Marketing has been functioning with too much “gut feel” and too little actual information, said OCBC’s VP of customer analytics Kasper Hansen.
Speaking at Marketing’s Analytics 2015, he pointed out marketers in general have three big questions to answer: How big should my budget be? How should I spend it across all my geographies, marketing channels, customer channels, etc? And how am I going to hit my targets?
The issue is how they go about answering this, and it tends to go along certain lines. “Here’s last year’s budget – let’s add a bit or take away a bit from that.” “Let’s do a percentage of sales” or “Let’s match what the competitor is doing,” he said.
“These are pseudo methods for coming up with how we can spend the money. But the proper way to look at it is by the payback you get from allocating the spend.
“But the thing is marketing groups don’t possess this information. They tend to go by intuition or gut feel. Then a groupthink or conformity starts to emerge. For instance, go for TV for a regional reach – this is what I’ve noticed after 45 years of sales. Certain industries have certain types of marketing that they prefer.
“If we rely on gut feel and intuition, how do we question more?” he asked, positing that there simply isn’t enough information in the marketing department to get the answer right – no matter how experienced the team may be. This is in comparison to the information that is present in data, which is a huge pool of resource that is often not mined.
“All of marketing’s measures at present, for example, ad spend and digital spend, these are not exhaustive.
“What we really want to know is what is a direct causal link to the input given,” he said, adding this was what analytics could do.
For example, he advocates using models such as response curves – which is the heart of any analytics process – in tracking the cause and effect relationship between an activity or spend and outcome. For example, a price versus sales chart.
“If we can draw out a response curve – if you can draw that out then you can estimate what the optimal budget will be. It should all flow out of this.”
There are several things to look out for as well.
For example, there’s the danger of simple analysis. He cited the example of an electronics firm he previously consulted for.
“They had some data showing that display advertising was not as effective as search or affiliate websites. When they looked at last-click analysis, that was how they apportioned it. That didn’t feel right to them, and the marketing team asked themselves why they spent so much on display advertising.
“If you saw the end results, it would seem that search or affiliate sites drew the most results. But if you took a more statistical approach and looked at the results one month prior, you would have seen that display advertising played a key part.
“It fed interest and people would go educate themselves on the product or service and a month later think about buying the product.”
This is an example of looking at something over simplistically, he added.
Another thing to look out for is not to let analytics function in a vacuum.
The issue with most companies, and why the analytics function has not taken off, is it tends to be a once-a-year expensive function done by external vendors. But by the time the analysis reaches decision-makers, it is no longer useful.
He advocates having an in-house model and having the key decision-maker or the “person who understands the business and asks the questions” to the person who does the analytics.
“This is valuable as you can make decisions quickly and affect the business cycle.”