a view of a grocery store produce section from above

The Seamless User Story is the Enemy of Good Design

Have you ever been at a presentation where someone presents their design and walks you through a user story that looks really awesome? By the end of the talk, you are hooked on the concept. Wow, what I can do with that in my toolbox!

This video is a perfect illustration of what I am talking about. It is a presentation by Tim Rodgers from Rehab Studios talking about his vision for the future supermarket experience. The design looks great. And in his user story, it works great. I want it for my next shopping trip.

[Tweet “Challenges of the seamless #user-story in #design | @PSFK @timnotjim #ux]

My Take

But take a step back and read between the lines. Think about what happens when something doesn’t work perfectly.

  • In the presentation, the app gives the shopper the one perfect recommendation for what to cook that night for his hot date – right on his smartwatch.
    • What if the recommendation is OK, but he wants a few choices? Many of us are like that. How can it show him 3-5 options but also why the top recommendation is really the best? And on a smart watch sized interface!
    • What if he wants to know WHY the algorithm considers this the best choice so he can be confident it hits the right buttons (e.g. Is it Jenny’s favorite dish? Is it what they had on their first date?)? How can it show him these details without requiring a whole bunch of navigation or a huge display on his wrist?
    • What if the algorithm is way off base because it is using some profile attributes that don’t apply to the current situation? How can it let him adjust the attribute weights or change the data in some of the attributes?
  • In the presentation, the app recommends butternut squash, which happens to be perfectly ripe in the store.
    • What if it isn’t? What if the system thinks the produce is ready because a less than ethical produce manager said it was? Does it have to evaluate the store’s Yelp ratings and how could it use that information?
  • In the presentation, the system picks a delivery time by synching with the shopper’s calendar.
    • I don’t know about you, but I don’t put everything on my calendar. What happens when my secretiveness breaks the algorithm? How does it implement an interactive system to negotiation a new delivery time on a watch?
  • In the presentation, the seamless payment and delivery assumes (as do many automated self-checkouts) that the shopper didn’t stuff something in his pocket. Of course, that can happen anywhere. But it is a problem that the system does not address.

None of these possible snags show up in the presentation. It would be very easy to assume that the system is perfect if you just watch the presentation and the seamless user story it tells. And I am sure that there are many people watching presentations like this who don’t have the user experience background that you do who can see through to the potential snags. They are easily fooled. That is why these kinds of presentations are so dangerous. When the path through the forest looks perfectly clear, there is often a wolf hiding behind the nearest tree. There is a patch of quicksand directly in front. And there are alligators hiding in the stream.

Your Turn

Does this resonate with you? Have you ever been fooled by a seamless user story presentation? Share some stories with us.

Image credit: beofonemind

3 thoughts on “The Seamless User Story is the Enemy of Good Design”

  1. Used to happen all the time in the speech technology world. Back in the 90s and even the early part of this millenium, people would show beautiful little pre-recorded demos of speech recognition systems performing without a hitch, and many, many companies bought the technology, not realizing the problems they were going to experience because automated speech recognition was not ready for prime time at that point. So sales and marketing often contribute to the problem. But if people went to “live demos” at conferences, they’d have soon seen how “buggy” the technology really was – the demos often fell flat and didn’t work. I remember one case in particular, when Microsoft was bringing out its speech recognition developers’ system….big conference, big marketing, with ads like “voice-enable your web application in 20 minutes.” The demos at the conference (even taped beforehand) were really awful…and those of us in the speech human factors business just shook our heads and lamented in advance that there would be a LOT of nightmare speech applications coming down the pike….and there were.

  2. I did not view this as a seamless user story. My first reaction watching the presentation was, “Why do I even need to go to the store?” If my primary interaction with the store is deciding whether or not to buy the ingredients for a recommended dish, why do I physically need to be there? I do not know how to pick out a perfect butternut squash so why not leave that up to the experts when they pack up my order for delivery?

    The presenter mentioned looking online for grocery store pain points, but I don’t think he mentioned observations. I went to Whole Foods on Sunday (while traveling as we do not have one in my hometown) and observed an interesting customer flow. The store was crowded, but the crowds were only at the prepared foods areas. I saw 15+ people in line for the buffet breakfast, but during my trip I did not see anyone buying typical ingredients to prepare such a breakfast (e.g., eggs, bacon, pancake mix).

    Approximately 8 people were in line for the prepared lunch/dinner foods and I watched people buy pre-cooked salmon, pasta, turkey, roast beef, and vegetables. I myself bought glazed brussel sprouts and rosemary sweet potatoes because it just seemed easier than making something myself when I got home. However, when I went to buy some raw pork chops I did not have to wait in line at all.

    Based on these limited observations, I wonder if the future of retail is not to recommend a recipe for a fish stew, but to recommend the delivery of a fish stew prepared by the grocery store.

  3. Your analysis of the design considerations with UIs that provide recommendations is spot on. I’ve worked on such applications and these are sorts of issues we spent a lot of time addressing in user research, design, and evaluation. A more rigorous demo would cover the key attributes of any successful recommender system and demo how this system embodies these attributes. Unfortunately, such a demo would be the enemy of good marketing.

Leave a Reply

Your email address will not be published. Required fields are marked *