In today’s world data is just like oil: once extracted they need to be refined to be of any value. The refinement can lead into several products that must then be transported to end-users to fulfill their task of making your business (engine) run. Easier said than done, as many insights professionals would know, and, thus, the request for providers’ ability to transform data into business recommendations is booming.
The struggle is real: It is getting harder and harder to make sense of all the data and information available - and every day petabytes of new data is added to the global data lake. We are reaching the limit of our capacity to make sense of all this information and turn it into captivating, insightful stories; insight teams have their nose under water facing increasing demands with stagnating or decreasing budgets (ESOMAR).
No wonder our clients say the understanding of their business and the ability to provide business recommendations are key in collaborating with us, and the translation of data into business decisions at a pace matching consumers’ changing needs and aspirations is a key competitive differentiator.
Still, a lot of CMIs feel their survey results do not deliver quite the business recommendations they are looking for.
You already know everything relevant to know about your own business, your competitors, your industry, and consumers at large - so the trick is to fix the final, crucial piece of insights: your survey data.
There are three common explanations why survey data do not turn into the decision-driving insights tool you need. They have in common that they are 1) common, and 2) preventable.
Once a CMI for a furniture brand told me: “if I ask consumers, they say they want the red couch, but my sales data says they want the dark gray”.
This is human nature at play: we wake up every morning with the best of intentions to do better, eat healthier, be a better person, and exercise more. Then reality hits and we end up making the same choices we made yesterday and before we know it we’re crawling back to bed thinking that tomorrow we will do better.
We’re just the same when it comes to our consumer choices: some consumer choices we make in a millisecond and we rarely reflect on it afterwards. And never ever do we make our purchase decision on a perfect foundation of knowledge.
In a market research situation we are suddenly being very reflective about our choices and attitudes and, thus, your task is to minimize the risk that your questionnaire produces this bias in how you ask respondents. This goes for all types of market research from concept tests to pricing studies. I wrote a blog post about that here.
Check for this in the questionnaire design process and/or ask your supplier to review your questionnaire.
Another factor that can completely prevent you from making the right business decisions from your survey data is your respondents. If you experience a disconnect between your survey results and your sales data, this might very well apply to your.
Rethinking their sampling approach and target audience was exactly the medicine needed for our client CO-RO when they first came to us: they could see the mismatch between research data and reality, but they couldn’t solve it with their previous supplier. We helped them and you can learn how and the difference it made right here.
Review your sample setup: is your sample representative of your target groups on all verticals like age, gender, geography, income, etc.?
You might read this with a smock feeling that this never happens to you or your team. Or you might smile because you recognize how common it is.
It’s not rare that we experience draft questionnaires with a disconnect to reporting requirements - or additional reporting requirements being announced only after the sampling has been completed and those not really matching what’s been researched.
I remember this global food & drinks brand where the survey included brand image statements and concept testing questions. After presenting the results the client came back and asked us to expand the report to include a driver analysis. Those you’d usually make based on a set of satisfaction questions.
The morale of this story is that you need to define the type of analysis/reporting you want before finalizing your questionnaire to make sure they match.
Bar charts are not inferential analysis.
If you are a part of 30% of companies with more than 1,000 employees that have a CMI team of less than 10 people or you just feel the general pressure of doing more with less that characterizes the industry (ESOMAR), it may feel tempting to cut reporting down to being descriptive.
Don’t do that.
You cannot tell from a bar chart what is right for your business. Consult your market research agency if you’re in doubt whether your survey design can actually answer your business questions.
In summary, your survey results will only become the decision tool you need if you make sure your research is well-designed. That’s tricky to many CMI teams as it might feel tempting to just launch questionnaires and then figure out the details during or after the sampling has completed. But bar charts will never become valid inferential analysis. So check your survey design carefully. Those extra hours spent can mean a difference of millions of Euros for your business.