It is Time for a Revolution (Again)

At the CASRO Research Conference sometime around 2010, I was giving a talk on the topic du jour.?I don’t remember the specifics but the gist was there was some perceived threat to online research as we knew it then.?I stopped my talk after saying the phrase “traditional online research” to ruminate on how bizarre that phrase seemed.?A nascent industry barely ten years old was now being credibly referred to as traditional and had emerging threats to it.??We had come through an era of tremendous transition and turmoil.?I remember how folks who opposed online research called it’s adoption the end of humanity as we know it…?Well, humanity still exists.?The key to the revolution was that online survey research capabilities enabled companies to generate data-driven decisions more effectively and more frequently. It was an exciting time.?

Another 10-15 years on, nothing much has changed within the industry.?Sure, there have been new ways of procuring sample developed, more efficient methods of dealing with sample introduced, and more ways of taking surveys now than ever before.??But the process and output is basically the same as it was 10, 15, or even 20 years ago.?The issues the industry faced then are basically the same issues that the industry faces now.?The mindset hasn’t really shifted.?We basically do surveys in the same way with effectively the same sample as before, just more efficiently, well sometimes.?And that would be ok, except for the fact that the world has changed.?Without a radical shift in mindset, the survey industry, followed soon by the market research industry will become largely irrelevant.?Not wiped out but irrelevant.?We are already seeing shifts in that direction.

What do I mean by that??I have already alluded to the fact that I see the real revolution behind online research was that companies and ordinary people were given the ability to get more information and consequently make better decisions. ?Research was available to all and in quantities never before possible. It elevated the role of market research in the decision-making process.?But data and information are coming at us these days from all directions.?We are in a data / information revolution and the stars are the IT departments not the research departments anymore.?IT departments manage the data lakes, provide the infrastructure for the different data sets and create the reporting tools that get the right information to the decision makers in easily digestible form.?Data /information can come from a myriad of different sources these days with traditional online research just being one of those sources.?And, I would argue its importance has diminished greatly already.?

For reporters the key questions are who, what, where, when and why.?Key decision makers are interested in all of those same questions and they look for data that gets them the answers to those simple questions.?Survey research is still the best at figuring out the “why” but other types of data and information have overtaken survey research for the other questions.?For example if you want to know anything about how you are doing versus your competitors, POS data is the best with some limitations on the retailer and brand side.?If you need to overcome those limitations, receipt data does not suffer from some of the biases of survey data and given its availability, why not use receipts to answer those questions that POS data can’t??It is available when you are.?You don’t need to wait for a survey to gather the information.?This is but one example where survey data used to be king and no longer is.?

Observational data is everywhere these days, particularly in a digital environment. ?Streams of data about the activities, likes, purchases, etc of an individual can be gathered and analyzed in real time.?Large databases of information about virtually everyone, everywhere abound.?Larger companies are spending their limited resources gathering information of this sort rather than expanding their research operations.?Even where survey research still is the best method to use, newer analytic techniques are getting pretty good at guessing why folks are doing things even without survey data.?Add to this advances in ML/AI and it is looking like folks in the survey research industry need to do something to retain their relevance quickly.?

So what should be done??The online survey revolution elevated the role of market research in decision making and it helped open the eyes of the C-suite to the power of data in decision making.?The C-suite has run with the orientation of data-centric decision making and they have moved beyond survey and market research.?They think purely from an information orientation.?What information do they need to make decisions or optimize their business.?Survey research providers and market research professionals have a powerful position to use if they can see their way to doing so but they too need to adjust their mindset.?They provide information not survey results.?And surveys alone will not answers the questions.?Market researchers and data providers need to adjust their thinking to encompass the entire data ecosystem.?

Surveys will continue to have their place but in order to thrive in the new world, they need to provide better information, faster.?In addition, survey data need to connect to other data streams and serve as the backbone of the connectivity across all of the data streams in order to return to prominence.?

It is extremely important to solve the quality issues that have plagued the industry as a first step. The survey industry has had initiative after initiative concerning quality.?A lot of motivated and talented people have been involved in the effort.?We have learned a lot in that time.?At this point, we know how to provide quality data, for the most part, but it comes at a cost.?Clients want quality data and seem willing to pay for it but there are virtually no mechanisms for them to be able to understand what constitutes high (or low) quality data.?It is not in the best interest of some in the survey research business to provide these mechanisms.?And those third parties that may be able to deliver this information do not want to anger those that might pay for their services so the situation remains muddied.?In the absence of these mechanisms clients see all sample as equal and in that case sensibly go for the lowest cost.?ML/AI approaches to quality illumination show real promise.?The amount of information being flagged as suspect by these techniques is somewhat scary, sometimes going above 25% of the sample.?There needs to be the adoption of a quality standard that is real and easily understood by vendors. ?And once this standard is adopted firms in the industry need to react to what is happening.?Clients need to understand that they will need to pay more if quality is truly identified and need to react by putting their money behind this effort.?

Four of the major industry bodies have announced that they will be joining forces to address the situation.?I applaud the coordinated effort.?We do need to ask ourselves what is different this time that will give us hope.?I argued at SampleCon that what needs to be different is that not only do we need to do the work but we need to focus on the mechanisms of implementation and how to incentivize the various actors to implement the solutions.?Finally it needs to be made clear what is good research and what is bad research.?If the industry achieves this then real progress will be made.?

In parallel, speed needs to be a focus. Fast turnarounds are necessary in this always on world.?Given our past, I have no doubt that we can rise to this challenge.?But not only does data need to be delivered quickly to have impact, it needs to be delivered in a way that allows the users to use the data in a meaningful way.?The information flow is so strong these days that practitioners often have a hard time making sense of It all.?The right data needs to be highlighted in the right way at the right time.?When something meaningful happens, the user of the data needs to know and more importantly be aware of this information in sea of other information.?

This is where the industry can step in but it requires courage.?Why??There are infrastructure costs that may blunt short term profit but will be hugely beneficial in the long run.?But this also plays into our strengths.?IT departments may be the star at this point but there is often a lack of understanding of what is important from a substantive viewpoint?Combining forces with IT departments research providers can be quite powerful, integrating the technical know how with the methodological and contextual prowess.?This will unlock a tremendous number of opportunities.?

But in order to unlock these opportunities to thrive we need to resolve how to bring different types of information together.?Survey and other panel based data (receipts, scanner) can act as the backbone or glue that serves to bring different data streams together.?Data architecture needs to take the different types of structures in order to be effective.?No data type is perfect.?Most report at different levels and measure different dimensions.?But imagining how things fit together can open up a whole host of possibilities.?Firms in the space need to imagine what is possible and start expanding their offerings (credit card, receipts, POS, GPS, scanner, social media, etc.) that connects their core data to other data types.?This will drive the next data revolution and put market research back at the forefront when it comes to driving decision making.?If we don’t do this we will become more and more marginalized.?

It is time for the next revolution!!!?We need to drive better data, provide it faster, and become the backbone to the information ecosystem that is used to drive decision making now and in the future. ?I am certain we have the passion and ability to win this one but it won’t be easy.?There will be winners and losers but at the end of the day our survival is at stake.??