Speak data fluently. CMOs need to be skilled in the language of marketing technology, AI and customer data to deliver value.
CMOs spent nearly one third of their marketing budgets on technology in 2019, according to Gartner. Yet some experienced marketers feel they don’t know the key marketing and AI technology terms that would allow them to understand the technologies that are critical today. Even leaders known for their digital savvy have had to grow into the role.
JP Morgan Chase CMO Kristin Lemkau offers a case in point. In a podcast conversation with former P&G CMO Jim Stengel she said, “The technology piece was the one I think that stunned me the most, of how much I had to deeply understand my tech stack…To operationalize that was going to be the key to everything. The whole point of shifting from product to customer is actually a tech job of driving personalization at scale.”
Though there is no substitute for deep understanding of technology and how it brings value to marketing, there are shortcuts. Here, Persado offers a reference guide to help leaders develop quick fluency in the most current marketing technology and AI terms.
Machine learning uses algorithms to identify patterns in data and predict likely events from them. It is one of the oldest forms of artificial intelligence, dating back to the 1950s. The algorithms “learn,” in the sense that their predictions get more accurate over time as they gain access to larger pools of data. Effective machine learning solutions therefore rely on large pools of accurate data and strong statistical models.
What can machine learning do for marketers?
Machine learning is increasingly used to improve the ways in which brands engage with customers. It is often used in tools that predict actions customers are likely to take before they take them. The same tools can also identify pain points that interrupt the customer journey or lead to attrition. When organizations use that information to improve customer experiences or design products, it translates into faster conversions, greater loyalty, and less churn.
Read More: “What Is Machine Learning?”
Data integration is the process by which an organization extracts data from the many in-house and external systems that collect it. The goal of integration is to combine data in a consistent format so it can be used and analyzed for insights. Integration is a prerequisite for any organization that wants to develop a cross-system, cross-channel view of the customer.
Why do CMOs need to understand data integration?
Challenges to integrating multiple, diverse data stores can increase the total cost of ownership (TCO) of martech systems and decrease ROI. Smooth data integration ensures that best-of-breed systems work together in harmony. Ask about native integrations when assessing martech vendors to avoid costly surprises.
Application Programming Interfaces (APIs)
APIs are a set of tools and methods that allow different software programs and data sources to communicate. APIs come in many flavors including SOAP, REST and JSON. Think of them as the computer equivalent of a language interpreter. They ensure that messages from one program or application are understood by another.
Why do CMOs need to know about APIs?
APIs enable marketing teams to buy technology and marketing AI solutions from different vendors knowing they will be able to make the solutions interface with one another. CRM systems talk to different email and analytics systems through API integrations. APIs also enable faster and more effective innovation of new products for the customer. For example, APIs allow a brand’s new app to access in-house data without a glitch.
Read More: “5 Things Every CMO Should Know About APIs”
Natural Language Generation (NLG)
NLG refers to the way in which computers analyze data and turn that data into written narrative and words. The sources of data can be voice recordings or written text. Once analyzed, they can be used to generate written or verbal responses using words and syntax that a human would use.
How does NLG apply in marketing?
NLG allows brands to deliver customized content across thousands of customers and interactions. Financial services firms use NLG to take raw financial data and generate personalized portfolio summaries. Brands use NLG in the Persado platform to create personalized marketing messages across channels such as web pages, social ads, and direct mail.
In its simplest form, experimental design is a method of laying out data so it’s possible to conduct statistical tests on it. Experimental design allows researchers to test multiple variables at once and identify cause-and-effect relationships between inputs and outcomes.
For example, a mobile phone company that wanted to determine how emojis impact an SMS campaign could use experimental design to test multiple possibilities at once. The data team might test four variations – the happy face, the exclamation point, the winking emoji, and no emoji at all – to see which version performs best. If the team used an A/B testing method for the same purpose, it would have to run more than 5X as many tests.
Why should CMOs understand experimental design?
The combination of effective research design and predictive analysis used in experimental design allows marketing teams to identify the specific components of ad campaigns and marketing creative that are most likely to engage customers or drive specific outcomes.
Attribution models analyze the various channels a customer visited before some type of conversion in order to accurately allocate the credit for influencing the conversion.
Since customer journeys are often multi-channel, brands need different ways to measure a channel’s impact. Instead of assigning all the conversion credit to only the last point of contact (“last click”), which would disproportionately undervalue any other touchpoint before the last click, an attribution model enables the brand to choose a first-click model or any other custom weighting model instead.
Why should marketing leaders pay attention to attribution models?
Attribution models are important for CMOs because they show which parts of the marketing mix work, where they work across the customer journey and channel mix, and as a result which parts of the mix should receive more or less investment.
Technology portfolio management
Technology portfolio management is a systematic approach to managing technology investments, projects, assets, and people. Disciplined portfolio management ensures that the business gets the most out of its technology assets. It also reduces the acquisition of solutions that are incompatible with or redundant with existing assets.
Why should CMOs learn about technology portfolio management?
This is a new area for CMOs but it’s table stakes for CIOs, given their domain over enterprise IT portfolios. As more technology spend comes under the control of the CMO, however, tech portfolio management skills will be critical to prevent redundancy and secure data assets.
Read More: “Portfolio Management: How to Do It Right”
Putting it all together
Here is a paragraph describing how one company uses NLG that should start to make a little more sense now that you know a few key marketing AI and technology terms:
“For example, our Wordsmith platform accepts structured data via uploading a CSV directly, passing the data to our API as a JSON object, or through one of our integrations that connect to the API like Tableau, MicroStrategy, or Zapier.”
Though an Executive MBA for martech and AI has yet to be created, it’s possible to learn key skills and concepts on the job to ensure marketing technology and AI deliver real business value. As a result of learning how to evaluate martech and AI more effectively, JPMorgan Chase’s Kristin Lemkau was instrumental in pressure testing Persado’s AI solution to ensure it delivered the requisite marketing and business ROI.