Attention CMOs: Get Technical or Die
"Lately there’s been a lot of talk about the power struggles between the teams traditionally responsible for information technology and those responsible for marketing. Heidrick & Struggles sums up the collision of these two functions very well."
70% of email marketers don’t have time to think of subject lines, and only 5% use advanced analysis techniques"
Lately there’s been a lot of talk about the power struggles between the teams traditionally responsible for information technology and those responsible for marketing. Heidrick & Struggles sums up the collision of these two functions very well:
“Yesterday, CMOs and CIOs had little in common except places at the executive leadership table. Today, they are being driven together by the proliferation of technology platforms, the torrent of big data, and the almost limitless choices and power customers now enjoy. Tomorrow, the two roles could converge in a single position.”
CMOs are grounded in a rich strategic understanding of market dynamics, the positioning and role of a brand, and leveraging engagement across all customer touchpoints — and this offers a big advantage over the CIO. On a related note, in 2012 Gartner predicted that by 2017 the CMO will spend more on IT than the CIO. But to really get there, CMOs must meet the CIO head-on, by becoming more data-driven and knowledgeable about data science and enabling technologies.
As a CMO myself, I don’t believe that we must become data scientists. However, we need to know the potential of and how to use data-driven technologies to succeed. Predictive analytics, machine learning, and automated personalization and optimization are found in all of the most exciting marketing technologies of 2015.
In fact, these are also what’s making email marketing sexy again — and not only for companies with lots of money to procure advanced email marketing platforms and the big teams to keep them running. The advent of “plug-n-play” add-ons for the most common email platforms is bringing technologies based on data science to mainstream email marketers — with minimal need for involvement from their technical colleagues.
For example, subject line optimization can be the bane of an email marketer’s day. According to Phrasee, 70% of email marketers don’t have time to think of subject lines, and only 5% use advanced analysis techniques. One may argue that there is no excuse for this critical task to be neglected by any email marketer, especially with the plethora of articles and data on what works and what doesn’t. However, dig a little deeper and what you find is that most experts say the impact of subject lines varies for every business, and to figure out what’s best for yours, you must test and test again.
Enter predictive analytics. Why guess and test when a machine can capture infinite amounts of data about your customers and campaigns, run that through an intelligent language database, and produce better subject line variants for you? It may sound futuristic, but it’s already a reality. What’s more is that after a campaign, that data goes back into the model, and machine learning makes every subsequent subject line recommendation better and better.
Sounds good, right? But what happens when you as the CMO need to face down your CIO or present to your CEO a convincing argument for why this technology is an important investment? You’ll need to have some understanding about why the predictive model of an application like Phrasee is better than another subject line-testing application out there.
Another challenging topic email marketers face is send-time optimization. Many email platforms offer segment-based options that target more ideal send times, but individual behaviors and changes over time are ignored. New technologies like AudiencePoint aim to improve upon this by optimizing send time at an individual subscriber level. However, for marketers to reach this cutting edge, they need to know why other methods of send-time optimization are inferior to those powered by predictive analytics. What data is used to produce the individual profiles? How does the algorithm weigh different behaviors? How does this level of personalization impact ROI? Email marketers should feel comfortable with the answers to all of these questions.
I’d like to mention one final example related to this, since it’s near and dear to what I do every day: personalization of dynamic content in emails. Products, articles, special offers, and other content can all achieve a significant uplift in email campaign performance when customers are targeted more precisely than with traditional methods (i.e. predictive one-to-one personalization is better than segmentation, even microsegmentation).
However, one of the challenges we run into is that email marketers and their CMOs have a hard time shifting from a rules-based to a predictive mindset. They are used to fine-tuning the criteria for content targeting and creating all the variations that go with the different conditions they have created. It’s a struggle for many to understand that our technology is capable of automatically creating an infinite number of personalized recommendations for every individual, simply based on what’s available on a website.
Because they aren’t fluent enough in how these advanced technologies work, marketers will ask for dials and levers to adjust the recommendations model — not realizing that this is the worst thing they can do to a recommender system that is self-optimizing. They cannot imagine that the machine knows better than they do, even though the data shows over and over that it does. Email marketers who take a bit of time to learn the basics of predictive analytics and machine learning can more easily trust such a system enough to get started, and then the results do the rest.
In conclusion, for CMOs to appropriately embrace new technologies, keep their companies ahead of the competition, and convince others in their organizations to bring more data-driven applications to their marketing stack, they need to have a firm grasp on the underlying technologies that power their success. They don’t need to be implementation or tactical experts, but they do need to know how things work and fit into their marketing mix. It’s no longer enough to be translators of market dynamics and masters of marketing strategies. It’s time for marketers everywhere to be knowledgeable about data science too (and keep those CIOs off their backs!).
What do you think?
Will the roles of CMO and CIO converge? Who do you think will come out on top?
Are predictive analytics, machine learning, and automated personalization / optimization the most exciting advances in email marketing in 2015?
What can be done to help marketers better understand and embrace data science and related technologies?