Proof Analytics delivers a marketing mix modeling (MMM) analytics solution. It allows companies to understand the relative impact and ROI of their marketing initiatives, and optimize accordingly, correct?
Yes. MMM has been around for maybe 30 years as a marketing thing, and it’s been a big deal with Fortune 500 companies for about that long. It’s nothing more than applied regression analytics, the same math used in most areas of science, including things like pandemics or climate change or anything that involves a large network effect involving many factors.
You offer automated MMM. Why is automation important?
The human brain can’t handle more than about three or four variables at any time, and if there’s any lag, that complicates matters even more. Traditional MMM, done by human data science teams, has been popular, but it has operational problems. It’s very expensive, very hard to scale. The human teams are made up of very scarce talent, and because it’s based in human-powered analytics, it’s very slow.
It could take many months to get the insights, and you’re trying to use the insights to change your future, to do something different. That becomes very problematic when the future has already happened by the time that the insights arrive.
By automating it, we killed all those problems. We offer a real-time solution. Every time new data comes in, Proof Analytics re-computes and shows you what’s going on. It’s month to minutes, quite literally.
In terms of cost, we’re talking about an order of magnitude less than traditional MMM. That allows us, most of the time, to deliver Proof Analytics as a managed service.
Do agencies have access to these insights, as well as your client companies?
Yes, the solution allows agencies and clients to collaborate safely and securely in what we call the Proof Exchange.
Of course, it’s not just marketing that has a measurable impact on business success. How does something like communications enter the MMM picture?
We have a client right now that’s embarking on a very ambitious approach. They want to look at how everything they are doing across the company is impacting the overall customer experience. This is not just marketing; it’s communications; it’s product; it’s service and support, you name it.
All of these things have impacts on customer beliefs and behavior, how they think about the company, and whether they trust it. All of this manifests at different rates over time. Time is the important factor. If you’re just looking at what marketing or communications does for you over 30 or 90 days, you are missing out on a huge amount of value. We had a client look at ROI on three different investments over 90 days, and it all looked the same. But when you enlarged the time frame to 120 or 180 days, two investments had dramatically better ROI than the third.
In the case of great communications, there is absolutely a short-term benefit in terms of awareness, the equivalent of the top of the funnel for marketing. But the primary benefit of great PR, great analyst relations and great employee communications is building trust and confidence. It’s not a coincidence that Edelman’s index is about trust. Building trust takes time, and the benefit is not going to be immediate; it’s going to be lagged.
ROI on communications, then, is something quite tangible in your view?
Yes. If I was going to put $10 million into communications or marketing, I would skew it hard today towards comms. And yet the communications industry, because they’re afraid of analytics and measurement for the most part, is really walking away from their greatest chance of glory. They’re ceding the field to marketing.
Confidence in a company is consistently related to bigger deals. The correlation between trust, and how fast someone makes the decision to buy, is off the charts. How does great communications monetize for the business? That’s how it happens: by making sales, or other parts of the business, vastly more effective than they otherwise would be.
Is your service dependent on analyzing the kinds of large data sets associated with enterprise-level companies?
Actually, this is not a big data problem, and doesn’t require a big data solution. This is a small data problem. This is one reason you don’t need AI to solve this problem; you need machine learning to identify patterns, but that’s about it.
Just like a scientist studying climate change, for example, you start with a hypothesis, expressed in a series of questions. For each of those questions, you have to create an algorithmic model to ingest relevant data and give you feedback. That’s exactly what’s going on here. You can’t determine whether you have good data, or enough data, in the absence of analytics. It’s important to have the data, but success here isn’t data-driven, it’s analytics-led.