Proof Analytics with Mark Stouse | Ep. 189

Sima is delighted to have Mark Stouse, the Chairman and CEO of Proof Analytics, joining her today to dive into his background, the years he spent building Proof Analytics, and the exciting work he is currently doing.

Mark’s story

Even though Mark understands data science, he is not a data scientist. He is a business leader who got hooked and wanted more after discovering the power of analytics. So he began climbing the hill.

Business problems

It takes a lot of effort to get a small amount of data. So Mark started focusing on business problems associated with analytics. He decided to adopt the role of a whisperer figure and began building that role within all the different global organizations he ran.

Building a team 

He had to spend millions of dollars each year, mostly on salaries, to build a team that could provide the latency needed for the various businesses. But it worked!

A valuable discovery

The huge and intricate mega-models that Mark and his team created were time-consuming, and business leaders could never quite wrap their heads around them. That led Mark and his team to discover the value of what is now known as the Minimum Buyable Model.


Mark and his team wrote some code early on to allow them to federate models in case they needed to make them bigger, more explanatory, or more inclusive.

A better way

It all took a lot of work. So, Mark looked for a better way of doing things. Something that had never been fully automated or aided before was multivariable linear and non-linear regression. Yet, that is a lean-data analytic that still answers most of the world’s questions.

Big data

A problem many businesses are facing today is that they have tons of data, but very little of it is big data.

The problem with time lags

When looking at the impact of marketing on sales, almost everything marketing spends on increasing sales productivity has different time lags associated with each investment. Without knowing the normal time lag for each type of investment, you cannot know how to plan or find an ROI.

Proof analytics

At Proof Analytics, they used focused and targeted AI to assist and accelerate their modeling process. Once in production and linked to the necessary data sets, their models become mostly autonomous, so business owners need not be regularly involved with them. Their models get recalculated as new data gets presented, so people can see the historical reality, forecasts that get generated, and also, an updated reality as the future becomes the present.

Mark’s customers’ favorite part of the tool

Mark’s customers like that they can load data into the libraries and then attach file-level security to that data. That makes it impossible for anyone to see the data in the file, yet the algorithm can still see it.


Mark conducts sessions with his customers to help them triage. That is important because even with everything done with the Proof Analytics tool, it is still not fully democratized. So even though it does not require a data scientist to run it, a data analyst or someone switched onto the topic can help marketing teams have a much better experience with it.

The ideal customer 

The ideal customer for the tool is someone already doing all the work the old-fashioned way and looking for a way to do it more efficiently.


Email me your thoughts



Proof Analytics

Mark Stouse on Twitter

Mark Stouse on LinkedIn