When do you use market research to inform decisions vs when do you use behavioral data?
I absolutely start with behavioral data first - assuming that by "behavioral" you mean product usage data, regardless of the question I am trying to answer. Behavioral data is unbiased (it is users in their "natural" environment with your product), and it is always much more valuable to observe behaviors vs. stated answers. You want market research to confirm or disprove a hypothesis, which you develop looking at the data you already have. After you understand usage patterns (ideally by your relevant segments) you can pinpoint the areas in which you're still not really sure about and design research against these.
If you have the ability to use both, it tends to help paint a richer picture. That said, a good rule of thumb is that behavioral data can help you understand what people are doing, and market research can help you understand why they're doing it.
Let's imagine you lead product marketing for a subscription food delivery product (like a DoorDash's Dash Pass or Postmates Unlimited) and are seeing cancellations increase. That's a big behavioral data-based insight. Now, you might be able to use behavioral data to form some hypotheses as to what's going on. For instance, has there been a surge in demand recently (e.g., from COVID-19 increasing food delivery), and that's increased average wait times, making people unwilling to pay for a premium product? Has a certain cohort of customers been ordering less, meaning they might feel they're not saving enough in waived delivery fees?
All of this is helpful, but it won't definitively get at the root of why customers are cancelling. Market research can then help you test those hypotheses you came up with to put more color behind it.
Beyond these, there are other macro things to consider:
Behavioral Data
- How people are behaving with a current product or feature
- What people are doing within your environment
- See how people interact with messaging once in market
Market Research
- How people react to a new product concept
- What people are doing in the broader category (not just your product)
- See how people react to messaging and catch red flags or confusing things before launch
Used on its own, market research is a reasonable tool for getting gaining and sharing knowledge on customers, markets, and competitors. What market research won't do is tell you if you have the right solution. For that, you need to run experiments to test if there's a good fit between the problem and solution. Experiments provide you with the data you need to shape the solution.
Here is what the best teams do:
- Capture granular behavioral data that describes the complex relationship spanning the customer journey
- Embrace deep qualitative and quantitative research and analysis methods
- Task cross-functional teams including designers, developers, and product managers with learning about customers and their needs, aligning their goals around customer loyalty
- Understand how individual and connected behaviors directly impact key metrics around engagement, retention, and customer value. And leverage that information to determine the right metrics to track and work to influence.
- Use that data to personalize and customize the product experience to suit the specific needs of specific groups of users
- Learn as fast as they ship, and ship as fast as they learn
- Close the feedback loop by immediately implementing changes and measuring their impact in as little as hours
Organizations will typically have the most behavioral data on their own customers and visitors. They know what customers are doing (and not doing) on their platform... where they drop off in a conversion funnel, what they click on, how frequently they log in, the product/feature mix they have used/purchased. This wealth of information can inform where to focus on if conversion/renewal/repeat/expansion within your own customer base is the goal. In some cases, for example if you're experiementing with a new user flow using an A/B test, behavioral data may be all you need to make a decision to move forward. But behavioral data is limiting as it only deals with people who are on your platform and features/flows that already exist.
Market research can be used to complement behavioral data to provide additional context. It can provide the "why" behind behaviors and decisions people make. There are also other benefits. For one, you can reach a more representative group of people you don't already have access to to get an un-biased view of what your prospects or the market at large think. You can also get feedback on ideas earlier on in the development process. A great example would be feature prioritization. You could survey people to understand which features from a list you're considering would make people more likely to purchase/use your product.
I'd generalize this by saying market research is better for informing decisions around strategic direction and pre-development; behavioral data is better for informing post-development optimizations.
In our experience, market research is generally better suited toward future-focused strategic decisions than behavioral data (e.g. where can we expand our product offerings and what problems should we be solving?). Behavioral data can usually offer a clearer view into how to improve experiences and deepen engagement with existing products (e.g. after they do X, Y is a natural next step so let's send them a notification to nudge them in that direction).
You can even link your survey data to behavioral data to get even more benefits. The three of us just completed a huge consumer segmentation analysis of our total addressable market. Now we're seeking to activate our typing tool on our site survey and link it to behavioral data so we can see exactly how each segment behaves on our site. This will give us the advantage of the forward-leaning insights from the segmentation as well as the current behavioral patterns of each segment today so we can learn how to deepen engagement now and expand usage in the future.