Demystifying marketing mix modeling (MMM)

Key takeaways

1. MMM helps you understand what’s working

Want to know which of your marketing efforts are actually bringing in customers? MMM is like a detective who figures out which marketing channels and campaigns are the real stars. By understanding what's working, you can spend more money on the stuff that's bringing in the bucks and less on the stuff that isn't.

2. Good data is the key to MMM success

Imagine trying to build a house with broken bricks. That's what using bad data for MMM is like. To get the most out of MMM, you need accurate and complete information about your sales, marketing spending, and other important factors. Cleaning up your data might take some work, but it's worth it in the end.

3. Make MMM work for your whole business

MMM isn't just for the marketing team. It's a tool that can help everyone in your company. By showing how marketing impacts sales and profits, MMM can help everyone from the sales team to the finance department make better decisions.

4. MMM is always learning and improving

The world of marketing is always changing, and so is MMM. New tools and techniques are being developed all the time. To get the most out of MMM, you need to stay up-to-date and be ready to adapt.

5. Communicate your findings

Discovering all these amazing insights from MMM is great, but it's useless if you keep them to yourself. Share your findings with your team and other departments. The more people who understand the power of MMM, the better.

Read full episode transcript

Transcript

Daniel King (00:05):
Hey there. Welcome back to another episode of the Marketing Intelligence Show by Supermetrics. I'm Daniel King, a key customer account manager here at Supermetrics. And today we're diving into a topic that's been gaining a lot of attention lately, marketing mix and modeling, or MMM, despite describing popularity, MMM is not always well understood by marketers. That's why we've put together a very special fireside chat to clarify what MMM is all about and how it can benefit your marketing strategy. Joining me are two industry experts, Matt Farrugia, global Chief Customer Officer, and Co-founder of Mutinex, and Kevin Alphonso, senior marketing manager at Seek. In today's episode, we'll explore how marketers can determine when they need MMM. So let's dive in. So before we kick off into anything, let's start with a very quick introduction to the team that's here today. So Kevin, can I get you to quickly introduce yourself, your role, and Seek?

Kevin Alphonso (00:57):
Yeah, thanks Dan. So, hi everyone. I'm Kevin Alphonso. I work at Seek. We are an online marketplace in the employment space in a ANZ and in APAC as well. I work within the marketing team at Seek, and I look after the paid media function within the Seek marketing team. So look after both above the line marketing as well as we've got an in-house digital marketing team and also look after ad tech and marketing analytics as well.

Matt Farrugia (01:24):
Thanks, Dan. But Matt Farrugia, co-founder and chief customer officer for Mutinex who are Mutinex. We provide a marketing mix modeling platform called Growths, and it's designed to optimize marketing investments, ultimately make better decisions around marketing. And we work with, we help businesses manage their data to gain actionable insights as well and improve ROI with a focus on granular, real-time decision making as well across a wide range of industries around the world.

Daniel King (01:54):
Amazing. Thank you very much guys for that. Matt, what exactly is MMM and why is it so important do you think, in today's climate?

Matt Farrugia (02:02):
Yeah, good start, Dan. Great questions kickoff. I think, well firstly, market mix modeling. MMM is referred to and not medium mix modeling. They are different market mix modeling. It's a statistical analysis technique. It's used to measure the impact various marketing tactics on sales and to forecast the impact to future marketing strategies. And it uses historical data to understand the relationship between marketing efforts across some multitude of channels, tv, radio, digital, et cetera, and other marketing channels and sales outcomes. And it analyzes these relationships to help organizations allocate marketing budgets and make decisions around growth more effectively. I'll just talk about some of the key components quite briefly around market mix modeling. I mean, key components are things like data collection, the model development, the analysis and interpretation of the model optimization. Then ultimately decisioning making decisions from that data to, and we'll talk about what type of decisions and how to use it in a minute.

(03:04):
So the second part of your question, why is it important? Why market mix modeling is important in today's climate? Quite briefly, we love to use the analogy at mutanix at MMM in today's market for the modern marketer is what a GPS is. GPS is tool pilot. Without it, you're flying blind and you don't know where you're going. So it has become critical for businesses in today's climate due to several factors. If I explain why it's important, things like a complex marketing ecosystem, the landscape that we operate in now in marketing is increasingly complex with increasing amount of channels popping up every month it feels like across digital, social, et cetera. And MMM helps to understand the effectiveness of each channel. It's also become critical because making decisions from data is now essential across all businesses. And a lot of customers not only are just drowning in data, but trying, and a lot of 'em are desperate to find the signals from that noise and make the right decision.

(04:09):
And MMM is able to enable just that. And then there's some other key reasons, which I can elaborate a bit further later, but things like evolving consumer behaviors, increasing competition privacy regulations and digital attribution channels is increasingly becoming a very big reason why with regulations like GDPR and what we've seen in CCPA in the US and increasingly now in Australia and the demise of third party cookies, because MMM reliance on aggregate data rather than individual data or individual user data makes it a far superior and valuable tool for measurement. And then finally, I'll just call out, the other reason why it's critical for brands today are things like economic uncertainty and a business needing a way to help navigate those ups and downs of the economic climates. And finally what we hear a lot is organizations wanting a holistic view of marketing impact in a single view across all their efforts is increasingly becoming a big reason why MMM is important today.

Daniel King (05:17):
Absolutely. Yeah, I think again, having these conversations that super metrics, a couple of things you pointed hit the nail on the head on there in terms of the noise, the sheer volume and scale of data and trying to just make decisions based off of that. So you spoke specifically about some of the issues that we're seeing in the market. What's changing? What are some of the key challenges that businesses are actually coming to Mutinex to look at MMM specifically to try and solve it?

Matt Farrugia (05:47):
Yeah, look, if I go a little deeper, I mean there's quite a few, but by nature of the marketing landscape being a lot more fragmented with so many different channels and opportunities for, if you put a dollar here and a dollar there and a dollar there and you've got multiple channels operating your marketing tactics, every one of those channels individually if you report on them, looks amazing. And every one of those channels can seem like they are attributing to a sale and a success. So one of the channels is really understanding the impact of different marketing activities on sales and overall business outcome. So MMM absolutely helps attribute sales to specific individual marketing efforts and provides a clear picture of which campaigns and channels are delivering the highest return on investment. Another challenge, as you've mentioned earlier, economic climates make it difficult to allocate budgets.

(06:41):
Budgets might be decreasing where there's opportunities to lift budgets when economic climates are taking a downturn. We saw a lot of that through covid, but one of the challenges is how do I improve or improve my budget allocation? Because allocating budgets across a multitude of channels as well can be really, really challenging. So MMM enables businesses to identify, again, which marketing activities are most effective and allows them to allocate budgets more strategically and more frequently. The other challenge I'll just say as well, I mean when I mentioned a holistic view and we hear a lot about unified view and it's the kind of promise from many analytics platforms unified view, and it's a term that I kind of irks me sometimes because it's widely overused, but gaining a holistic view of marketing efforts, anyone on this call that's managed multiple digital channels in a campaign is really, really difficult because integrating data from multiple marketing channels online and offline to get a really comprehensive understanding of performance is bloody hard.

(07:50):
And it has been for a long time. So MMM does aggregate that data from all that activity and does connect it to provide a unified view and also provides a view on how the different channels interact and contribute to overall performance. And then there's also, I mean I can speak a lot about to the channels challenges that it solves, it solves all the challenges of the world for marketers. But if I was to highlight some priority ones, and Kev will elaborate shortly on some of the ones that they're solving, what we're seeing a lot lately is reducing wasteful spend, really, really managing or understanding your wastage of where you're spending, knowing where you're overspending and underspending so you can improve, how you can reallocate that spend or drive better experimentation across the business. And also the other one that we're seeing a lot is really diving into an understanding of how market changes and seasonality are impacting marketing because we're noticing a lot more, as we're all noticing the market market fluctuations and seasonal trends and competitive actions are happening more frequently in today's world.

(09:02):
So MMM can absolutely help solve for that challenge. And then finally I'll just call out another really big challenge which we're seeing become a priority is MMM is really driving a major improvement in stakeholder communication across an organization for the value of marketing investments to stakeholders outside of marketing. It's really driving and helping with the literacy of marketing and finance in particularly, but also really driving value in stakeholder communication for other functions like organizations', product or sales even all the way through to procurement as well as what we're seeing. They're some of the key challenges that we're seeing MMM. So there are others, but we can touch on them in a minute.

Daniel King (09:53):
Amazing. Thank you so much for that, Matt. Again, you sort of raised a couple to the top there that I wasn't even familiar with in terms of how MMM can sort of not just solve for the marketing use cases or the marketing requirements, but also company-wide and other strategic initiatives as well. So very interesting to hear. I think this is sort of great point to hand it over to Kevin actually and hear from Kevin. Kevin, I'd love to understand a little bit about what brought you and the team at Seek specifically to actually start considering MMM. Let's jump in there. I'd love to hear from you.

Kevin Alphonso (10:24):
Yeah, sure. Maybe build background about Seek as well. So we are Australia's leading marketplace for employment. And now as you can imagine, we have a whole lot of visits coming to our sites as well as business metrics as well. But very often for advertisers you can kind of attribute visits more to your lower funnel marketing and your performance marketing, but you can't get a complete picture on some of the other parts of your mix as well, like brand marketing on ships and the likes as well. And that's where MMM really helps I think, because it gives you a complete impact of your entire marketing funnel. And I like to distinguish something which is attribution and contribution. Sometimes you can look at channels in isolation and it will tell you the channels are doing well, but how do they work across that funnel and contribute to each other in the entire life cycle?

(11:16):
So that was something we were looking for. The other is just evolving our own medium maturity as well. When we looked at our toolbox, we had lots of tools for operational metrics, but individual tools, we've got amazing data at seek brand health metrics, brand live studies, but MMM takes it to that next level where you look at it and, sorry, mine, I've got to say this word again, holistic, but A looks at it that way. But also it allows us to separate out marketing's impact from some other factors as well, like base sales, external factors, and just talk about marketing's incremental impact on business metrics only. And then we had a big campaign launching in market last year. We had a new brand platform coming in, so we thought it was really the appropriate time to kind of bring in this capability. It would enable us to optimize that campaign throughout its life cycle and give us some kind of insights in terms of whether it's an ad stop saturation curves and continually optimize that campaign.

(12:17):
And then most importantly, as I think with all advertisers, we are constantly being asked to kind of talk about the ROI of marketing efforts in this macro environment that we are all in there. Are there more scrutiny on cost as well? And you really get a seat at the table when you talk in the language that some of your other stakeholders outside marketing are talking about as well. So MMM takes a lot of those boxes, but these are really some of the challenges that we were trying to tackle along with some of the external factors that Matt spoke about coming out of the pandemic. We are using a wider range of marketing channels now, but also consumers behaviors have changed where they're now across a whole range of platforms and publishers than we were using even two years ago. And then external factors like privacy controls have made some large touch attribution or multitouch attribution a lot harder to do than they used to be before. So MMM kind of mitigates a lot of those factors as well.

Matt Farrugia (13:14):
And if I was just going to add some of those points as well, Dan, I think Kev made a really good point there. I think aligning the language or taking the metrics that matter, and we'll talk about metrics in a second hopefully, I'm sure there's a question, but I mean helping inform the language around marketing's impact to the wider business and speaking the language and unifying that law, making that language a bit more widely understood is so important and so often missed. And what I mean by that is we, we've all seen it and probably been part of it over the years. I mean if you take a report today to the C-suite and it says this is how marketing is performing well, we've lifted these metrics around brand metrics and we've got this many likes and social media and this many shares and comments, what does that actually mean for the bottom line? What is that actually doing for growth? If anything, what that's doing is really pigeonholing and limiting marketing as positioning marketing as that growth engine for a business is what it actually really is. So MMM does really help enable and push marketing to have those more valuable conversations about contribution to the organization's success metrics and bottom line.

Daniel King (14:29):
Yeah, absolutely. I think it's a great point, Matt, and I think you actually beat me to the punch and I think now is actually a great time to follow up with that in terms of what are those specific metrics that actually matter when it comes to MMM and what are those KPIs that businesses are ultimately chasing?

Kevin Alphonso (14:46):
I'll tell you what, and I'm not a cop out, but these metrics kind of differ for different businesses. I would say that's where the upfront thinking is really important even before you start to scope this project. But ultimately I would say you want to focus on the metric that matters to the key stakeholders in your business and what matters to the bottom line. And that's where we will differentiate between business metrics and lead metrics. So visits and some other metrics that Matt mentioned are lead indicators, but ultimately choose the metrics in your model that are actually closest to revenue. And your point of sale I would say is what matters. And that will differ for different businesses. Some sell online, some you might sell offline, but yeah, and you want to choose as tough as it is, you want to just choose one metric ultimately because ultimately you want to have that one-to-one relationship between your marketing investment and then dollar value of revenue.

Matt Farrugia (15:47):
Yeah, a great way to explain it Kevin, and I think if I were to build on that, I'd emphasize it. Absolutely. As Kevin said, it does depend on the priority on the business and the category you're in and also what maturity you're at with marketing data and even tech in terms of context to MMM. But first and foremost, metrics that matter are the metrics that your business exists for to drive. Yeah, so as Kevin said, they do differ category to category. If you're in financial services, obviously you do marketing to drive home loan applications. If you sell soft drinks, you do marketing to obviously build brands and sell units. And in the context of market mix modeling to build on that, the metrics that market mix modeling insights and reporting can provide. If I were to list a few in terms of, and again, these differ and are varying priorities for different organizations, but ROI is obviously a metric.

(16:49):
And just to explain that briefly, what do we mean by that? I mean, you can measure the profitability of your marketing investment when you're comparing revenue generated to you've spent on marketing activities quite simply. But then it's other things, like we said, we can break down. The sales contribution can become a metric. The incremental sales, which Kevin mentioned earlier, you can isolate. It is so critically important for a marketer today to understand and know your incremental sales even at a monthly level. And that's also a metric in define with MMM. And then there's marketing efficiency, then there's media spend effectiveness, and I'll just call it a couple more elasticity is also, there can be metrics around elasticity, which what I mean by that for those unfamiliar, it's the sensitivity of sales to changes of your marketing spend, help you understand how to allocate budget around that. And the final one in terms of a metric, which I'll call out, there's also lots more, but these are some of the more common ones, is ad stock or decay rates are a metric as well. Like ad stock for those unfamiliar are the prolonged effect or the rate of decay or effect of marketing activities over time and how quickly the effect of your advertising diminishes. So that helps you with creative messaging for example. So they're just some of the key metrics that we see are really, really important. Driven by MMM.

Daniel King (18:13):
In terms of your experience of this MMM implementation, what were the biggest key learnings for you as you got this off the ground from where you started and to where you are today?

Kevin Alphonso (18:23):
Yeah, I'd say data quality is probably one challenge that we faced. And we are a business that has a lot of data and pretty well organized data we thought, but we did come across some unforeseen challenges because you're talking about historic data going back, say what, four to five years? And when you considered a lot of that data overlapped with a period of up and down change in the market, given that it took over the covid period as well. We're also working with multiple digital platforms, agencies, and vendors. We did come across some challenges with maybe some inconsistencies of data and incompleteness. This is something that I'd say because a lot of people doing it for the first time might encounter as well. And obviously we are a great partner in EC that kind of helped us. They've got a great platform that also allows us to kind of share data in a very secure way. Dan, your team and Supermetrics also obviously enables some of that data transfer. So I'd say the data quality and the infrastructure, we had to set up connectors that maybe we didn't have before and maybe intervene in some cases to get that data as robust as possible. Data garbage and garbage out as they say, but that quality of that data is so important. So just a few, I guess, hiccups along the way, but also some new learnings for us also. But once set up, everything just gets smoother as you go along.

Daniel King (19:49):
Okay. And Matt, what about your experience in dealing with customers that come to you or people that come to you looking to move next to their MMM? What are those specific learnings or takeaways that you've seen over the years that you might be able to give back to the audience today in terms of that implementation stage, your first time being introduced to MMM?

Matt Farrugia (20:10):
Yeah, look, I think Kevin pointed out a couple of the main ones, but if I was to look at that, I mean first and foremost, I mean when you see data quality and integration, and I also preface this by saying, well, I'm going to talk about the challenges of implementing an M solution. I mean, a lot of these challenges I'm about to talk about are exactly why mutanix exist. We obsess about these challenges every day. We deal with them on a daily basis and solve these challenges. And the first one is data quality. It is pooing poo out, so to speak, politely. We've all heard that, but we've got to ensure that data is accurate, it's complete, and it's coming in from the multiple sources that's required. And what I mean by multiple sources, there are sales data sources and finance data sources, et cetera.

(20:58):
And marketing as well. I mean, marketing comes from your, there's probably a lot of internal platforms in your organization, but also there is paid media data sources that come through various places and also vendors. So there are means to help solve for those as well. And improving methods there, I might just say. But inconsistent or missing data can lead to incorrect outputs. So yeah, I mean, I just want to stress that you've got to be working with really solid data management systems with clear processes and protocols for collection, and then it's rinse and repeat. The other one is like alignment with business objectives. I know we touched on this early, but without aligning with the business upfront, it's marketing pushing poo uphill, if that makes sense. And driving their own agenda into and upwards in the organization. So learnings from us is that we've seen great success and Kevin has done an incredible job at this as well at seat with his team.

(21:54):
Aligning with those objectives right from the start means that this solution is not there just to solve for marketing. It's there as the growth utility for the organization, which is really, really important. These MMM programs or solutions can be resource intensive on the customer side. It's okay to have, I mean, it's one thing to have an appetite to do it and a need for it, but it's another thing if you are choosing a vendor or a solution that has a method or an approach that does require a heavy resource burden on your organization. And I can probably attest many on the call here who are in marketing and manage teams and hire teams and build teams. It's not as if you have a myriad of analysts doing nothing and ready to go when you say you want to kick this off and applying for additional FT is difficult these days in increasing headcount.

(22:48):
So you've got to have that case building that case to have that. So working with a vendor that really understands resource allocation and works with you, so they're not a burden as such is really, really important. I mean, we spoke about sales and a bit obviously attributing sales, limited access to relevant data. Absolutely. That's a consistent one I'd probably say there. There's an opportunity that I've seen with brands who enter this program, especially with us. I can only speak on customers who work with mutanix have really used our solution as an opportunity to gain access to relevant data and also reassess their internal kit around analytics as well. So it's kind of had that double value point there. Inconsistent data, absolutely. We did speak at this and lack of expertise in marketing. I'll touch on that. There's historically, with the historic approaches to MMM, when there's been a, let's do this massive audit, let's build this massive presentation, and in the final endpoint or output of this an MMM study over six to eight to nine months has been a 500 page PowerPoint pack.

(24:00):
That takes a lot of expertise to read through, disseminate that information and find those insights. That is absolutely a big challenge that we see. So again, when selecting a vendor, I highly encourage all of our customers and also prospects or just anyone who's actually talking about approaching NM solution is really, really look at how your internal team, a willingness to develop or upskill and build capability and look for vendors that really are focused on educating and handholding and crawling before you're walking and then running, but very soonly, but having the resources there to support you in your journey throughout that way, because it is a journey and it does evolve over time. I mean, the last one's really important, buy-in from senior stakeholders around measuring effectiveness. Absolutely. And it goes back to my earlier point, speak to the organizational metrics that are important. It is revenue, it is sales. What are the metrics that make that organization live and exist? And then it's linking your MMM program to those metrics. I often see these myths. It's so common when it's solved, you look back and go, geez, that was so simple. And that's what we mean by really, really aligning the language and really, really aligning those universal metrics to the organization.

Daniel King (25:21):
Amazing. Yeah, some great stuff there. And I just want to go back to you, Matt here. Just MMM seems to have evolved over the years. What has really changed in the last couple of years to make it so relevant and so impactful in the current climate and given the existing needs of the industries that Tex works with?

Matt Farrugia (25:42):
Yeah, it's a big question, Dan. It's a great question. I think that you get asked this quite often, I think how MMM has evolved. I mean, yeah, everything, the demand side has evolved obviously from what we've spoken about is the market's changed, media's changed, marketing's evolved, et cetera. There's a real need there. But I think if I look at how MMM changes, there's been some massive changes to MMM over the years. And mind you, I mean, market mix modeling is nothing new. It's been around for probably more than 50 years now. One of the more recent changes is that Mutinex was born. No, that's just a self plug, but I'll go onto that later. But I think if I was going to talk about just some of the key themes for change and how it has changed, I'd probably start by integration of digital data.

(26:27):
MMM. Historically it was a lot easier because it only focused on offline channels, tv, print, et cetera. Modern MMM today does integrate all channels, digital, social, et cetera, CRM, also all the way through to things like sponsorship. You could also look at earned media and all your internal owned assets as well, and a big one On the second point, I'd probably say the statistical methods or techniques have evolved significantly. I mean, again, over the years, early days relied primarily on linear regression models. Where today approaches are, they incorporate a lot more advanced statistical methods, like obviously machine learning techniques, bayesian models to improve accuracy. And those techniques obviously allow for more sophisticated analysis, but we're able to use those methods with the vast amounts of data. Now because of the advancement in tech, because the access to cloud and the sophistication and cost of operating in cloud means that you can use those models in ML and analyze just so much more data than what you could have only probably 10 years ago.

(27:41):
The other thing, the other key themes, and I won't label these too much just for the interest of time, but today, modern MMM solutions have much more granular data analysis because they use aggregate. We use aggregate data and even more at a more frequent approach, like monthly mostly also change because you can incorporate external factors a lot easier now, macroeconomic factors like weather, economic factors like interest rates, et cetera, and then the myriad of other macroeconomic factors, and then just a few others, the automated side of things with modeling and agile modeling. And then I can go on, but the last one I'll just touch on is the accessibility or usability of MMM today and the advancements we've seen with ui, I mean by nature of what Mutinex do, what we do and how we deliver MMM, we deliver MMM in a platform experience to what a lot of marketers and users might be akin to using tools like Google Analytics or Power BI and Tableau and the like. And yes, we do have the support teams who are humans around that, but it just means that we can get those insights and data and manage it essentially through a platform and keep it much more up to date and much more accessible to make more decisions more frequently. So they're kind of some of the key themes of how MMM has evolved. Well just

Kevin Alphonso (29:06):
Course

(29:07):
Something at Matt as a marketer, looking back at what we were presented maybe five, six years ago, you kind of had a lot of static charts. There was a bit of prediction, here's what your sales will do if I project out 12 months and that's what you had. But today, what we have in the platform from next to the Growth OS platform is there's almost actionable insights for different levels of stakeholders. Your csuite might just be interested in the top line numbers and the incremental impact, but someone for me who's actively managing budgets, I've got budget scenarios that are using predictive analytics now so I can kind of project for the next financial year planning, but then there's others in my team that might find a lot of actionable insights in terms of what creatives are doing well from an ROI point of view, which publishers should we drop on, increase things like are we over attributing in some channels? What's the right length of time for a particular channel to be in market? Are we approaching the point of diminishing returns? So that's making it far more actionable than it used to be, I think when we go back a long way ago. And it's in an interactive platform in the way that marketers operate on a daily basis at different levels. As I said.

Daniel King (30:20):
Thanks for the additional points, Kevin. That was great. I suppose Kevin, I'd actually love to ask you specifically here, what are the, I know we've touched on this again, Matt sort of touched on this, but what are those key learnings that you've really taken from this whole experience in terms of where you started and where you are today and where Seek is going in the next 12 months with MMM as well? What are those key learnings and what would you say to those audience members who are keen to learn from your experiences?

Kevin Alphonso (30:47):
Yeah, absolutely. I think number one, upfront thinking absolutely critical. Even at the point when we started think about MMM, we started to think about what this project was going to deliver from us, what would we want to get out of? It was really critical, and we started to not just talk about it within marketing. We started to talk to stakeholders outside of the media team as well. So we've got a separate marketing and data teams spoke to them. There were teams from finance, other parts of our business that we consulted with. So I'd say do your upfront thinking, because you're going to have to think about what your data structure is going to be and how it aligns to your current business model. The other I would say is data infrastructure. I know we spoke about that for a bit, but make sure that your organized internally in terms of having a team where your data resides and what you might need to do.

(31:37):
So you're prepped once you're kind of ready for rollout. The third one I would say is also start small and gain some initial success. What we are doing is we've got an active campaign in market, and we've started to make a few optimizations on that campaign, either adding in channels or opting out channels so we can build some success and some experiments, which will resonate quickly internally. So we are starting small with some of those wins, I would say. And finally, I'd say choose the right partner to kind of partner with you. You need a partner that understands your business, who can take all the context that you have about your business and kind of work with you to build the right model in MMM, because there could be a lot of different models and you don't want to go down the wrong path.

(32:27):
For us, what was really important as we kind of started to work with Mutinex was that they had a very clear work roadmap for a long time. So as our business evolves, we could see that there was a roadmap that would kind of adapt to how the landscape was changing, but also our business. And yeah, look, you've got to think about complimentary capabilities as well. We have data strong business, but there are certain things that Mutinex brings to the table that we don't have or we couldn't build ourselves through the journey as well. So I'd say just a couple of things for people to think about if they embark an MM m for the first time.

Daniel King (33:01):
Amazing. Thank you so much for that. You actually touched on it there yourself a moment ago, Kevin, in terms of how to get up and running, where do you start? What are the minimum or basic requirements to run an MM study?

Kevin Alphonso (33:12):
Yeah, look, you want to have a minimum amount of historic data so that, and I'll match the expert here, so I'll let him build on that. But I'd say making sure that you've got access to about at least four years of historic media spend and sales data. So that allows the model to kind of take a whole lot of history so that your predictions are stronger annual inputs, we chose that long a period solely because we want to get some data from the whole covid period when there can be a lot of kind of ups and downs in your data and shifts and spends. So we want to make sure that we got some data from prior to that as well, apart from having a fairly robust data and obviously having it as granular as you can as well. I lead to the weekly level I think is best. So I'd say having that in place first is good. And then just align to some of the points I mentioned earlier, which is making sure you understand your own business really well. What are the segments by which you would like to see your data structured? Because ultimately you want the output in the model, in the way you plan your market mix as well, so that when you get those outputs, they align to how you do planning within your marketing team. So I'd say just those two things probably very important at the start.

Daniel King (34:32):
And Matt, do you have anything to add to that?

Matt Farrugia (34:35):
Yeah, I think Kevin did cover it quite comprehensively. I might add just we do have criteria that we speak to for standing or implementing market mix modeling for customers. And the first one, as Kevin mentioned, data, I mean, a lot of customers do kick off and come into the platform with four years data, some even more. We do have minimum requirements that are less than four years, and also some specific data requirements that we work through and have a conversation with customers before they start their journey on what data is required. We can share those data requirements. I will say no without going too detailed into it, and I mean things like how long should I go back? But also, how granular does the data need to be? How frequent is it a weekly time series, et cetera? Granularity can always be added over time. Sometimes you don't want to boil the ocean too much and get every single data source and every single detail and granular point in that data, external data, et cetera. It's about standing the platform up and then enriching that data over time. But I would probably encourage it, if anyone's interested in learning more about those data requirements, hit me up or flick us an email, we can set up a time and just step you through what requirements they are and what requirements we see are needed to get you up successfully.

Daniel King (35:57):
Absolutely. What does that typical journey of a customer look like with zero MMM experience, right up to full exposure with Mutinex? So when they're absolutely starting with absolutely no experience of MMM, no understanding of what it is, and then how to get completely up to scale, up to standard, where Seek is today, what does that journey actually look like?

Matt Farrugia (36:17):
We worked with customers locally, globally, at all different levels of data maturity and appetite and capability. Customers that have had no experience whatsoever, but know that it could be solving some problems. I mean, we do initially really try to understand the customer's problems first, and that's where it starts. What problem are we setting out to solve? What questions we want answered, what metrics we need to align to? And then our team, we've got an exceptionally talented team across our marketing science team, our account managers, and also our sales team that are very experienced in having these conversations about what I've just said in asking those conversations. Once we understand what you're setting out to achieve, then we can structure or at least put the what's required to go forward and really provide the information or support that you need every step of the way. And each step is kind of gated.

(37:18):
We make sure that it's understood and all questions are answered along the way, and it is really pulling your hand along the way and providing the education and the support, but ultimately it's helping you build that business case internally so you can confidently take it to your stakeholders and communicate the case for implementing an MMM or MMM solution, not just because it's a nice to have or everyone's doing it, so should we, but really, really showing the, as I said, aligning it to the business metrics and setting out well what it is likely to achieve all the way through to what incremental gains we are likely to achieve or what we're looking at growing with MMM or what we're looking at saving or how we're operating more efficiently. So it's probably understanding what those most important use cases are for you first and then educating you around them and helping you build that business case along the way.

Daniel King (38:13):
Yeah, absolutely. When marketing budgets today are being so constrained when there's so many eyes on where that budget is going, where spend is, how do you go about garnering support for an MMM project? Again, going and asking for more when they're looking to give less, that is a hard conversation to have. So how do people go about doing that?

Matt Farrugia (38:32):
Absolutely, Dan, look, and we've seen this is also very, very common one, right? I mean, look at economic climates around the world. There's very few organizations are doubling their budgets each year. Accountability is rising. Organizations are expecting more from less. That is just a universal truth right now. So what we've experienced, and even going through covid, coming out of covid, all of that occurred. So it starts with if your budgets are constrained, we see a bigger need for having a market mix modeling solution in place. Why? Because you can quickly understand what will happen if you reduce your budget and the negative impact that may have to your business. It's very difficult historically, and still a lot of organizations find it difficult to really defend a budget reduction without the data. And we've seen quite a few case studies where organizations or CMOs have been requested to reduce their marketing budgets, but the senior marketers have had the access to our platform, have been able to leverage predictive analytics with our scenario builder and very quickly establish, well, if we reduce the budget by X, we need to accept a reduction in sales impact over here, and there's a trade off.

(40:00):
So very quickly you can objectively make those decisions, informed decisions and support and defend conversations around constraint budget. So in terms of going through the process to request budget for a market mix model, and how do you do that when your budget's reducing? Well, it's building that business case and really demonstrating what you're spending, what you would get, what are the use cases for a solution like market mix modeling, how you would demonstrate ROI and value. How will you demonstrate not just being more efficient and reducing budget, but also how you would demonstrate while you're doing that, you could also identify opportunities for growth. You could also identify opportunities to improve your creative through ad stock metrics and the like. So you can leverage MMM to navigate through economic pressures or economic downturns that a lot of organizations are experiencing a highlight. The efficiency gains, highlight the fact that you are implementing a capability that historically would require 20 new headcount, but you're doing that with very, very, very few headcount and your gains are much higher. So it's a pretty obvious argument when you present it that way. And if you focus on the quick wins and the roadmap for the plan of how you plan to engage this, propose it or the detailed commercials, et cetera, and then roll it out and evolve it. And the final one I'll say, which always helps customers get over line is competitors, right? If you are not doing this and your competitors are, what do you think is going to happen?

Daniel King (41:33):
Yeah, absolutely. I think that's possibly the best question. Yeah,

Kevin Alphonso (41:37):
I think it's about changing narrative, right? Because there's a paradox in that question because if you start to think about everything as a cost in marketing, then you're not really moving the conversation. Two words the way business leaders speak ultimately, think of it as an investment, right? Marketing completely is an investment. So is your market mix modeling investment as well. And once you start to change that thinking in your conversation with senior leaders, but also within the marketing teams, you don't see these as incremental costs. Think of when you think about the questions it's going to answer so that you're delivering an uplift in sales over and above your base sales, but also what happens if you don't keep investing in marketing? What's the impact of that on your base sales as well? Suddenly you're talking a very different language, and then no one's ever going to question you about cost anymore because you are talking like a growth engine to use a term that Matts used before, I'd say, but encourage people to think about it that way. Talk about the gains. It's going to give you the fact that this is going to optimize your spend, help you make better decisions because there are probably pockets of opportunity already in your marketing. Maybe you're over attributing, maybe there's a bit of duplication in channels. Maybe there's a halo from one segment to another. This is what it's going to uncover and start to solve business problems, not just marketing problems. And that's where you'll start to see a diminishing concerns around incremental cost in such kind of technology and capabilities.

Matt Farrugia (43:09):
Dan, I love it. Answer questions better than I can.

Daniel King (43:14):
Absolutely, absolutely finesse that one. Thank you so much, Kevin, for taking the time to jump on today. And the same to you, Matt. It's been an absolute pleasure speaking with you both today, and I look forward to chatting again in the future. But with that, have a wonderful week everyone, and I look forward to seeing you all again in the future.

Kevin Alphonso (43:30):
Thanks, Dan.