machine learning and marketing tech, surveying over 50 founders or executives in AI marketing companies. In the article that follows, we break down our major findings and what they might mean for martech at large.
When Machine Learning in Marketing Becomes MainstreamWhile understanding specific marketing applications for AI was the main point of our survey, we also made sure to ask executives about timeframes for the technology’s development and adoption. We asked “By what year would you suppose that marketing applications (even for small businesses) would almost ubiquitously involve some degree of machine learning or AI?” This question was open ended, and respondents could answer with any date they like. Curiously, roughly one third of respondents converged on “2020” (17 responses) as their date of choice. Next most popular was 2022 (8 responses) and 2025 (6 responses). While we would normally expect executives and founders to be “bullish” about adoption of their technologies (many of these founders spend their days convincing investors and early-adopting customers that their technology is the wave of the future), we still weren’t expecting such an aggressive estimated timeline for adoption. While this optimism might not imply that all marketing tech will involve AI, it pose interesting questions about how “normal” AI functionality will be in the years ahead. As it turns out, different applications and different types of businesses might be more poised to reap these benefits in the near term – which takes us to the other aspects of the research.
What AI Will Be Used for in MarketingIn addition to asking companies what their company is doing to apply AI to marketing, we also asked them where they believe the profit potential will lie in the coming five years. Not surprisingly, there was a lot of overlap between the areas of focus of the companies in our survey, and the areas where they see profit potential. While we take this inherent bias with a grain of salt, it seems reasonable that companies raising millions of dollars for their technology have chosen to focus on application areas that will genuinely be useful to their market. This dependency might lead us to trust the sentiments in this section of the survey more openly. The most popular responses for five-year profit potential was “Recommendation / Personalization” (20% of total responses), followed by “Customer Segmentation / Targeting” (18%) and “Programmatic Advertising” (14%). The emphasis on one-to-one marketing (as opposed to “shotgun” marketing) seems to only be fueled by quantifiable marketing channels and attribution tools – facilitated (in large part) by the great shift from print to digital advertising. With machine learning in the mix, patterns can be recognized from this flow of data in order to determine new and distinct customer segments, determine the ideal timing of marketing messages, determine the highest converting “up-sells” and “cross-sells” for a specific user, and more. It’s interesting to note that all three of the top three answers involve technologies to help dial in marketing messages to the appropriate audience. “Programmatic advertising” is a contention term that refers roughly to the process of buying ads based on certain criterion from an advertising platform, rather than through an agent or agency. The proliferation of programmatic advertising (on platforms like Google AdWords, Facebook, and others) is aided in part by the robust ability to target individual users and specific behaviors or demographics. There seems to be a consensus that this domain is where AI can do a better job than humans – and there it’s ROI will be proven on the battlefield of business in the years to come. The 4th, 5th, and 6th responses (“Content Generation” 10%, “Search” 8%, “Decision Support” and “Forecasting” 4%) didn’t have anything to do with customer targeting. It is interesting to note that many of the companies who filled out our survey (~26%) considered “analytics” to be their core value proposition, while the top 52% of responses to our question about high ROI applications in five years involved applications having to do with targeting. This may mean that some of the companies surveyed see their tool as mostly useful for analytics now, but that they believe targeting will be the true future ROI as data becomes more available and “kinks” are worked out in building these tools (many of the companies surveyed are less than five years old).
The Companies and Industries Best Poised to Leverage AI in MarketingIn our interviews and research, we haven’t able to find strong evidence about which individual industries will be poised to benefit most from AI in marketing. However, there does seem to be strong evidence as to which types of businesses will benefit most, and it doesn’t seem to bode well for service businesses or much of the B2B world. Over and over we saw a regurgitation of the same “themes” among businesses poised to benefit from AI in marketing:
- Quantifying customer touch points (digital publishers and digital advertisers have the easiest time here)
- Volume of campaign results and volume of sales data (B2C companies with short sales cycles have a supreme advantage here, particularly if they sell primarily online)
- Integrated marketing and CRM systems that allow for robust tracking and reporting (data needs to be accessible to marketing, data science, and executive teams)
Last Updated on