5 Things You Know About Marketing Mix Modeling and Attribution That Are Actually Wrong
By Michael Bertie, SVP Analytic Consulting
I am a bit of a pop-culture nut. Always have been, always will be. One of my all-time favorite bits was David Letterman’s Top 10 Lists (Editor: If you have some free time, enjoy the rabbit hole that is Letterman’s YouTube channel). Consistently funny, often topical, and usually inclusive of some truth, the Top 10 List was a must-see segment for me back in the day.
What does this have to do with Marketing Analytics you may ask? Simply that it dawned on me that a Top 10-style list (or Top 5 be that it may) might be the perfect vehicle to address the myriad of misunderstandings or flat-out wrong perceptions that still exist today regarding MMM and Attribution. There are many viewpoints out there regarding the viability and applicability of advanced analytics so, like the OG Letterman lists, let me set the record straight.
With that, ladies and gentlemen, live from the Ipsos MMA office in Norwalk, CT, here is today’s Top 5 List:
#1 MMM is only good for measuring the short-term
For a long time, this was the #1 argument against Marketing Mix Modeling. “Sure, MMM can shed light on what happened last quarter or last year, but I really need help in understanding the long-term.” Truth-be-told, we debunked that myth long ago. In fact, measuring and predicting the impact of Marketing investment over the course of 1, 2, or even 3 years is a critical insight we share to virtually all Ipsos MMA clients today. Balancing short-term and long-term decision-making is a must for all industries today.
#2 All of our data is perfect so delivery should be fast
Beauty is in the eye of the beholder – not unlike parents cradling their newborn, Marketers often gaze lovingly at their data with a sense of wonder, simply overwhelmed that anything could be so beautiful. IT teams – like delivery-room doctors and nurses – know that the reality is often much squishier and needs lot of TLC. Fortunately, expert help can help smooth the way but it often takes time and patience. Expectations need to be managed, roadmaps need to be developed, and easy wins need to be celebrated before your baby bird is ready to fly the coop. But once its ready, get ready to soar.
#3 If Attribution is not using individual data, then its worthless
We all know that the Marketing data landscape continues to evolve and change, seemingly daily. We have been talking about this reality with clients for years now. Today, more than ever, these tectonic shifts in data tracking have major implications on the measurement of Marketing investments. Nowhere is this more prevalent than individual-level data. What was once considered a must-have for attribution is just as often now considered a roadblock, due to availability, cost, complexity and/or lack of resources.Enter Agile Attribution, Ipsos MMA’s approach to granular attribution, leveraging daily placement-level data, straight from 3rd party platforms and providing insight into execution across publishers, audiences, buy-types and beyond. Customized to align with your media taxonomy, Agile Attribution is the key to bridging the gap between data challenges, detailed measurement and driving measurable value from your marketing
#4 We can do this in-house, or we can just use a 3rd party tool – we don’t need consulting support
I am not saying that you *can’t* build your own internal MMM or Attribution model. It has been done and I am sure it will be done again. However, there is a reason that we here at Ipsos MMA do not manufacture and sell widgets – it is not our area of expertise and we would not be well-served to try to replicate the efforts of companies that have building their own widgets for almost 40 years.Internalization – be it a homegrown solution or an off-the-shelf open-source alternative – has its place but the reality is that we come across very few clients that have the resources, time, skills, and patience to implement an effective solution that meets their needs. Moreover, when they do, the results run the risk as being viewed as departmentally or internally biased toward organizational objectives. The reality is that successful Marketing measurement programs require more than data and modeling – bringing the insights and recommendations to life in a way that accurately measures incrementality and is supported by consulting and change management is key, yet it is too-often overlooked.
#5 MMM and Attribution are too expensive
Budgets are always under pressure. Do more, faster, with less. It is the constant refrain that Marketers hear, day-in and day-out. With that, every dollar counts so focus on building programs that drive measurable value to your organization and focus on the Total Cost of Ownership when alternatives are raised. Moreover, the cost of a well implemented MMM and Attribution is massively overshadowed by the measurable value in terms of incremental sales and brand equity achieved through the optimization and planning processes.Talk to others who have successfully traveled this road before – in fact, we are happy to connect you to our client network. Anything worth doing is only considered expensive when it does not deliver value but when successfully deployed by an industry leader, Marketing measurement is always ROI positive.
So, there you have it – tonight’s Top 5 list. There are many other fallacies floating around about Marketing Measurement – we are happy to chat any time and to answer any questions you may have. Until then, can I interest you in some Stupid Pet Tricks?