Using Mental Models
understanding mental models will change your approach to product management
When building a product, even if you don’t realize it, you and your team create a conceptual model of how your users are supposed to engage with your product.
However, when that conceptual model differs drastically from the user’s mental model of your product, challenges can arise for the user as they experience the product.
Let’s Back Up
Wait, what are mental models exactly? They are the internal representations constructed by the brain to grasp the world around them effortlessly.
The brain efficiently digests all of the sensory and information perpetually bombarding it to make quick decisions. And mental models create certainty, even if it means aspects of it are false.
Now, what does this have to do with product management again?
The result of collaborating with your product designers and engineers during a product launch is an artifact that looks something like this:
That is the conceptual model of your product. It’s rooted in logic and has a coherent flow. Things are supposed to happen in a specific, expected way.
The conceptual model is your product bible. It’s the go-to reference for product iterations. You’ve determined which parts of this model are most important for the audience to adopt and even labeled these ‘key performance indicators.’
But, what happens when your user’s model looks more like this?
Let’s call that the realized model.
You build features onto broken models — a Frankenstein approach. You measure the wrong metrics. Or your key performance metrics are underperforming for a feature, but in reality, your audience will never engage with that element because it doesn’t get priority in their mental model. But most importantly, a significant difference in models could leave your customer feeling like this:
What Can We Do About It?
By determining the realized models users conceive of our products, we can build to address the gaps in misunderstanding from the realized to the conceptual or fix the confusion in the conceptual to match the realized. But we can’t do anything until we measure our users realized model.
Mental models are often incomplete, contradictory, and difficult to measure¹. Like many aspects of product management, it’s as much of an art as it is science. Distilling a complete and precise picture from your users is impossible, given the complexity of human mind a person can’t even accurately articulate their mental processes; however, measuring what we can of realized models is directional enough to guide meaningful design and product decisions.
“Mental models are deeply ingrained assumptions, generalizations, or even pictures of images that influence how we understand the world and how we take action.”
Tips to Determine a User’s Mental Model
1. Get Out Of The Building
Coined in the Lean Start-Up movement, you and your team must get out of the building. Leave the office and find your customers. Target customers within each of your critical segments since their intent when using your product may impact their realized model. Try to meet them where they would use your product in their natural environment.
2. Recognize Your Bias
As the product gatekeepers and contributors of the conceptual mental model, it’s important to be aware of the bias an individual brings to qualitative research. Confirmation bias — forming a hypothesis before an investigation, by pre-constructing what you believe to be their the mental models — causes you to look for actions and signs that uphold your original point of view while throwing aside contradicting data. Before going into research, mentally prepare yourself by disregarding what you think you know and remain open to discovering information that conflicts with what you think you know.
3. Observe And Take Notes
Begin by making the person feel comfortable and by asking open-ended, introductory questions. If your product is live, ask the individual to interact with the product as they usually do and carefully observe. Avoid scripts.
Merely study their interactions and behaviors. Key things to look for: Where do they start? What do they skip over? At which points do they speed up and which areas do they slow down? If it’s an app, notice their physical actions. What facial movements do you witness? How does their body language change throughout the experience? Applying the Pareto principle, focus on examining the 20% of features used by 80% of users.
4. Identify Patterns And Gaps
What patterns do you identify with users? Are there shared patterns amongst similar types of customers? Or do different patterns exist within your audience segments? Because humans create mental models to simplify cognitive overhead, patterns may form due to either assumptions made or gaps created in place of incomplete information.
According to Individual and Group Decision Making: Current Issues, mental models provide a heuristic function by allowing information about situations, objects, and environments to be classified and retrieved in terms of their most important features. Especially when rapid comprehension and response are required.
Therefore, it’s useful to look at where users are taking mental shortcuts in your product and using their own rules of thumb — find out what they believe to be true even if it isn’t. What they skip over, such as a feature in your app, is input to guide your decision: try different flows and messaging to drive users to that particular feature or cut it all together and expend your efforts elsewhere.
5. Ask Questions
After observing behavior, if there were actions taken that did not line up with the conceptual model and caught you off guard, go back and ask questions. Be cognizant that what they tell you may not match up with their mental model but the information you collect is still useful. Behavior takes precedence over words. Actions speak louder than words, after all.
Identify how much information is presented at once for a user to digest. Since the average person can only keep seven items in their working memory at any given moment³, decreasing the cognitive load of your product can be impactful. Your in-depth on-boarding explains all of the app features, but it could be doing more damage than help by overloading the user with too much. It’s likely why Google’s homepage is limited to a single search box.
6. Try Story Cards
Another way to stack rank which features are most valuable to your audience is the story cards method. Write each possible action a user can take in your product onto a separate notecard. Randomly mix up the notecards and have users arrange the cards in the order that seems most logical to them. Or even have them discard those that are unimportant. Again, look for gaps in flow between the conceptual and realized. By proactively having the user assemble the product in a way that is most logical to them instead of reacting to the predefined flow, they will avoid the order effects bias.
7. Look Elsewhere
According to Jakob’s Law, users spend most of their time on other sites and prefer that your site work the same way as all the other sites they already know. Pro tip: observe and understand the mental models your users have constructed of their top sites. Evaluate the precedent set by forth by primary players in the market both directly related to and tangential to your industry.
Product designers establish patterns, such as the widely used hamburger menu in the top left corner, which users come to associate and expect. In the case of the hamburger menu, people associate the iconography and placement with navigation to additional information. How do your users find more information? Is it in the same place as the industry or completely different?
Riding the wave of what users expect from the industry can create efficiencies, or it can be useful to differentiate your product by establishing a new experience that disrupts what your audience expects to happen. There’s an argument for both.
8. Compile And Prioritize
As I’ve written about before, effective prioritization is a fundamental skill for product managers. Not every observation requires action. Take what you learned in the above exercises and boil it down the essential takeaways. Layering this qualitative data with other factors like quantitative data, product strategy, competitive landscape and much more, you will have integrated information to draw on as you carve out the next round of product development.
9. Do It Again
It’s important to revisit and measure the realized models of your users on a continual basis particularly when you ship a new feature or redesign your product. These artifacts should be a living document.
One Last Thought
Mental models are meaningful as a form of input of qualitative data to inform decisions and influence your product roadmap.
The more we get out of our headspace and into that of our consumers, the more worthwhile products we deliver.
p.s. thanks for reading! If you’ve used mental models, I’d love to hear about it in the comments below.
References:
blog post content
