Henry Innis is the co-founder and CEO of Mutinex. In a typically frank exchange of views, Henry and Ellie discuss the long evolution and more recent explosion of MMM, the strengths and weaknesses of modelling, the various challenges of making a modelling project sticky and meaningful, the art of pitching for MMM, the potential and complexity of the vast US market, working with MMM and agencies, and new technology horizons.
You can listen to the podcast here:
Follow Managing Marketing on Soundcloud, Podbean, TuneIn, Stitcher, Spotify, Apple Podcast and Amazon Podcasts.
Every single one of the Fortune 100 does MMM. Every single one has an MMM in place. There’s not one that doesn’t.
Transcription:
Ellie:
Hello everybody, my name is Ellie Angell and welcome to Managing Marketing, a podcast where we discuss the issues and opportunities facing marketing, media, and advertising with industry thought leaders and practitioners.
And remember, if you are enjoying the Managing Marketing Podcast, please either like, review, or share this episode to help spread the words of wisdom from our guests each week.
Today, I am expecting wisdom because I’m joined by Henry Innis, the founder of Mutinex. Welcome, Henry. Thanks for joining me.
Henry:
Thanks, Ellie.
Ellie:
Let’s get right into it. I’d like to talk about the history of what Mutinex does, both as Mutinex and going further back than that.
So, in one way or another, I’ve been personally in my career, exposed to what you could call models, whether that be media mix models or attribution models or econometric models (call them what you want) for well over 20 years now.
So, the concept of attributing media or marketing investment directly to commercial outcome is nothing new. But the way it’s discussed now does really feel like a new frontier. And people are really … the adoption is going up and up.
What’s the primary driver of that kind of explosion in the industry, do you think? I mean, it could be macroeconomics, it could be commercial pressure, it could be internal pressure on marketing, it could be tech, it could be increased awareness. I mean, it could be all those things, but what’s your view having started Mutinex?
Henry:
Well, I think I start with what’s the problem that every marketer faces. And I think for me, the empathy around the problem’s very, very important. And I have this very visceral picture in my head that was once described to me by a brilliant marketer who said she was sitting in the room, and she was in front of a CFO, and she was being asked, “Well, if we put our dollars here or here, what happens?” And she had to say the fatal words, “Let me get back to you.”
To me, as business speeds up, as the economies to some degree speeds up, marketing undoubtedly has sped up. What that has actually put a requirement on is faster and faster answers. We have a demand on us now to answer more holistically across more complex ecosystems at a much faster pace than we ever have before.
And so, to my mind, I think that is what’s driving our requirement. So, forget the Mutinex proposition. I think that that fundamental problem and tension in the market has created a series of companies and challenges to the existing attribution market. Whether it be us or there are a number of other providers as well, to be fair, who are actively trying to solve that problem because we can all sense that tension in the market.
And so, I think markets to some degree create companies as opposed to companies create markets. And I think it’s the height of arrogance to think that you create the problem. I think that we have understood that problem and empathized with that problem and want to solve it.
I think if you start to look at, well, why is MMM starting to be a good solution to that problem, as opposed to 20 years ago, as you rightly point out – the core reason why I believe is the right is of cloud computing.
So, before the cost to run and simulate models and to understand kind of what a model was doing, what we would be able to understand out of a model to simulate what was kind of going on, what would happen if we did this, was very expensive and labor intensive, and largely relied on humans running those calculations.
Today, that job has shifted to computational power, which has its tradeoffs. You get a lot more in terms of your ability to provide complex models that can accurately represent a business, but you also have far more complex techniques and statistical techniques that may not be as accessible as they were before.
And even looking at the open-source models that were launched at Meridian, was launched recently, Robin’s been around for a while. I spoke to a customer this morning who said they were running Robin, it was great, but it gave them 70 possible models to choose from, and they had no idea how to select any of those. And I think that’s inevitably true.
Ellie:
Yeah, that last comment really chimes with my own experience where it was like how long is a piece of strength, and finding the actual optimal became really, really hard.
But I hear everything you’re saying and certainly there is a need in the market for sure that has created cloud computing. I mean, you are sort of talking to tech as being the explosive force there.
What else are your customers saying to you? The ones who’ve now been embedded for a year or two years, what are we solving for and what’s do we still need to solve for?
Henry:
Well, I think there are three classes of problems that you solve for. So, firstly, is how do I justify and defend my budget at the boardroom level? Which is commonly becoming a question of frequency as well.
So, previously, I used to have to defend that process once a year. So, I would use a model once a year in preparation for that. Now, the frequency of that defense and that justification has dramatically increased because marketing has sped up, business conditions have sped up. So, I think that that frequency is a factor. We estimate it’s about monthly that defense has to happen. I think it will go to weekly or daily. I don’t see much that we can do that sort of speed, but I’m yet to see the value coming through there.
I think the secondary component is optimization. So, how do I actually figure out how to improve or extract more out of my baseline?
So, we had a very good example of a customer who did that. They identified in a recommendation from the MMM in a particular market, one of the APAC markets that television would’ve supported them within this geo region.
They then ran a test and control, and they lifted above their baseline by about 6% within that region in terms of typical volume that they would’ve done. And so, you start to see good results like that in the optimization side. And when you compound those sorts of results over the course of a year, you can actually get quite a lot out of an MMM to do that.
I think anyone who claims an MMM can boost a business by 50% is ludicrous. And-
Ellie:
That’s the base of the businesses, in which case you wouldn’t need an MMM to start-
Henry:
Correct. So, I think where MMMs become very, very strong is if you’ve got a very complex market mix and you need to find the optimizations that a human probably can’t easily see. And you can’t easily see just by looking at the data. And so, you get a lot out of that side of things.
And then I think you get a lot out of the insight side of things. So, I think with the rise of our LLM product, for example, MAITE, I’ve seen it being used a lot more to query insights and to validate a hypothesis and go, “Is this supported by data?” And being able to navigate that really, really quickly and powerfully. It’s almost like insight testing meets econometrics to some degree.
And that’s a more new use case I’ve seen, something that’s probably getting me quite exciting in terms of the value that it brings. I measure my business on usage, so I can tell you two things about my business.
One is I know the accuracy of the models that are running across the board. We put a lot of observability and testing around that. And then the second component is that I look at usage religiously across the business. And usage is the responsibility of both the marketing science partner and the product teams to make sure the platform is easily usable. And we can see people doing things that kind of indicate to us that they’re getting value out of the product.
Ellie:
It’s really interesting you talk about usage there because … and it’s linked to a thing that is maybe running through this conversation more than once. You talk about using the product to defend budgets and to optimize budgets.
Implicit in that is a degree of change management. And if the model is agnostic, which of course, it will be, it’s data-driven, presumably you still have to tell marketers sometimes things that they don’t want to hear about their budget or things that they don’t want to hear about the traditional media mixes that they want to give up.
That was always a big problem for me. And often, it was further down the chain of seniority was just trust – either the trust or the belief or the willingness to make changes based on what a model recommends wasn’t there because people were very protected over their patch. And that led to a drop-off in usage and it ended up just gathering dust on the shelf.
Henry:
A hundred percent.
Ellie:
I’m not even sure I’m going on a question with that other than to say, how much do you still encounter that? Or is it really with the power of the technology and the power of the product, it’s just a non-issue anymore and people are just accepting?
Henry:
Well, even Google has, like encouraging organizations to spend on search even though it’s the largest monopoly possible.
Ellie:
Well, look where last click got us all with … everything is attributable to that.
Henry:
But I see every enterprise business to some degree has people in … I almost call it like cheerleaders are very, very common in enterprise SaaS businesses. Someone who’s there to encourage you to help you understand how to take the decision to really be your partner in that.
I think when I first started out, and we first launched [inaudible 00:10:24] as it was known at the time (terrible name). But we had everything pirate-themed, but it was, oh gosh-
Ellie:
I know, I remember the news days.
Henry:
But I distinctly remember the change in thought process that came from this, which was, we had a very light customer success function. We stuffed up customer success probably three times, I’d say in the early days for Mutinex.
So, it’s very easy for people to kind of look at Mutinex now and go, it’s a business that works with things like that. We made a lot of really bad mistakes early on, some of which cost us customers earlier.
And we lost a pitch for a very large tier one business. I don’t think we lost it by too much. And we were lucky enough that in that pitch, we had somebody who was kind enough to give us really good and constructive feedback. And she met with me quite a lot actually, she’s a friend today.
And she basically said, “Your tech is awesome.” Tech works really, really well, it’s what we want. But what we needed was a marketing science person there who could sit there, coach us along with it, and focus on bringing the product to life. And you need to bake that into what you do.
And that was a seminal turning point in our business, was embedding those marketing science people. And to some degree, because we don’t have the cost of data wrangling, data modeling, all of those crazy things that are effectively plumbing that you are paying for with consultancies and things like that, you can actually put a much better ratio onto that change management piece in terms of your P&L management, and your P&L management.
So, in a funny way, even though you’re a SaaS business, you can actually mostly use your kind of leverage in SaaS to provide a better servicing ratio on the other end, which I think is a part that it’s really hard to consider like a consultancy versus SaaS.
I know of another business in the states, Recast who does a fantastic job. They’ve got a similar philosophy to us. And they’re really, really bright business, great business, really good people where they kind of have focused on a similar approach. And what I really, really love about that is you can just spend more time actually helping on that side of things.
And so, one of the things I think is very interesting in RFPs generally in this space, is the RFPs often focus on a dichotomy of consultancy versus SaaS, as opposed to where I think it should be. Which is, how do we understand how much time we’re going to get on change management and helping us operationalize the model, and how quickly are we going to be able to refresh the model, and what’s the cost of kind of refreshing the models and things like that.
Because the hidden cost in MMM, they tend to come from spending all the money on modeling and all the money on the report deliverable and not enough on embedding operationalizing and monitoring the models ongoing.
And so, that to me is a central challenge. And one that I think as the industry transitions – to some degree, this is an industry transitioning from consulting to SaaS models. And so, we’ve got to start to grapple with those sorts of things.
And you saw the same thing in CRM in late 2006 type approach just before Salesforce really started to take off and tear through a lot of the custom CRMs. You had everybody building customized CRMs. What you were looking for was how much would it cost you to stand one up. Every enterprise was kind of spending 5, 6, $7 million on that.
And what everybody woke up to was, it was much better to license software as a service. And I think what GPT’s waking us up to, I think what the wider market is waking us up to is actually models as a service are probably the next frontier, intelligence as a service is the next frontier for us to start constructing on.
And so, I get pretty excited about that future, but I think the ratios and the way that you operationalize that model is very, very different for clients.
Ellie:
I also think the change management piece really is interesting to me, also, just an observation coming away from the tech a bit and talking about people involved.
Historically, I think marketers have operated in a very siloed way when it comes to MMM solutions or modeling solutions generally. Whereas now, and we were just talking about it before we started recording, we’re seeing a lot more top down.
Instead of building these things bottom up and the celling being the head of marketing, we’re seeing top down CFOs, CEO taking an active interest in this, pressurizing the marketing teams to answer these questions of how marketing is attributable.
That’s flowing through in pitches I’ve seen, it’s flowing through in the way in which these products are embedded. Even though they’re SaaS, you’ve just got different people pushing from different angles, and that’s all helping the change management piece, I think.
Henry:
I think yes and no. So, I think on the yes camp, it’s helping immensely to get more organizational buy-in, to make strategic decisions with these models. I think on the no side is that organizations fail to build data-driven cultures, not because people don’t like data or don’t want to embrace data, but because they discard data in favor of politics.
And so, when that happens, when you get lots of different stakeholders occasionally, on the band side that I’ve seen, you get an injection of politics into an otherwise apolitical process. And I think if you’re in an agency kind of landscape, agencies are brilliant at politics, that’s kind of what they do. And they have an entire department dedicated to it. Account management’s, literally the department of politics.
Ellie:
That’s a bit harsh, but okay, fair enough. Is the department managing the politics.
Henry:
Department of managing the politics.
Ellie:
Whether they’re creating them or not, I hear you.
Henry:
But yeah, it’s the department of managing the politics. But it’s so interesting to me. For an organization like ours, we are probably ill-equipped to deal with that. I’d say we’re lesser equipped to deal with that kind of political environment.
And actually, one of the things that I say to people, there’s actually a dichotomy. If you have a highly politicized environment, we actually may not be the best solution for you. There may actually be better alternatives. There are a number of other great vendors in the Australian market who are actually really well-equipped to deal with that, and who can help navigate that change too.
So, for me, it’s a question of a lot of the times, you have to make a decision when you’re kind of looking at those wide stakeholder groups and go, “Is this because there’s organizational buy-in around building capability to empower and drive marketing? (Then it’s the right client for us). Or is this because there’s a lot of competing silos and actually it needs an external voice to bring them together.” And then it’s probably actually a benefit for somebody else.
Ellie:
Look, this holds so much wraps up in there. When I was talking about protectionism before, this was … politics rules. I think all of what you said is true, and sometimes an external voice can, and sometimes just proof points from a service or a SaaS based model or whatever, it will gain traction over time.
But also, I think … and it’s something that marketers are often accused of is lack of bravery, lack of courage. These organizations, political or not, need a point of arbitration at some point to say, we’re doing this and we’re pushing, and that’s where I see CEOs and CFOs coming into good effect.
Henry:
I agree.
Ellie:
Sort of shaping that, and sort of forcing that change. But I think until … you’re probably right in that some organizations may well just not be ready yet or may never be ready, just the way they’re structured to cope with the way in which you go about it.
But I’m interested in your comment about agencies too. I am interested. Well, it’s one of these questions, I’m dipping all over my list of questions here because the way that we’re meandering, but media agencies are great at politics. I mean, it’s fascinating.
I think you’ve been quite smart in developing partnerships with media agencies. I mean, either white labeled or maybe more explicit in some cases, because clearly, agencies are, or at least have been at one point, direct competitors of yours in this space.
Henry:
I think they still are.
Ellie:
How do you find working with them?
Henry:
So, I think a few things. I think that by and large, our working relationship with media agencies has actually been really good. I think the MMM component of what they’re doing is relatively small, and for the most part is not something that they’re super interested in compared to billings and stuff like that.
But my general feel is that agencies are looking for better infrastructure to help prove their worth. And I think that’s a relatively important aspect to the long-term survival of the media agencies. I think that the media agencies do get unfairly beat up on at times.
So, one of the things I always observe is 99% of people in media agencies passionately care about their clients, want to do the right thing by their clients and are trying to navigate an incredibly complex media landscape, often without data tools and tooling to do that efficiently and effectively.
So, it’s no wonder in that environment, you see a huge amount of agency people overworked purely because they’re having to cope with such high volume of information and people. I think my position to the agencies has always been pretty simple. If a client wants to work with you for MMM, they should do so. And that’s exactly what they should do.
If a client needs the MMM to be independent for whatever reason, because confusing execution and measurement often is a really hard thing for most clients. Most client CFOs won’t accept measurement and execution being the same vendor. Which is a tough-
Ellie:
That’s a common but tough problem.
Henry:
It’s a common but tough problem, and it’s the same reason why if you go to the consulting world, it’s actually illegal for someone to be in audit and business consulting for the same client (illegal). I actually think the same legislation’s probably coming to the media agency sector, I think we’re three to five years away from that.
And so, I think in that world, what you really want to go is, “Well, what do the agencies need from an effective MMM partner?” They need someone who’s not going to bog down their people in data ingestion problems. They need someone who’s not going to be sensitive about sharing model outputs back into their planning tools. And they need someone who’s not going to be sitting there competing with them for consulting head hours.
And I think in that world, a SaaS based model like ours makes a lot of sense because we’re solving the problems in a way that can fit well into their ecosystem without needing to be a disruptive voice. And I think that’s why the model has worked. I will say that I think that the perception of our business being agency-reliant is wrong. Only about 15% of our volume comes through the agencies. 85% is direct.
So, I think what we do a lot of, and what people do see us doing a lot of is we train the agencies because we think it’s much easier to do that at scale. And also, I have a view that actually, if we want the agencies to trust our offering coming in and not be an opposing voice, transparency builds trust.
And so, often, what I say to agency CEOs about training and why we should come in and train the agency is, you’re probably going to end up competing with us at some stage on an MMM project that you want to win for whatever reason.
If we come in and train your agency, you’ll know my exact playbook, my exact product, the exact way I’m going to pitch it. You’re going to know all of these things and so you’ll know how to beat me. That’s your worst-case downside.
And so, if you come at it with that kind of level of transparency, I think too many relationships in media and marketing generally come from a position of frenemies and fear. And I think we’ve just tried to take a very different approach to it. If someone’s going to copy us, they’re going to copy us. Like I just don’t see that as being a real downside.
Ellie:
No. Yeah, it’s a different way of looking at it. And I’m surprised by that, I must admit, I didn’t have the perception at all that most of your business came from agency. I mean, I’m actually kind of surprised it’s as high as 15%.
But what I did see is that, like I say, you’ve sort of made some of these partnerships to enable agencies to deliver those services that they probably can’t scale up for themselves.
Henry:
Yeah. I think that’s a big part of … and also just make it easy to integrate with their ecosystems. I think there’s a number of agencies where typically, the interface between an MMM and an agency is a little bit like ships in the night to some degree, and I think that’s a bad outcome for clients.
I think what you want is you want your MMM vendor to go, no, it’s a priority for us to try to get media outputs and coefficients back into the agency for planning purposes.
Ellie:
I know from my own experience in media agencies, even when the model was being done by the agency, that ships in the night analogy was true. I mean, it was still two separate departments. I think my other observation is that over time, the hierarchy of needs for media agencies in regards to modeling has changed.
I think in the past, the aspirations were much more this is going to be a massive revenue stream for us. This is going to cement our place at the top of the table consultative-wise, and it’s going to make us sticky to our clients. And then it was about proving media worth.
I think your observation about media agencies need to be able to prove themselves now as well from a data perspective, I think that’s higher up their list of priorities now. And I think they’ve tried and found it very, very hard to attract the right talent, to build the right systems, to do what you do as a bespoke player in this field. I think agencies have found that really hard.
And then that’s before they come up against the issues of clients not believing the agencies should be “marking” their own homeworks. Even though I 100% subscribe to the notion that actually, the vast majority of people in agencies want to do the right thing, I do think media agencies get beat up on. I do think media agencies are entitled to make money and exploit contracts and act commercially just like anyone else. It doesn’t make them villains.
Henry:
No, I think media agencies are full of people trying to navigate really complex problems for clients, a complex media ecosystem, and it’s been a thankless job for too long. And I think it’s very easy for MMM vendors and SaaS players to constantly beat up on the agencies.
And I think the position of Mutinex and the way we want to come at it is, even if at this stage we’re not seen as … we were seen as occasionally competitive and things like that, I think long-term position of our company is, how do we be a positive force in that area, because I came from that world of agencies and I have a lot of friends still in it, I believe in it, and I believe in the people’s intent in it. And I think it’s an industry worth protecting.
And I think too often, we kind of are rushing to … I kind of hate the fact that a lot of the trade press and stuff like that is always speculating on the demise of this, the demise of that, and things like that. I think it’s more important for us to actually start to try to talk up our industry, talk up what we are actually contributing to the industry.
It’s very rare that I would see media not being a relatively positive force. Not always a super strong force across a business, but it’s very rare media is losing a company revenue overall. In fact, I’d say I’ve seen it maybe 5% of the time I reckon, if that. But we have this world where everything is kind of framed in the negative.
And I think the way that vendors deal with each other and the way that SaaS industry deals with the agency industry is also framed in the negative. And I think what we need to do is try to start to drive a bit more positivity, try to actually try to find ways to work together for the benefits of our clients.
And if that sometimes makes it easier for someone to switch away from us or switch away from Mutinex, so be it. Principle’s not a principle unless it costs you money.
Ellie:
I hear that. And I love the integrity behind it and the security of it and the confidence of it is great. I do think negative … I mean, there’s always been tall poppy syndrome. I think in our industry there’s always been a bit of negativity. I do think media agencies … I think the craft of media is something that’s not respected enough, frankly.
I do think there are perceptions of the media agencies just being chickens in it with a calculator type of thing, and that extends across … I experienced it myself. It extends across clients and it dents all sorts of things. And it makes people’s lives in media agencies really hard because they are trying to account for really complicated things.
So, I think as long as they’re in … if I had to put a fault at the door of media agencies, is they’re often so defensive about other people being in that turf with them. So, the approach and the philosophy you’ve just described to it, is going to still be an anathema to some agencies, I think.
But I agree that the more they just embrace that kind of, “Come on guys, let’s just work it together and we’ll win some, we’ll lose some, but ultimately, we’re doing the best that we can.”
Henry:
That’s okay. Media agency CEOs, if you’re listening to this, I’m a lover, not a fighter.
Ellie:
Me too. I grew up in media agencies, I’ve got a lot of time for them. So, it’s interesting to have this debate with you.
Let’s change gears a little bit though. I do want to talk about pitching a little bit. So, we’ve worked together on an MMM pitch recently for one of Trinity P3’s clients that I ran, it’s now one of your clients, of course. You won the pitch.
And I’ve run a lot of pitches of various different types of agency, various different types of service. But pitching a service of this nature, for me, was very different to pitching an agency. I made a lot of changes to the back end of what I did or what I do to accommodate these things.
What are you seeing in the pitching market? Is it common? I mean, I’m guessing it’s quite common for you to have to be competitively tender, but what are the pitfalls and watch outs that you’ve seen and the way people go about it.
Henry:
So, not always common for us to tender actually. I’d say covers about 30% of our volume. So, I think where we get a lot of our business is actually referrals from clients. So, clients tend to refer us, the vast majority of our business would come through that.
I think on the tenders, it tends to be mostly related to the intent of why MMM is happening. So, I mean, I’m sure you recall this, but in the pitch, the most important thing in our first conversation for us with that particular client was why is this happening?
Like what’s the kind of underlying driver for what you want to get out of this? Because it’s so easy to focus on models and measurement and all the fancy stuff, but the best way you can help a client is understand why they want measurement in the first place and why they think MMM is a solution to that measurement problem. Because too often, it’s easy to talk about all the features and things like that. It’s very, very hard to try to understand how you might contextualize yourself in a client’s world.
I’m seeing two types: the first is a customer who’s probably relatively new to MMM, and I think this pitch was more on that sort of spectrum. And who wants to actually go on a journey to embed it in their business, and to embed a capability into their businesses and is excited about that.
The second that we often get approach for is someone who’s very experienced in MMM, but perhaps wants to speed it up or get a different layer of granularity or have a different speed to answer and response that they otherwise aren’t getting. And I think both of those categories are equally valid.
And speed is not just how quickly you can refresh a model. It’s actually how quickly you can get an answer out of the model. And I think that’s a key differentiator. People confuse speed with how quickly am I pressuring the model refresh button, but actually, it’s more to do with the speed of answer. And if the answer is out of date, then it’s not a good answer.
And so, those are kind of the two I generally see. I think the RFPs for MMM have gotten much better over the past three years. So, three years ago, it was all around what are the deliverables, if that makes sense. Today, it’s all about the capability, the technology. And in some pitches, all about the model testing and governance.
Model testing and governance is something we would see in 20% of pitches, roughly. I would like to see it in a hundred percent of pitches. I’d strongly recommend even the clients look at getting synthetic data sets that they can use to test a vendor out and see kind of what the vendor comes back with, how they test, how the process runs, and how seamless it is. Because all of that sort of stuff, it matters.
And one of the most sophisticated pitches that I’ve seen was a global tech company who did exactly that. And they actually were very well-organized, very well disciplined. You could tell that they were an excellent client as a result. And so, I see it kind of falling into those two camps. I think the really exciting thing is you’re typically as well getting a lot more people on the pitches.
So, previously, the MMM pitch might have been the media manager and an analytics person. Now, it’s a typically, CMO, CFO, someone from finance and marketing team, brand managers. And that’s really exciting to see how further up the chain MMM has got as a priority, and I think how much further up the chain being financially driven in terms of marketing and being outcome driven in terms of marketing has gotten.
It’s clearly penetrated much higher, which is extremely exciting for a world where I’d love to see more outcome-led marketing.
Ellie:
Oh, I think it’s exciting for marketing for sure. And the pitch that we’re referring to, there was absolutely from the top of the business, a lot of interest in that. They were very much of the first subset you described in terms of being at the start of that journey. I ran a whole dev workshop with them before we even engaged in a formal RFP to talk about some of those questions. Why?
What are the barriers? What are the opportunities and what’s your constituent kit of parts in data terms that you’ve got to actually … how ready are you for this? And based on that, what’s the journey that you are going to need these guys to go on with you?
And that led into some of the things we’ve been talking to before about level of consultancy versus level of SaaS, the complexity of the build of the model and how ready they are to be a part of that. But it was interesting, I mean, no two are the same for sure. It’s always complicated for sure.
I like your thinking around the synthetic data set. So, it is not something that I felt in that particular pitch that they were going to be ready for, but that’s actually probably not the right way to be thinking about it. It needs the validation, right?
Henry:
Yeah, it’s more just, it’s more just testing. I think there’s been such a proliferation of vendors in the space. There’s obviously some clear standouts across the market. And I think with a lot of the small vendors, it’s just very easy to make claims and they just should be verified. Because there’s every chance that a small vendor may have a technological breakthrough. Certainly, we did, there’s no reason that someone else couldn’t.
But I think all that needs to be happening is claims need scrutiny and the best way to scrutinize it is just to test the models.
Ellie:
So, new business is obviously the lifeblood as it always is; pitching, it’s an interesting space to be pitching in. Talk to me about that in context of the U.S. The U.S., from what I read in the press, it’s a growing market for you guys.
I’ll ask you some something more specific. I mean, what are you finding out there obviously is an interesting one, but I don’t know how you would … I mean, clearly, technology plays a role in this, but the U.S. is a massively complex market, I mean, compared with Australia.
Does that figure in your process, does that figure in your thinking? How do you even go about building models in a market of that size and complexity?
Henry:
That’s a good question.
Ellie:
It’s kind of beyond me to … it’s hard enough over here with five states and 25 million people, let alone over there.
Henry:
Well, the first thing you realize about the U.S. is it’s an incredible market in the sense that it’s sophisticated from a data collection and provisioning perspective. So, things are generally in a warehouse and stuff like that. That side of things is more sophisticated.
Ellie:
So, it gives you a head start, is what you’re saying?
Henry:
Correct, yeah. So, our average onboarding times are about half that in the U.S. than Australia. So, which gives you a sense of how well set up they are. Pretty much every single one of the Fortune 100 does MMM. Every single one has an MMM in place. There’s not one that doesn’t. They have varying levels of how those are operationalized, but they all have them. And that’s really exciting.
So, the market is very big, it’s very complex. But I think the principles of the models and our philosophy on models has always been ramp up the statistical power of the model within a given customer. The way that you can do that is through granularity to some degree, and by breaking up the sales targets to a large degree.
And breaking up the sales targets gives you more observations as long as that sales target doesn’t get too sparse. That’s kind of the principle.
And so, I think that on that basis, the U.S. actually works better because you’ve got more products, more targets to break up in more states. So, you can kind of actually superpower the models a little bit more.
Ellie:
It’s quality over quantity is what you’re saying. Like the quantity is huge, but the quality you’ve got to play with is better?
Henry:
Kind of, yeah. Like there’s more volume and remembering in stats, volume to some degree does count for quality, because volume is number of observations.
Ellie:
Of course. The higher the volume, the more-
Henry:
MMM, everyone bangs on about MMM being causal and stuff like that – nonsense, it’s an observational technique. We’re observing time series across the time, and then we build causality by structuring the models in certain ways to reflect real world known dynamics. That’s how an MMM works. But just because you put structure in a model to mirror causality doesn’t make it causal in and of itself.
Now, observational data, in science, what are we kind of taught? The more observations you have, the closer it gets to causality. Now, you’ll never have enough of that in MMM to make a true causal claim. Anyone who says MMM is truly causal is lying. But you will have enough to kind of start to at least make some better observations and have better distribution of those and things like that.
And so, I think the U.S. is a really exciting market for that perspective. I think that all the work that we’ve done around data collection, ingestion, making that really seamless, making the organization of it really seamless, actually plays better for us in a bigger market than a smaller one.
And I think it’s been a huge contributing factor to winning some of the big CPGs, some of the big tech companies that we’ve won. And I think it’ll continue to do that.
That being said, it’ll be completely false of me to say that we haven’t gone there without our challenges. We went in there not fully understanding some of the geo-specific dynamics around sport and stuff like that in particular, and we’ve made some stuff ups there.
So, I wouldn’t necessarily characterize, even though we’re going really, really well from a perception perspective, it hasn’t all been roses over there. We’ve stuffed up some stuff, made some mistakes, learned a lot.
We’re pretty lucky that we’ve been able to kind of course correct on a number of things. And I think the final thing I’d just say is the team that’s great.
Ellie:
You got [crosstalk 00:41:19].
Henry:
So, John’s interest … AKA, so John and I were on the main streets of New York up on 57. So, as I call him, he calls me Hen, Henny, Hen, and he’s [Jas Innocent 00:41:19], so that’s what we call each other in NYC.
Ellie:
You guys are pros, aren’t you? Very cute.
Henry:
But he’s fantastic.
Ellie:
Yeah, that’s fantastic. Okay. Very briefly because I’m conscious of your time, let’s just talk about … you’ve brushed on this already, but your next frontiers – I mean, intelligence as a service is really interesting, but just briefly, what are your big buckets that you’ve got on your horizon?
Henry:
Well, I really want to build a growth co-pilot that can actually help media agencies, clients, anybody in the industry navigate the challenge of growth and the complexity of growth with clarity, so we can actually get back to making decisions.
My fundamental view on the business of media at the moment is that 80% of our time is spent in data administration, and that’s actually what’s killing the industry. It’s a noose around our neck if all of our time is spent in data admin versus value creation for clients.
I would estimate the MMM business to be probably in one way or another there’s 10 billion of rev on it around the world. And then say another 10 billion probably floating around in ancillary services like data prep and stuff like that.
My objective with Mutinex to some degree, and this might sound crazy, is to shrink that industry to be a $2 billion industry, because I think it’s actually too big. It’s too bloated and it’s not adding enough value. Let’s make it a $2 billion, really sharp industry that’s got hyper clarity on delivering value.
And in doing so, we can free up that money, those people, all that thinking time to actually go back to thinking about customer problems. How do we grow these businesses? How do we grow them with clarity and confidence?
And so, if we can build technology that does that as its purpose, I think for me, we’ll have achieved a business that’s meaningful and we’ll actually leave a legacy that’s positive.
Ellie:
Potentially transformational.
Henry:
And so, I think that’s really, really important. And for me, I think probably over the last year, I mean, it’s been pretty well-documented that I probably had my struggles in leadership last year, let’s put it that way. And so-
Ellie:
Everybody does.
Henry:
Well, and to some degree, I think that’s because I wasn’t necessarily … we went through this rocket ship of growth at Mutinex, and it was kind of an imposter syndrome moment. And I wasn’t quite sure of myself last year to many degrees. I wasn’t quite sure of where I wanted to go, whether or not I could lead the business in many ways.
Whether or not I was even right to kind of … whether or not I should have a vision for this business and stuff like that. You go through huge moments of self-doubt, which probably isn’t evident when you’re say stupid things in the press as I did last year, which are a bit too confident and brash – that’s kind of where that comes from. It comes from a place of insecurity more than anything.
And so, I think that one of the things I’m excited about is how do we get back to positive, focused vision, focused on product, focused on how we can bring good back to the world of media marketing for customers, agencies, industry writ large, and stick to those principles.
And I think if we do that, we’ll at least build business as a positive force. And so, that’s what I really want to do. I think that I love the industry. I think that we add a lot of value to businesses. I think we don’t always allow ourselves to shine in the right ways. And I think we need to start to start to build technology that allows our industry to shine brighter.
Ellie:
So, that’s a great place to finish. I really love the ambition, I really love the passion. I think it’s an exciting vision. Most of all, though, I love your honesty. It’s refreshing, frankly. And the courage you have to show your own vulnerability is really … it’s a great thing, Henry. It really is.
It’s great to have that kind of honest perspective. So, that’s been a great discussion. Thoroughly enjoyed it, and thank you so much.
Henry:
No, thank you.
Ellie:
And all the best with Mutinex going forward, your new frontiers.
Henry:
Thank you.