What Good AI Should Do For Your Business

by | Oct 1, 2019

How workflow focused automation can deliver a bigger bang sci-fi AI

As science fiction has moved inexorably closer to actual science we’ve been promised that artificial intelligence will change the world in all sorts of profound ways. Various outlets have predicted that algorithms will cure cancer, drive cars, and unravel the labor-based economy. All of these ideas are very real, but they also cast a pretty utopian light on what AI can do for your business and your marketing organization. AI has incredible potential, but most of that potential isn’t in the magical, it’s in the practical.

The end result of employing AI can be the kind of transformational technologies described above, but the real aim of AI, especially the AI for marketing that we build at Simon, is to simplify workflows. If our AI does its job, it will help marketers do the things that they already do faster, more efficiently, and at a larger scale than would be possible for a human being. Our goal is to reduce the total cognitive load associated with major tasks so that marketers can refocus their energy and resources on creative and strategic aspects of their work. By contrast, many vendors offer intractably vague but supposedly AI-driven solutions that make big promises only to underdeliver on their transformational claims. Perhaps the simplest way to understand these two competing is to contrast them.

Many vendors offer intractably vague models for their AI-powered solutions. These tools claim to tackle ambitious challenges in business and marketing, but too often they fail to live up to the high bar of expectation they set for clients. For instance, there’s no doubt that the ability to accurately predict customer lifetime value at the individual level would be a tremendous asset to any business. However, the reality on the ground falls far short for most clients.

Estimating the average LTV of a consumer cohort produces some viable results that can help a brand to make key messaging and investment decisions. However, anticipating with any degree of accuracy the lifetime value of an individual customer simply involves too many unknown variables. These predictions appeal to marketers who would love to make targeted one-to-one decisions, but the reality is that without more information about the individual consumer than any brand could possibly gather, it’s impossible to tell if they’ll make four new purchases in their lifetime, or none at all, at least with any degree of actionable certainty.

One the other hand, AI does lend itself to substantial practical solutions to business challenges, and employing it creatively offers the ability to dramatically scale valuable programs and campaigns that would otherwise be unmanageable or too resource heavy. This could include AI that helps to reduce the cognitive workload by helping to send a message to specific sets of customers, or accomplish some other specific set of actions based on a specific set of triggers and signals. While somewhat less glamorous than predicting the future, this is the type of AI that can help to transform marketing organizations.

Good AI for marketing uses data science to simplify workflows. But what does that look like in practice? AI-solution that actually simplifies workflow will answer at least one of these specific questions: Who should I message? When should I message them? What channel should I use to reach them? and What should I say? By sorting through large numbers of inputs, a good AI-solution can streamline and scale some of the critical decisions that marketers need to make in order to best serve their audience.

A task-based AI helps to solve a discrete problem or complete a task, saving marketers labor and improving scale. Because it’s driven by an existing workflow a good AI can make connections and deliver messages that add value rather than simply making connections based on a series of assumptions?

At Simon we believe that the positive potential of AI is limitless, but we also believe that realizing that potential requires us to apply AI to tangible problems and existing processes and workflows. AI has huge potential to augment and enhance the work of marketers but first we need to separate the good AI from the bad.

To learn more about Simon Data’s AI-powered solutions, visit our platform solutions page here.

Ready to get started? Get in touch for a demo