[ 12 ]

Memory: Expectations and Filling in Gaps

In this chapter, we’re going to consider the semantic associations our customers have. By this, I mean not just words and their meanings, but also their biases and expectations (Figure 12-1).

Some of the questions we’ll ask include:

• What are the frames of reference our audience is using?
• What were they expecting to find?
• How were they expecting this whole system to work?
• What are the stereotypes, mental models, or schemas influencing those expectations?
• How do the customers’ stereotypes differ from our expert schemas or stereotypes?
• What changes can we make to ensure that we’re meeting customers’ expectations?

Meanings in the Mind

Let’s go back to the idea of stereotypes, which I mentioned in Chapter 3. These aren’t necessarily negative, as the stereotypical interpretation of “stereotype” would have you believe—as we discussed earlier, we have stereotypes for everything from what a site or tool should look like to how we think certain experiences are going to work.

Take the experience of eating at a McDonald’s, for example. When I ask you what you expect out of this experience, chances are you’re not expecting white tablecloths or a maître d’. You’re expecting to line up, place your order, and wait near the counter to pick up your meal. At a more modern McDonald’s, you might also expect a touchscreen ordering system. I know someone who recently went to a test McDonald’s and was shocked that they got a table number, sat down at their table, and had their meal brought to them. That experience broke this person’s stereotype of how a quick-serve restaurant works.

For another example of a stereotype, think about buying a charging cable for your phone on an e-commerce site, versus buying a new car online. For the former, you’re probably expecting to have to select the item, indicate where to ship it, enter your payment information, confirm the purchase, and receive the package a few days later. In contrast, with a car-buying site, you might select the car online, but you’re not expecting to buy it online right then. You’re probably expecting that the dealer will ask for your contact information to set up a time for you to come in and see the car. Buying the car will involve sitting down with people in the dealership’s financial department. These are two very different expectations for how the interaction of purchasing something will go.

Case Study: Producing the Product versus Managing the Business

Challenge: For one project with various small business owners we noticed that, broadly speaking, our audience fell into one of two categories:

1. Passionate producers

   They loved the craftwork of producing their product, but weren’t as concerned with making money. They were all about making the most beautiful objects they could and having people love their work. They loved building relationships with their customers.

2. Business managers

   They didn’t care as much about what they were selling and its craftsmanship; they were looking much more at running the business and making it efficient. They weren’t as client-facing.

Outcome: We realized there isn’t necessarily one set of expectations from small business owners. Rather, there are two hugely different patterns here, with unique expertise and expectations. One group loved working up forecasts in spreadsheets, whereas the other group wanted nothing to do with that. One group was great at customer engagement; the other group preferred to work behind the scenes. Each group had different expertise, and therefore needed products, services, and language that catered to its respective strengths. This story goes to show that identifying the different types of audiences you have can really help to influence the design of your product. More on audience segmentation coming in Part III!

Putting It All Together

We want to consider all facets of people’s preprogrammed expectations: how something should look, how that thing should work, how to get to the next step in the process, and generally how customers think the whole system will work. As you work to decode your customers’ expectations about the product or service experience, I want you to look out for how different novices can be from experts. As we’ve been discussing with language, it’s crucial that we meet users at their level.

Case Study: Tax code

Challenge: For one client, we observed accountants and lawyers doing tax research. Specifically, we looked at how they were searching for particular tax code information. While the information was historically organized by the type of publication (e.g., journal, book), the end users’ expectations and mental organization revolved around very different dimensions (US taxes versus. international taxes, estate tax versus corporate tax, guides versus tax law, etc.). The interaction models that were available to them at the time just weren’t matching the multidimensional representation they were seeking.

Recommendation: In creating a tool that would be most helpful to this group, the designers needed to revise their model to be more in line with users’ thinking and provide them with filters that matched their mental model (Figure 12-2).

Figure 12-2

Tax experts have a multidimensional organization of tax law information

Real-World Examples

Going back to our sticky notes, let’s think about memory and which findings relate to that lens of our mind (Figures 12-3, 12-4, and 12-6):

“I want this to know me like Stitch Fix.”

This one involves a frame of reference. (For context, Stitch Fix is a fashion service that sends you clothes monthly, similar to Trunk Club. After you give it a general sense of your style, the service uses experts and computer-generated choices to create your shipment, which you try on at home, paying for the items you like and sending back the others.) There’s definitely some emotion going on here, suggesting that the user wants to feel known when using our tool, and is perhaps expecting a certain type of top-flight, professional customer experience. But I think the main point of this finding is that it suggests the user’s overall frame of reference in approaching our tool. Knowing the user’s expected model of interaction is helpful for us to discern what this person is actually looking for in our site.

“Can’t figure out how to get to the ‘checkout counter.’”

Here, it sounds like the user is thinking of somewhere like Macy’s, where you go to a physical checkout counter. You might argue that this is vision, since the user is searching for something but can’t find it; you might argue this is language because they’re looking specifically for something called the “checkout counter”; you could argue it’s wayfinding because it has to do with getting somewhere. Or is it memory? None of these are wrong, and you’d want to consult your video footage and/or eye tracking data if possible. But I think the key point is that the user’s perspective (and associated expectations, of a brick-and-mortar store) were way off what a website would provide, implying a memory and expectations issue.

This comment makes me think of a fun tool you all should check out called the Wayback Machine. This internet archive allows you to go back and see old iterations of websites. This idea of a checkout counter comment reminds me of an early version of the Southwest Airlines website (Figure 12-5).

Figure 12-5

Southwest.com in its first iteration

As you can see, Southwest was trying to be very compatible in its early web designs with the physical nature of a real checkout counter. What they ended up with was this overly physical representation of how things work at a checkout counter, with a weigh scale, newspapers, etc. All of these features were incredibly concrete representations of how a checkout counter works. Even though digital interfaces have abandoned this type of literal representation (and most of these physical interactions have gone away, too), it’s important to keep prior patterns of behavior in mind when designing for older audiences whose expectations may be more in line with the checkout counter days of old.

“Expects to see ‘Rotten Tomato’ ratings for movies.”

I think this one also points to an expectation of how ratings work on other sites, and how that expectation influences the user’s experience with our product. This is also another example of where too literal a reading of the findings may mislead you. When you read “expects to see,” beware of assuming the word “see” indicates vision. In this case, I think the stronger point is that the user has a memory or expectation about what should be on the page. You could argue that the expectation is linked to their wanting to make a decision, so I would say this one could fall under either memory or decision making.

What You Might Discover

In our exercise, we looked at users’ expectations about other tools, products, and companies; how our users are used to interacting with them; and how users carry those expectations over into how they expect to work with our product, or the level of customer service they’re expecting from us. These are the types of things we’re usually looking for with memory. We want to look out for moments of surprise that reveal our audience’s representation, or the memories that are driving them. We also talked a little about language and level of sophistication, as these can indicate our users’ expectations.

We want to understand our users’ mental models and activate the right ones so that our product is intuitive, requiring minimal explanation of how it works. When we activate the right models, we can let our end audience engage those conceptual schemas they have from other situations to do what they need to do. This helps build more trust in our product or service.

Case Study: Timeline of a Researcher’s Story

Challenge: For one client, we considered the equivalent of LinkedIn or Facebook for a professor, researcher, graduate student, or recent PhD looking for a job. As you may know, Facebook can indicate your marital status, where you went to school, where you grew up, even the movies you like. In the case of academics, you might want to list mentors, what you’ve published, if you’ve partnered in a lab, and so on. We realized there were a lot of pieces when it came to representing the life and work of a researcher.

Recommendation: Doing contextual inquiry revealed what a junior professor would like to see about a potential graduate student, and how a department head might review the results of someone applying for a job in a very different way. We learned a lot about users’ expectations regarding categories of information they wanted to see and how it should be organized, and how that differed from a typical resume or curriculum vitae (Figure 12-7).

Figure 12-7

Hypothetical web page for an academic profile

Concrete Recommendations

• Ask customers about the underlying source of their expectations: What are you basing your expectations on? What else have you used that works like this?
• Build not just an audience persona (more on this in Part III), but a set of assumptions that persona holds regarding word usage, how things should work according to their expectations, and how they are framing the problem.
• Document visual attention biases and the words/actions users look for, word usage and the meanings associated with those words, the syntax of their sentence construction, the answers they are expecting from the system, and the flow that users expect to have.
• All of these together can provide powerful suggestions for how best to match the system to end users’ needs.