I think overall it's just good to trust your instinct as a marketer, trust the data, but trust your instinct as well on what to do.

Juha Nuutinen (32:40):

Yes, yes. Instinct is very, very powerful and this experience and instinct should not be underestimated. And our goal is not to replace this experience and instinct, but actually give all the necessary data points, all of the data points from different sources to be included in the decision making. If I needed to choose pay only based only AI solution and experience, I would go actually with experience. But I think experience combined with all of these automated insights and really integrating all of the data sources in easy to grasp this summary view, I think that's the best way to go.

Outi Karppanen (33:27):

So summary, do MMM, but do you trust your instinct as well because then you have something to back it up because it's asset easier to go to your CEO and say, why I need 2 million more in marketing when you have data to say that it brings more sales rather than say, well, because I want to.

(33:44):

So

(33:45):

Data talks

Juha Nuutinen (33:46):

Well said

Outi Karppanen (33:47):

Looking ahead, what do you think is the future of MMM or what's happening at the moment in the market?

Juha Nuutinen (33:55):

Okay, so first positive thing is that you have seen a massive increase in demand for marketing mix modeling and that's very easily seen if you just type in marketing mix modeling the Google trends, you can actually see that the demand has tripled or quadrupled in three to four years. And I think that's a big thing is this privacy. But the big thing is also this really the ad platforms supporting this transition toward MMM. So marketing mix modeling. So that's really a good thing. So I of course expect that to continue and strengthening as a trend. I think this is just first phase, I think for the, I expect the trend to triple again in the next three years supported by the ad platforms and also these technologies that are giving now reliable read to this new segment of advertisers, digital advertisers running their e-com D2C stores. Technically, I think at least we are not going to do anything different.

(35:08):

Like I said, we care the most about our customers. I argue that we are the company that cares the most about their customers. We are not going to change that. Quality is still going to be a number one thing for us and unlike getting value to our customers, but I think what we can do better is getting to the value quicker, more easier. When we started the company eight years ago, I think the marketing mix modeling onboarding process was like six months. And I think sometimes it was like thousand hours of work in total to get there. But nowadays, we regularly do these onboardings, especially for digital marketers running e-commerce stores within one week.

Outi Karppanen (35:58):

That's the trend of it's becoming more accessible and easier to use, not to be then confused that I think it does require experts on analyzing and doing it so that it is robust and reliable. And I think that's, from my aspect, what I'm seeing now is there's becoming a lot of tools that you could maybe, if you know about data, you could play around with yourself, but then there are reason why there are experts in the field who understand the data better.

Juha Nuutinen (36:27):

And now that the onboarding is squeezed to one week, we want to actually squeeze it to one day. Yeah. Actually what I want to do, I want the customer to sign an order form and start a data sink in the background. When they come back from lunch, they already have the results basically access to their MMM results.

Outi Karppanen (36:49):

That'll be the dream.

Juha Nuutinen (36:50):

Yes. Your

Outi Karppanen (36:51):

CEO asks you, what are we doing? What's the result of marketing you like? I will look into it ur.com and go on inside.

Juha Nuutinen (37:00):

Yeah, exactly. Exactly, exactly. And we want to make it riskless because it used to be that, okay, minimum investment, 100,000, hey, we are actually doing even three pilots nowadays, by the way. Yeah, go to saleforce.com/supermetrics landing, and if you are actually a supermetric user, if you are an e-com D2C running mostly digital channels, you might actually be eligible for the free pilot of Supermetrics customers.

Outi Karppanen (37:36):

So if you're not using yet, and now it's your opportunity to dip your toe into it,

Juha Nuutinen (37:41):

Zero investment, one week onboarding and yeah, I think you cannot fail.

Outi Karppanen (37:47):

Yeah. So as parting words, besides that amazing effort, what advice do you have for someone who's just starting their MMM journey if they haven't yet done it, someone's listening in and they're like, I'm interested, I want to start doing it. Well, besides contacting you guys, but what should they do? What's the advice?

Juha Nuutinen (38:10):

I think just get started. It's like nowadays, and don't try to eat the elephant in one piece as a whole, right? So like kick in the doors that are half open already, right? So I would recommend to start with a scope that can be super easily achieved and automated. So hey, just take your digital, even if you are doing a 20% or 30% of your media mix in TV or out of home, it doesn't ruin the analysis even if you leave these channels out from the pilot and at later

(38:53):

You can just go in with, Hey, I'm going to go in with Google ads, meta ads, TikTok ads, it's going to comprise maybe my AV or affiliate channels. Hey, it's going to be already like 70 or 80 or even 90% of my media investments. And hey, I'm going to add my e-commerce sales data. I'm not going to add my stores, which are still half of the sales, but I'm going to access my e-commerce sales with this scope. You can actually get the results in weeks. You might actually get the free pilot that we discussed, right?

Outi Karppanen (39:28):

Yeah.

Juha Nuutinen (39:29):

So hey, do that. Of course, you can add your TV and out of home later. Of course you can add your offline stores later. That will give you an even more accurate read. But just not getting started I think is the big mistake that you can do.

Outi Karppanen (39:47):

I think we've been saying this as well as something is better than nothing. So it's a journey. And we talked about the calibration of MMM that is on the roadmap when you've already started doing MMM to some it might be a reality after a year you started doing hopefully quicker, but still taking bite bite-sized portions of that elephant that is MMM, and going little by little by little because something is better than nothing. And I've said last click probably is not an ideal last click attribution, but that's still better than nothing because if you know nothing

Juha Nuutinen (40:26):

And oh my God is our, Hey, we love last click. The reason we love last click is because it's very reliable source for calibration on these channels like shopping pmax or anything related to page search. It's very reliable source for calibration. So that's why we love it. But oh my God, does it give you bad guidance for your YouTube and Meta and TikTok investment? You should. Oh my goodness. Yeah. Or your tv. Of course.

Outi Karppanen (40:59):

That's the perfect place to click on it. Super Bowl was just yesterday. And how many people clicked on Super Bowl ads? Not that many. Most likely. Well, CTV helps, and I think from our end, what we're trying to do to help people build their data foundations is now having custom data import. So you can have your TV and offline media spend come automatically through to your chosen, whether it be BigQuery or whatever you're using as a data warehouse. And we call that all data foundation and we help that. And then marketers, you have access the quality of the data and then they give access to you guys or whoever their partners in other things as well that you do with data. And that's my guidance to anyone in marketing of what to do is well first check you have the data and get the data because if you don't have data,

Juha Nuutinen (41:50):

Yeah, I think let's make it easier. First check, do I still have my Supermetrics license? Because if I have my Supermetrics license, we'll take care of the rest.

Outi Karppanen (42:02):

And is it taking all of your data, marketing data and then self will take care of the next. But hey, thank you you have for joining us your amazing insights and comments. And this has been a great episode.

Juha Nuutinen (42:16):

It was a pleasure to be involved. Thanks. Nice to meet you Outi, and thank you so much for this.