Artificial intelligence solutions in marketing and other industries often have a problem:
It is sometimes hard, or impossible, to explain why AI systems make the recommendations or predictions they do.
It makes sense, if you think about it. Machine learning algorithms might use hundreds, thousands, or millions of factors to arrive at a single recommendation or prediction. It can be difficult to untangle just how the machine got to the outcome it did, even if that outcome seems ideal.
But this poses problems for businesses. What happens if your amazing AI tool makes what looks like a great recommendation—but that recommendation ends up being bad for business? How do you explain to your boss, CEO, or the board why that particular recommendation over another was made?
It's a problem that AI-powered platform simMachines tries to solve with its explainable AI models. The models are used by marketers to do everything from customer lifecycle modeling to lead prioritization to A/B testing and measurement.
Just as important as their results, the models tell you the "why" behind every prediction and recommendation they make. We spoke with simMachines CMO Dave Irwin to learn more about the solution.
In a single sentence or statement, describe simMachines.
simMachines is an explainable AI machine learning software company that provides a workbench for enabling data scientists and non-data scientists alike to develop explainable models that match or outperform any other machine learning algorithm, while providing the "why" behind every prediction.
How does simMachines use artificial intelligence (i.e. machine learning, natural language generation, natural language processing, deep learning, etc.)?
Our company leverages similarity-based methods combined with metric learning to provide machine learning algorithms and dynamic predictive segmentation that weights the features for every prediction in order of importance.
What do you see as the limitations of artificial intelligence as it exists today?
The lack of explainability or the "black box" problem is a severe limitation for use cases where explainability is critical.
What do you see as the future potential of artificial intelligence in marketing and sales?
Explainable AI brings tremendous efficiency to reducing the hours and cost associated with modeling, segmentation, and analysis, as well as greater precision, speed to insight and relevancy as the machine reveals greater depth with regard to the drivers of predicted customer behavior for campaigns, measurement or real-time applications.
Who are your ideal customers in terms of company size and industries?
simMachines supports large scale brands in retail, financial services, and media, as well as supporting marketing cloud providers, data processors, agencies, consulting organizations, and ad tech platforms.
What are the primary use cases of simMachines for marketers and sales professionals?
Use cases include customer lifecycle modeling (e.g. acquisition, churn, cross sell-sell), lead prioritization, dynamic predictive customer segmentation, A/B testing and measurement, trending and analysis, identity resolution, and e-commerce fraud detection.
What makes your company different than competing solutions?
We provide explainability at a local prediction level. Every individual prediction contains its own weighted factors in order of importance and nearest neighbor objects for rich analysis, segmentation, personalization, decisioning, compliance, and auditability at speed and scale.
Any other thoughts on AI in marketing, or advice for marketers who are just starting to explore the possibilities of AI?
Invest in piloting workbenches for model building and segmentation for understanding the business impact on efficiency and performance. These tools are designed to make it easy for end users to work quickly and efficiently using the latest in AI technology. Expect to use several software tools for different applications and the sooner you adopt and learn, the greater your ability to react to competitive pressures and meet customer expectations.
Paul Roetzer is founder and CEO of Marketing AI Institute. He is the author of Marketing Artificial Intelligence (Matt Holt Books, 2022) The Marketing Performance Blueprint (Wiley, 2014) and The Marketing Agency Blueprint (Wiley, 2012); and creator of the Marketing AI Conference (MAICON).