Market Mix Modeling: Why You Are Failing at It and What You Can Do About It

RajulJainWhile market mix models (MMM) might sound quite elementary to our peers in marketing, PR professionals have only recently discovered their value. This sudden interest in this statistical tool can be partly attributed to the need to prove our worth in the marketing mix to clients and partly to the increasing sophistication in PR measurement and analytics that our field is now realizing. As IPR’s Director of Research, Dr. Sarab Kochhar recently noted, “Virtually every aspect of business is now open to data collection and this broad availability of data has led to increasing interest in methods for extracting useful information and knowledge from data.[1]

And yet, I find us still struggling with statistical tools and programs such as MMM. As a Senior Manager of Predictive Analytics at a global PR agency, I conduct several training sessions with account managers and senior executives to increase their understanding of this tool. PR pros across the industry are now being asked for tangible reports and metrics to rationalize budgets while competing against other communication channels. In this post, I summarize some key points from these training sessions.

What is MMM: In simple terms, MMM is a tool that uses sales and marketing data to establish the return on investment (ROI) of each channel in the marketing mix helping organizations develop the most effective spending level for each channel. In other words, these models not only demonstrate the contribution of each communication program but are sophisticated enough to build future projections enabling budget optimization.

Key challenges: Most PR practitioners who have worked with clients on MMM would agree that most often these programs reflect poorly on PR. There are a couple of reasons for why the odds are not in our favor when relying on MMM:

  • Firstly, the entire notion of ROI is something that is often very difficult to demonstrate in PR. Please don’t get me wrong- I am a data person and completely support using these financial metrics as much as possible. But, it is not always possible. Our Key Performance Indicators (KPIs) such as brand equity, reputation, trust and credibility cannot always be represented in those hard terms.
  • Another related issue that works against PR is this term—working versus non-working dollars. In plain terms, all money spent on consumer-facing activities is classified as working dollars (e.g. ad placement, sponsorships), while all money spent on creating output (e.g., pitching, production and editing costs, agency fees, planning, measurement and analytics) is classified as non-working. As you can imagine, our working to non-working dollar ratio does not always match the desired 80:20 rule that applies to other marketing channels (a popular but completely misguided notion). Making matters worse, most MMM do not account for this distinction and club together PR with all other channels in the mix suggesting that PR has the lowest ROI.

As a result, we often find ourselves getting caught in the vicious cycle. We receive a lesser budget in the next fiscal cycle, so we are able to do less, and hence demonstrate less, and thus, get even lesser budget the next time. Sound familiar? If you have ever been or are in this situation, here are the top five questions you should ask when you see the output of a MMM and top five talking points to have a smart and effective discussion with the clients:

Top five questions to ask:

  • What’s the ROI of PR activities?
  • How has ROI changed from previous cycle?
  • What if the budget was $$ and not $?
  • How do sales/impressions move with weekly PR spend?
  • What is the contribution of PR within mass media?

Top five talking points:

  • Explain why the current MMM tends to under credit PR.
  • Demonstrate time series ROI movement of each channel.
  • Present a couple of scenarios with different budget distribution.
  • Ask the client about desired working: non-working spend ratio for PR.
  • Pitch other metrics (e.g. brand equity, reputation, trust) to measure value of PR.

While I have run several MMM for my clients and as a data analytics enthusiast, I personally am not a big advocate of the way in which they are currently being used. It is almost like judging how fast an elephant can climb a tree when that’s not really what it’s meant to do. But, if you do find yourself losing the battle in MMM, hopefully this post has given you some insights to turn the wheels around.

Rajul Jain, Ph.D., is an assistant professor at DePaul University and a senior manager at Ketchum Global Research and Analytics. Follow her on Twitter @talktorajul.

[1] http://cw.iabc.com/2015/07/08/actionable-research-depends-on-asking-the-right-questions-sarab-kochhar/

Heidy Modarelli handles Growth & Marketing for IPR. She has previously written for Entrepreneur, TechCrunch, The Next Web, and VentureBeat.
Follow on Twitter

3 thoughts on “Top Five Questions to Ask About Mix Market Modeling

  1. Thank you, Dr. Jain,. I have a better understanding of your permission with the benefit of your response. I agree…the modelers want to reduce the amount of sales attributed to “other” so they are eager to include what they can, including PR. Unfortunately, they don’t know enough of how PR differs from everything they currently use. The key is to break down their resistance to what is only “semi-controllable” and “uncontrollable.” Advertising, as we know, is controllable because it is paid. PR is only semi-controllable (at best) and requires the model to factor “negative” variables like critical news coverage or bad reviews on social. In most cases where we cooperate with the client’s modelers — whether in-house or outsourced– we begin with a simple discussion of how PR is different than anything else in the mix. That sets an important foundation on which to build.

    Best wishes,

    Mark

  2. Thank you for your response, Mark. I agree, MMM can be effective tools in demonstrating the value of PR in the marketing mix, if they are set up in the right manner. In my experience, most modelers equate ROI to aggregate return on each channel, which sort of diminishes the value of PR compared to other programs that receive relatively larger chunk of the marketing budget. As you mentioned here, ROI could imply different things and some of them are not really PR-friendly. I definitely vouch for PR to be represented in MMM so we can evaluate and hence prove our worth but in a way that does not disadvantage us.

    Thank you for sharing the link to the article. I have read it and have also included it in the reading list for my data analytics class that I offer at DePaul. It is an excellent piece that supports what I want to bring out in this post: MMM are great tools when they are interpreted correctly. But I also believe, value of PR should be demonstrated in more ways than just MMM. At Ketchum, we were able to turn the wheels in our favor once we started injecting the right inputs to modelers and began asking the right questions. We were also able to complement MMM with other KPIs that are more in sync with what we do and accomplish for our clients (e.g. reputation studies). Hopefully, this post adds to that discussion and helps PR pros have a meaningful dialogue with clients around this topic.

    I would love to connect with you and talk more about how PR can benefit from MMM.

  3. Dr. Jain,

    Thank you for sharing your insights and experiences with Marketing Mix Modeling. As client organizations become more and more familiar with the ways by which data analytics guide marketing investment decision-making, PR is either in the mix or we’re out. While it’s fair and accurate to suggest that not all PR is marketing and that not all PR is media, a lot of it is and if PR is not included in the marketing mix model, “out” means “out:” Fewer resources. Less recognition. Less tangibility in connecting PR with the business (at least in the minds of many marketing communications executives within organizations that use modeling).

    I must say that my experiences representing PR in marketing mix models seems to be very different than yours. In this IPR paper–co-authored by several of your colleagues at Ketchum–we offered examples of how PR delivers the BEST return-on-investment in marketing. Let me make an important distinction here: PR delivers the best ROI because of the low out-of-pocket expense and the slow decay rate of PR. “ROI” in this case is “sales generated by each dollar invested.” PR rarely drives the highest volume of sales but we almost always drive the best ratio. Maybe I’m misinterpreting your definition of “ROI” but “efficiency” is a factor in ROI calculations.

    The IPR paper is entitled ISOLATING THE EFFECTS OF MEDIA-BASED PUBLIC RELATIONS ON SALES: OPTIMIZATION THROUGH MARKETING MIX MODELING I offer IPR readers a link to the paper here:
    http://bit.ly/1Lyeotf

    I’d love to hear from other PR practitioners with Marketing Mix Modeling experience. And I’d love to connect with you, Dr. Jain, to share our experiences with one another.

Leave a Reply