A Scientific Approach to Better Digital Marketing
Aug 14, 2019
Most of us have heard these polarizing words from Henry Ford:“If I had asked people what they wanted, they would have said faster horses.”
Except, there doesn’t appear to be any evidence that he actually said it.
The debate over providing customers with what they say they want vs. what they don’t know they want is always fierce. A Steve Jobs fan might say something along the lines of “If Steve Jobs would have asked customers, they would have wanted a faster Blackberry not an iPhone”. On the other hand, a market-oriented leader will tell you that innovation is driven by asking the customer. Whether you believe that innovative products come from customers or from geniuses like Jobs/Ford, I would argue that real customer insight is what spurs innovation. These insights may come from customer feedback or genius brains or just by stroke of luck. Now, if you don’t consider yourself lucky—or a genius—you may feel like you’re left with gathering customer feedback.
Surveys immediately come to mind when thinking about customer feedback. But as we know, what people “say” they want vs. what they “really” want are two different things . This is where experiments and science comes to the rescue. For digital marketers trying to get insight into customer behavior, a scientific approach is warranted. Following is a framework that will help you do exactly that.
Start by creating hypotheses and go about proving/disproving them. In a work environment we run into a lot of people with strong opinions. Even when you show them the data, they will find an excuse as to why something doesn’t apply in this specific situation. I knew an executive, who until late 2016, did not believe that mobile web usage was larger than desktop web usage. Even when shown the data, the response was “I don’t think this is true for OUR customers”. Human beings are inherently biased creatures and it can be hard to accept data that contradicts our view of the world. How do we counteract this? Well, science can help. Everything should be treated as an assumption and then converted into multiple hypotheses.
Here are some examples:
- Creating mobile specific changes will not reduce the desktop traffic.
- Putting the registration process on one screen should increase the user account creation by 5%.
- Adding descriptive language to products will increase the click rate on products by 7%.
Once you have a few (or a few hundred) hypotheses, It’s time to run experiments.
Run experiments on the digital channels. Marketers should always be testing new features or new messages on all digital channels. Something as simple as A/B testing can get this job done. Gather ideas from the internal team about what may or may not work. As discussed above, the ideas or the beliefs should be turned into hypotheses. And then run the experiments on the website or other digital media channels to test them out. Consider hypothesis 3 from above. A marketer can have two versions of the product pages: one with descriptive language and another with everything the same except the descriptions. Now the users will randomly see either version A or version B. Collect the user statistics and click-through-rate for each scenario. After a few hundred iterations, you will have enough data to confidently prove or disprove the hypothesis.
Continually iterate and improve. The experiments should not stop after proving or disproving just a few items. The learning process is continuous and it should build on previous findings. That’s how science moves forward and that’s how digital marketing should move forward as well. An iterative approach building solutions quickly and failing fast lets the whole organization get more customer insights than any survey would.Continuous improvements are the building blocks of agile methodology that has reduced product life cycles, including at the digital agency level. Here at Oshyn, we use our Wave methodology (based on agile principles) in our Build and Maintain services to help agencies iteratively deliver higher quality results more efficiently.
In conclusion, successful teams stay diligent learning, adapting, and changing to keep up with the quickly-moving business environment. They look for innovation from all sides but don’t rely purely on gut instinct or strokes of genius. They work tirelessly at creating hypotheses, running experiments, and constantly improving products by being agile.