People remember less of your product than you think
I asked people to get an Uber using their memories alone.
After a few months of little social interaction, I decided to restart my small off-time research. Shortly put, a question appears in the world, usually triggered by pure curiosity. I then investigate that question through primary and secondary research. Secondary meaning reading articles, papers or similar studies. When it comes to primary research I‘m mainly interviewing people, hopefully trying more methods in time.
This time, I was getting back to my bike thinking: “ I have no idea where I left it again…should’ve taken a picture”. While struggling to find it, another thought crossed my mind “what do I really remember anyway, other than a messy collection of images, flashes, sounds, smells and feelings ? ”.
About Memory
I decided to refresh my knowledge on how our memory works then; this would be the secondary research.
Short term, long term & working memory
People remember five to sev…I’m kidding, that’s nonsense. Let’s look into something more interesting. We used to believe that short and long term memory work almost separately from each other. Initial studies associated long term memory with a file cabinet from which we retrieve information, and the short term one as a momentary keeper of information. A better association would be an enormous pinboard packed with old and new sensory information, images, spaces and numbers all interlinked in an intricate way.
How is this pinboard used though? Well, putting it simply, we get sensory information from the world, store parts of this in our short term memory, which we then use to retrieve existing information from the long term memory to help us understand it. The sum of all that stored sensory information and retrieved long term information is known as working memory. This enables us to act in the real world and / or store more information for the long term one. Attention, association, and goals all play a role in this process.
What makes something stick?
On one hand, important events, strong emotions or physical feelings are all strong triggers for storing information, sometimes without us even realizing. On the other hand, when an experience or a piece of information is associated with something we already know, neural paths are connected faster and become easier to retrieve, forming a long term memory. Remember familiarity and using standard interactions? those are basic design principles. The more patterns you relate to the memory, the easier it will be to recall every time.
How about products?
Products exist to support people in accomplishing something. They are not self-sustained entities. If they succeed in supporting people, be it shopping online, finding a job or listening to music, chances are that people will use them again. Goals were achieved through them. The more they use it, the stronger that memory gets. This will then interlink more and more information: getting around the product, shortcuts, personalisations, etc. Furthermore, the experience is likely to be remembered if it continuously connects with them emotionally by bringing a bit of joy, surprises, smiles or support when needed.
There is obviously much more to memory, but the research question started taking more shape at this point. Our mechanism to retain information is mostly based on complex paths formed by a mix of present and past information surrounded by emotions and context. If we are all supposedly building (or at least selling) efficient, easy to use, intuitive, accessible products, then:
What do people remember from a digital product experience?
I then had to pick a product. It had to be something known in the Netherlands, simple in its task and generally perceived as well done, design principle wise. I decided to go with Uber: it’s popular, it’s seen as well done and supports a clear goal: getting people from A to B. This is how a rider would use their service, from their own perspective.
Interviews
Having the question in mind, I moved on by posting my need of participants in a few social media groups, with both locals and expats. They were presented with the study goal and small reward in vouchers.
In order to schedule all interviews, I quickly created a Google sheet with available time slots and vouchers for people to fill in.
Most people chose the early time slots, right after responding to my posts. There was no screening phase at all, I don’t have the luxury to reject anyone at the moment. I did notice that most participants work or worked in a field that required them to communicate with consumers of their work, which might be part of the reason they agreed to support me.
Format
These were all calls in Facebook Messenger and Zoom. The 25–30 min interview was split into four parts.
- Presenting myself, the goal of the study and the way I’ll be using their responses. After agreeing again, they introduced themselves.
- Describing the scenario and asking about product usage.
- First glimpses of their memories of ordering an Uber.
- Diving into each step of the journey at a time.
It was important to mention the product as late as possible. Participants checking their apps beforehand would have resulted in weirdly precise answers. The scenario was quite simple, they need to get an Uber to go somewhere, I avoided to mention a specific location on purpose. Ideally, I would’ve taken a bit more time during the interview to set this imaginary location together with the participant (inspired by interview-based tasks). Otherwise it all may seem unrealistic and too playful right from the start. I was also interested in how much can I push people in describing visual details before they answer to just get rid of the question.
Participants
Out of the 35 responses, 8 people filled the participant sheet. There were 5 women and 3 men, of which 6 expats and 2 Dutch, all living in Amsterdam at the moment. Since there is no screening, their age, occupation or socio-economic status are all uncontrolled variables. They originate from different places in the world, from Macedonia, to France or Canada. Their jobs vary from teachers to web analysts or psychologists. Their usage of Uber differed too.
Findings
Everything was written down during the talks. Immediately afterwards, I transferred those into digital formats. Keyboard typing sounds during an online call might become annoying. I analysed them from two perspectives:
- Looking at the way people sequence and describe steps form their memory, compared to the ones required by the app. A bit of mental vs conceptual model.
- Specific memories about each step — if they didn’t remember one, I started asking more questions, guiding them towards it.
First memory
Without me asking, most participants started describing more about their latest experience, right after mentioning the last time they used the app.
Sequenced Steps
Afterwards, we moved towards their memory of the steps needed to order an Uber. These are then compared to the app. I did however split the second screen, since all participants kept referring to it in two distinctive ways.
The first thing to note is that none remembered the entire flow by themselves. Some skipped several steps, getting assigned a driver quite quickly, while others simply swapped a few. As expected, people who used the app sooner and more frequently remember more of the flow. Their memory of the product is less “noisy” compared to somebody who used the app six months ago. These are all mapped below.
Looking closely, entering a destination and setting a pick up point are always recalled as being consecutive, regardless of which one is first. People seem to remember that the first part is to ensure that the path from A to B is clearly set before continuing. However, Uber is encouraging people to press “Where to” , with location sharing turned on. The pick up point is confirmed only after selecting a car type. It cannot even be changed until then.
People had an easier time remembering selecting a car type (right below), rather than the small map (left below). Even after asking them, only two recalled the map with a time indicator (3 min) and the dark path, describing it as “a blue line” or “ “moving line”. In comparison, when asked, all participants remembered there were at least two types of vehicles to choose from.
Perhaps the fact that selecting a car type includes the price, and is necessary to continue, made this step stick stronger to their memories. The left part is basically an indicator of the current Uber availability and a visual confirmation of the trip, which cannot be influenced at this step.
Lastly, only a single person mentioned the rating step, after arriving at the destination. This could be explained by the fact that our attention towards products decreases significantly once a goal is finished.
Next, I’ll be describing a few individual step findings.
Car Types
After seven out of eight people remembered there are at least two choices of cars, I asked them to describe them as well as possible. However, it was not the visuals they were depicting, they were focusing on factors influencing their experience: car size, comfort and price.
Visual memories are mixed with thousands of other images we already saw. Experiencing comfort, having a good feeling of saving money or being happy to find the right car — as a participant mentioned going back from IKEA with Uber van — have a much higher chance to be remembered on the long term (remember the triggers mentioned in the first part?).
Driver Details
Most people simply mentioned that they get a driver assigned, see when she/he arrives and then simply get into the car. I then asked which details of the driver profile they remember seeing. Each element is analysed below.
To my surprise, not everyone was looking at the license plate as unique ID of a car. In addition, even though everyone remembered different details, there were always at least two elements from above mentioned by each participant. Six persons did also talk about looking at the app to follow the driver’s route and checking the time of arrival notifications — for them these were additional measures of finding the right vehicle.
Rating
Initially, only one person mentioned giving feedback to drivers. After asking if anything else happens after they get out of the car, they all remembered the missing step. First, everyone mentioned the 1 to 5 star rating. This is a widely used system used across products, which makes it easy to memorise. Seven people also remembered having tipping options. When asked, they all seemed to guess or remember to have three options and a free field.
Generic visuals
These are a few details unrelated to specific steps. Some are also based on the last question of the interview, asking people which colours or visuals they think of, when talking about Uber.
Pressing buttons or any UI element for that matter was rarely mentioned, a few did refer to acting labels directly as in “I press Find me ride”. Most people remembered cars being white or black, one mentioned them as blue. Four mentioned how they liked the big clear texts. Everyone who mentioned setting a pick up point also stressed the importance of having location turned on several times. One articulated that the map doesn’t really work otherwise.
This part took quite some digging in the visual memory binder, and I assume that a few answers were formed on the spot to skip the question. Asking people to describe colours, shapes and motions is where I believe the limits of remote studies are reached.
What does this all mean ?
I could spend even more time describing detailed findings, even after talked with only eight persons, for around 30 mins each. There is a pattern forming though, and I can draw an assumption with these preliminary interviews.
A good product provides people with the right clues for using the right memories in order to finish their goal. However, even a well-built product, backed up by a powerful brand is only partially remembered. Hiding clues and assuming people learn and get familiar with the product, due to its success, is a risk and a potential harm to their experience.
Specifically, let’s say that the product team at Uber decides one day to “streamline experiences” and “reduce amount of clicks” by removing the car selection sheet. People will remember what their default choice is, right ? Well, a bunch of people who remember car selection as being a clear step in their process might be really confused. Especially if their entire family suitcases are welcomed by a tiny city car.
Find out what people remember from you product’s experience. Can they walk through the overall sequence ? Do they do it in the same way you designed it ? Break goals into steps, and align those to their own understanding. This will increase their chance of “storing” how to use the product, for the long term. Afterwards, use clear clues to support their working memory in connecting what they’re currently doing to their past experience.
Obviously nobody will ask people to use a product without looking at it. Still, familiarity is earned in time, just like trust and respect. Start with consistency, association and memory triggers in context. Next time when you argue that people will get familiar with your flow by remembering it, remove half of it and see if they can still use it.
PS. Ideally, I would do this study in person, sketching the interface together with people while they are trying to remember it. That would be an awesome experiment to run.
Sources
- Think like a Researcher, David Travis
- Interaction Design Beyond Human Computer Interaction,by Helen Sharp (Author), Jennifer Preece (Author), Yvonne Rogers (Author)
- http://www.scholarpedia.org/article/Working_memory
- http://www.scholarpedia.org/article/Long-term_memory
- https://www.sciencedirect.com/topics/medicine-and-dentistry/working-memory