What is systematic market research?

At present, the consumer insights landscape — whether you look at corporate insights, the vendor landscape, or the consumers themselves — is between two worlds: digital and analog.

An industry formed on precision, understanding, marketing science, and getting to the “why” has seen the benefit of hundreds of innovations in the past five years. These innovations enable credible research to be done at the speed of business today, with interfaces that modern marketers can use. Yet too often, these innovations have been driven by thinking just about technology, not about people and process.

We haven’t gotten to the promised land yet — that place where insights teams can spend their time driving growth and moving financial needles while constantly improving the insights coming out of their connected systems.

This is why at Zappi we believe systematic market research is the only way forward.

What is systematic market research?

Systematic market research is an integrated approach to market research that connects different technologies, thought partners, and stakeholders in a way that delivers efficient and predictive insights over time.

So what does it look like specifically? In my view, systematic market research is research that is:

Your consumer won’t tolerate a one-size-fits-all product from you, and you shouldn’t accept one from your vendors either. More than just making sure any new system you implement works for your peers, you must ensure it’s configured to your business across these specific areas:

Accessibility in countries that matter to you

Each business has its core markets that matter. In some businesses, you will make proxy decisions in a lead market, whereas in other businesses all major decisions (like new product launches, changes to platforms, and ad campaigns) are tested at the market level. Systematization requires local set-up and relevance. Companies like Zappi can run work globally, particularly for clients who want to systematize.

Target audience and sub-groups you analyze

While you shouldn’t set up sample frames for every project (because that waters down your ability to run longitudinal analysis), you should work to screen on category and use filter questions to drill into specific demographics and user types to help you maximize opportunity. This should be done with as much consistency as possible, but not standardized until you’ve thought through your organization’s needs.

The specific setup and instance of the tools you use

Any good expert-led system (like Zappi, for example), will be a tool that answers the right questions to help validate and optimize a business decision. But it isn’t enough to pick the right solution. The magic happens when you configure that system to be relevant to you and your business. That includes areas like:

• Unique questions: These are a unique set of questions you always ask to get at specific topics that matter to your business, link back to a larger campaign, etc. — either pre- or post-stimulus exposure. A tool may give you some pre-set questions, but you need the ability to add in your own unique questions to suit your needs.

• Attributes, messages, and price points: By having an intentional strategy on the words you use to describe various aspects of your concepts, you’ll be able to draw longer-term conclusions because you’ll have a database of results you can compare to each other over time. You’ll also save yourself time and money by avoiding doing duplicative research. Why re-test whether a celebrity addressing a current event will work in an ad when you already know it is a bad idea, or that this idea requires a certain type of celebrity?

• Tagging your goals and key components for each study as part of the configuration and set up: You can tag your concepts in whatever manner is most meaningful for your business — such as ads designed to drive trial, increase brand awareness, acquire a specific customers of competitors, etc. Through tagging, you can link back to the goals of the ad, the innovation in the ad, what the survey metrics were, and more. With this level of personalization, you can do some very robust analytics on what works and why in a macro fashion with a few simple clicks.

The way the access of the tool is set up across your teams, stakeholders, and external partnerships

While there are a bunch of quick wins in front of you, like easily moving your testing away from a legacy platform with to a free norms build, it is very important that you think through the access requirements of your colleagues in other teams like marketing, innovation, other insights groups, and agencies.

In some instances (like idea screening, for example), you may wish to personalize and lock the tool but let marketing run their own work. In other cases (like ad testing, for example) you may want to give reporting access to your creative team but lock their ability to actually run the tests. All of this is possible on any credible platform today.

At a minimum, your data must be safe, secure, owned by you, and accessible to all key stakeholders.

All of the things I’ve described above will make sure that your ecosystem is connected, but also that the data is useful beyond an initial test. You don’t just want to answer a single question with a single test and never use those insights again. You can learn a lot by looking across tests over time. But you can only do so if you use one system for multiple tests and think through your personalized needs upfront from consumers to attributes, to questions and success criteria.

Unlike with the vast majority of research collected today, this approach gives you a consistent data map making it easy to run robust analytics across 100 innovation ideas to what works in the aggregate to drive health-conscious consumers, for example.

If you think about the ideal MR stack, you traditionally have dedicated platforms for ad testing, innovation, tracking, packaging, user testing, on-demand qual, etc. There is a lot of holistic learning that can come from linking all of this data together through integrations of partnerships and knowledge management systems.

In my opinion, no one has cracked the code on this yet, but we are starting to see the possibilities. For example, McDonald’s mined its innovation learnings and used them to inspire a breakthrough advertising campaign.Essentially, you want to unblur the lines of each of the partnerships in your stack. Your ad testing data should talk to your concept testing data. Your sales data ties back to the creative data.

If you’re approaching your research in a more systematic way, you’re setting your insights team up for success. They’ll spend more time connecting across the organization to embed the consumer into your business decisions and less time writing research briefs or running the same study over and over again. They can spend more time keeping up with new research trends and emerging technologies without losing your legacy learnings.

And, importantly, they can build on learnings over time. As I mentioned earlier, when all of your results from past studies live together in one place — and are based on consistent methodologies like leveraging the same audiences and research questions — you can study the results together and ultimately learn a lot more than you could from one-off studies. You can look at results in the aggregate to spot trends.

Thus, your understanding of the consumer improves over time.

We can debate System 1 vs. System 2, scales, sample sizes, and norm quality all day. But it is critically important that at the end of the day, we predict what will happen in the real world with a consumer lens.

When you adopt a systematic market research approach, you move from validation and past optimization into a world of prediction.

To do this, you need a plan in place to feed performance data (the data that says what actually happened) into your systems across advertising, innovation, user, brand, and packaging testing so your systems get smarter and begin to help you predict real-world results. You’ll also want to include other information like whether it launched globally, the numbers of units it sold, the percentage of its budget it hit, whether it was national or not, etc.

As an example, we talk to insights people all the time about using more System 1, like facial coding. Here’s the thing, I love facial recognition — it’s cool and it brings stuff to life. But it requires additional incentives and expertise to mine it for insights.

So instead of investing in facial recognition, we employed our data science team and some of our key customers and did a side-by-side comparison with the following hypothesis: Live emojis are just as predictive of sales as facial coding. Once we proved that was correct, we decided NOT to make facial coding part of our core offer but rather an add-on, because we didn’t want to risk respondent experience, cost, or time for something that didn’t add predictiveness.

I challenge all of you reading this to keep “is it predictive?” as your guiding principle. Because the end result is that once the system is predictive, we can all work in a world where insights managers are storytellers. They will be curators of culture and growth strategy and can advise their growth teams on how to resonate with consumers, change behaviors, and develop an affinity to their products and services.

The opportunity in front of us is incredible. There has never been a time where consumers are in more control than they are now.

The question facing us now is whether the market research industry can change fast enough to maximize on the growth that this space will see? Because the industry is evolving, whether companies are ready to change or not.