Five Channels

Destin, United States
Founded: 2007
Staff: 9
Online Marketing
Web Design
Search Engine Marketing
Social Media
Advertising & Media
Marketing Consultants

and 5 more

Sector Experience
Business to Business
Food & Drink
Professional Services

and 14 more


Interview: Why Retailers May Benefit First from Machine Learning and Marketing

29 August 2017 21:06pm

To gain clarity on the applications of AI in marketing, I recently flew over to San Diego and sat down with Daniel Faggella.

Daniel is founder of TechEmergence, the only market research platform focused exclusively on the applications of AI in business. I asked Daniel to answer questions that are relevant to other marketing leaders.

Jason Hall: What is the state of machine learning for marketing today?

Daniel Faggella: Machine learning has garnered a huge influx of attention in the last two years, and the increased attention has been both good and bad for marketing leaders.

On the one hand, machine learning has left the back-office of marketing giants like Facebook, Google, and Amazon, and has entered the normal business lexicon. By entering the popular tech lexicon, more marketers are aware of what machine learning can do, and more entrepreneurs and developers are working on it’s marketing applications.

On the other hand, machine learning is a vague topic - plagued with hyped expectations. This means that marketing leaders feel that they “should be doing machine learning,” when often they don’t have the budget, talent, or proper business model. It’s difficult to discern which businesses are poised to leverage machine learning - and which businesses needn’t concern themselves with it yet.”

I believe that eCommerce businesses may be the first to benefit from AI applications in marketing (when compared to other businesses). The factors laid out in the interviews and surveys with over 50 executives (mostly founders) from AI-in-MarTech companies are presented through the research conducted by TechEmergence.

Jason Hall: Why are the high transaction volumes in eCommerce and retail businesses so helpful for applying AI in marketing?

Daniel Faggella: Machine learning isn’t valuable in-and-of itself. It’s valuable only to the degree to which it allows businesses to meet their goals better than they could through other means.

One of the potential business benefits of machine learning is its ability to detect patterns in huge volumes of data, allowing machine learning systems to perform functions that they couldn’t otherwise perform. Applications include:

  • Find new combinations of demographic audiences and ad creative for PPC advertising
  • Determine ideal upsells or email offers for customers who buy a specific item online or in a store
  • Determine the best image creatives for banner or social media promotions based on thousands of split-tested variations across customer segments
  • More...

In order to find these patterns, machine learning systems must be presented with a constant stream of new data - and generally - the more the better.

Retailers (online or offline) have the advantage of having high transaction volumes. While a large consulting company or B2B software company may make $100MM per year by making a few dozen big sales, a B2C retailer of the same size generates hundreds of thousands of transactions per year. While there are legitimate use-cases for B2B AI marketing (some of which have been covered on MarTechToday and other sources), they aren’t nearly as plentiful as in the B2C world.

If a machine learning system is being trained to optimize the factors in generating a sale (or generating a sale with a high monetary value, or “cart value”), then more sales means more chances of detecting the underlying patterns that led up to that sale.

Jason Hall: What makes a “digital customer journey” essential to making AI work - and why is this more prevalent in retail?”

Daniel Faggella: While eCommerce businesses have a notable advantage over brick-and-mortar retail companies in the race to digitization, big physical retailers (such as WalMart, Home Depot, Starbucks, Ulta Beauty, and others) are increasingly collecting digital “touchpoints” along the buyer’s journey.

Because machine learning needs to extract its insights from data, companies that track the most data throughout a customer’s journey (from visitor or lead to customer, and from initial customer to lifetime value) are best prepared to leverage machine learning to coax out new insights and better target their offers and marketing efforts. Physical retailers with loyalty programs (Macy’s, Zappos, Costco, and others) have been collecting this data for years.

TechEmergence interviewed 51 executives of AI marketing companies, some of their first two questions were: ‘What industry sector are you trying to sell into most?’ and ‘What sectors are most poised to benefit financially from AI in marketing in the next 5 years?’ In both cases, eCommerce and retail ranked vastly above other industry categories on average.

Our executive panel put "retail / eCommerce" significantly ahead of healthcare, finance, manufacturing, IT services, and others in terms of near-term ROI for AI marketing tech (all in all they ranked eCommerce higher than brick-and-mortar retail, but both ranked among the areas of greatest opportunity). Digital customer journeys were listed as one of the most important factors in this AI opportunity in retail.

Jason Hall: Do you have any concluding thoughts on what business leaders should be considering today?

Daniel Faggella: All in all, I believe that most businesses (retailers included) shouldn’t be hell-bent on introducing AI into their marketing processes just for the sake of being “cutting edge.” This current phase of early adoption implies that most AI marketing vendors are still “piloting” their products on customers, and the most valuable, succinct AI marketing applications have yet to be fleshed out in a way that makes them available to most businesses.

In addition, small and mid-sized businesses generally may have a low likelihood of profiting fruitfully from AI in marketing within the next two or three years. Even for retail businesses (where opportunity seems to be most abundant), time is still required to tell which AI use cases will become valuable and lead to strong ROI. Smart managers in the retail sector should keep a close eye on the largest and most tech-savvy retailers to follow the innovations that seem to be making a clear impact.


Machine learning
retail marketing
Artificial Intelligence
Machine learning
retail marketing
Artificial Intelligence
Machine learning
retail marketing
Artificial Intelligence