Why Automated Marketing Mix Modeling?

Why should companies use Automated Marketing Mixed Modelling (AMMM)? What differs AMMM from MMM? Christopher Engman at Proof analytics explains the gains for companies using automation to increase ROI

In the marketplace most medium-sized to large B2B and B2C corporations are struggling with knowing how to allocate money between marketing, PR and sales. And going down one level – how to allocate money between all marketing activities available like retargeting, content syndication, webinars, trade shows, email marketing etc. When trying to figure out how to spend your money most people within companies are still relying too much on their brain. Science is showing that we can handle up to four variables. This is shown by Kahneman and a few other scientists. The human brain is good at associating, we’re good with memory. We’re not good with many dimensions and marketing and sales often have 15 to 50 different dimensions or variables, which is just impossible for the human brain and on top of that it includes time lags. There’s a big difference in time lag between different activities, it can be immediate or it can be a few hours. The Epic Split with Van Damme in the Volvo Trucks movie had several years of time lag.

What different analytics methods are there?

The different type of methods available in the market are first of all, which is the most common still, that we look in various systems like Facebook, Google Analytics, Marketo etc and look at their outputs. That’s data, but we’re still using the human brain trying to figure out what conclusions to draw from it and that’s not working. Another method very commonly used in the market is visualization, where you visualize the data for marketing PR and sales. It looks great but it doesn’t show how they’re driving each other and you can’t see what’s happening because there are too many variables.

Then there is the category Multi-Touch Attribution where you got a tool like Visual IQ, Google, Cube, there are many of them. In these tools you connect as many online channels as possible into one flow to see what’s happening on your own site. You’re also seeing click-through is coming from display ads, social ads and search ads. What you’re not seeing is anything offline like, PR or sales people. You also don’t see things that are happening on third-party channels so a big part of the customer journey is not seen in a multi-touch attribution solution. Companies that are using multi-touch attribution still are ending up overspending on things that have a very clear and immediate call to action with a very fast conversion. Anything that has a delayed effect is lost, anything that is offline is lost so you end up basing a decision a bit like looking at the world through a straw. Then you’re trying to figure out what’s driving what with time lags combining offline and online, using assets and not just your own but also other assets. That’s marketing mix modeling. Some people call it go to market mix modeling when you include sales and PR as well, or marketing mix modeling where you look at the whole marketing mix.

What is marketing mix modeling?

Marketing mix modeling is based on multivariable regression analytics where you produce a few outputs like the slope or the multiplier. The issues with traditional marketing mix modeling is that it is based on consultants that are typically PhDs or higher within mathematical statistics. It takes several weeks to months to produce the models. So they’re expensive, they’re slow and require a very limited talent pool in the world.

What has revolutionized this is Automated Marketing Mix Modeling where you in minutes produce the same outputs as you do in weeks or months and at a fraction of the cost. All of a sudden the big corporations can start to use it everywhere and every month, not just once or twice per year. Medium-sized companies can start to use it for the first time because it hasn’t been viable, due to the costs. It’s been out of reach but now it’s within reach.

Why Proof Analytics?

We are the first automated marketing mix modeling solution in the world. We have a team spread out over North America, Europe and Asia. So we’re a global company already. We work with some of the biggest brands in the world but also a lot of mid-sized companies. We deliver our results both on an annual and biannual basis but also on a monthly basis which acts like a live GPS with traffic information. Our analytical work is led by Magnus Söderberg who is a professor and his team of PhDs. So our analysis is done within minutes instead of weeks and months which is great. When companies use Proof you can start to do other analytics as well trying to figure out what’s driving awareness, what’s driving our site traffic, not just what’s driving our revenue. You can start to learn very fast inside your organization what’s driving what.