« View all blog posts

Lesson #4 from a data scientist: You don't have to give up the driver’s seat

Automation. For some people, the word conjures images of robots replacing human beings, much as “artificial intelligence” makes us think of Arnold Schwarzenegger in Terminator. Or how the words “90 minutes of excruciating pain” make us think of Arnold Schwarzenegger in Junior.

But automated marketing isn’t about machines taking over your job. Instead, it’s machines helping you do your job better. You never want to be in a position where machines are making your marketing decisions for you unless you authorize that. Automation should free you so you can spend more time in the driver’s seat. Maybe they should call it automotive marketing.

The human brain is an amazing machine, and there are some things that it fundamentally excels at. Synthesizing and processing information, for example. Making decisions in real-time based on a variety of different inputs. Aligning actions with business value and strategic objectives.


When are machines better?

But there are some things machines do better than human beings. Detecting patterns in data. Making associations between data sets that would seem to have no logical connection. Spotting lookalikes between customers that may look very different on the surface.

If marketers were honest, they would admit that poring over statistics is probably the most boring aspect of their job. Automation and machine learning aren’t replacing your job, but placing you in the driver’s seat in your data-driven organization and doing the behind-the-scenes work for you.

More data does not mean more insights

Many marketers come from the mindset that more data equals more insights. When we begin work with a new customer, they often feel obligated to import hundreds of fields in order to "capture everything." But what they’re really doing is creating unnecessary work. You don’t need everything. In fact, most of the answers that you’re looking for reside in nine percent of your data, not 99 percent. I like to say that data science provides marketers with guardrails. It extracts the signal from your data and keeps you on track so you don’t end up in the woods.

Another fear is that new technologies like machine learning will make a marketer’s existing skill set obsolete. But that’s not true. Machine learning doesn’t displace segmentation, for example, it creates better segmentation. It’s important to remember that the new generation of marketing tools are just that: tools. There is no magical “black box” that will instantly transport you to the finish line. Data science doesn’t tell you where to go, it just helps you get where you want to go faster.

While there is no black box, there should be a central place where all this science is happening. Otherwise, you run the risk of having automation create conflicting customer experiences because the decisioning lives in different tools. As an aside, our co-founder Aaron Raddon has a great article explaining the difference between Decisioning and Orchestration you should check out. For example, if a customer clicks on a link in an email that really speaks to their needs and then gets a different experience on the website because of different decisioning tools, it causes confusion. This is where a customer data platform is really important, because it acts as an intelligence hub for all your decisioning.

So, in summary, machines help with the mechanics of marketing, but human beings have to be behind the wheel. And that’s the lesson for today. Don’t forget to watch the snack break to hear from me directly and tune in for next week’s lesson on Insight without action is just information. Or, as Arnold would say, I’ll be back.