Design Thinking that she’s even visiting the Design Thinking Camp in Sofia right now (have a look).
This workshop’s main purpose was a team-wide intro to design thinking. Of course we picked a really interesting problem to solve: How can we contribute to Qualtrics (a recent SAP acquisition) surveys in new and creative ways? So the task was to think outside of the box and think about “filling out” surveys in new, exciting ways.
At the end of the day, the two teams presented two really interesting prototype ideas. These ideas had gone through a lot of ideation, thinking, discussion and finally were presented to each other in the form of a theatre play that we recorded for ourselves.
While we initially did not plan to implement these prototypes, I simply could not stop thinking about the idea that our team had. A few weeks later, over a weekend, I therefore came up with “Millow – the feedback pillow”. Crazy, but read on what it is and how it came to be!
Design Thinking 101
I am in no way an expert in design thinking (Denise is!) but I participated in a few workshops now and can try to outline the structure of such a day for you.
As you can see, there are three phases:
- In the discover phase, it’s all about understanding the problem to solve, getting onto the same page as a team and understanding the persona that you chose. Persona? Just a fancy word for the representative person we’re trying to solve the problem for. In our case – I roughly remember – we picked a 30ish woman that likes to shop in physical retail stores, she’s up to date with recent tech (smartphones etc.) and lives somewhere downtown Munich. We gave our persona a name, we drew her picture and we then interviewed a person that came closest from the other team to learn more about her needs. We finally even went down to a local supermarket and interviewed random people about their opinion about giving feedback while shopping.
- The design phase covered the other half of our workshop day and is all about brainstorming ways to solve the problem at hand, then agreeing on a good potential solution and validating this idea. We began with a couple of brainstormings (there are funny techniques for this, e.g. to warm up and to get into a creative mood), then grouped the ideas and finally voted on which idea to sketch as a prototype. We presented these ideas to each other via short theatre plays.
- While our workshop was now over, the deliver phase would now implement the idea, test it and finally deploy it. This can often take a longer time, depending on how challenging the technical solution is or how committed the people in the team are.
During our research, we learned that our persona is in general happy to give feedback, which would allow the retailers to improve their processes. But very often, there is simply no way to leave the feedback that moment. And it is not an option to give feedback minutes after an experience, as very often it is forgotten, time pressure kicks and people are in a rush to leave for work/home or they simply do not believe that feedback would be taken seriously.
We came up with a system that would be installed a couple of times in a retail store and it would facilitate the giving of feedback. It would also lower the barrier of giving feedback in a public space, make it a fun experience to break the ice. We initially thought of a boxing ball that you kick to start the feedback process via some kind of voice system. We also discussed puppets that you would press or pull down to start the feedback – one for good feedback and one for rather negative feedback. In the end, we thought that having a kind of pillow that you hug for starting the process would be best: the feedback pillow.
This is – roughly – how we meant it to work:
- The person having an experience that she wants to give feedback about, simply hugs one of the feedback pillows installed in the store.
- Via built-in loudspeakers, the system will say high and ask the person to leave her feedback via voice, which a service in the cloud converts to text, analyses the positivity/negativity/relevance and captures the feedback via the Qualtrics API.
- In case non-relevant feedback was given (our semantic text analysis would discover that) , we kindly ask the customer to start over.
- In case the feedback was relevant we would now reward the customer instantly – via some QR-code or NFC tag based mechanism. The pillow would thank the customer and a display shows a QR linking to a web page, where the customer can decide how she is rewareded. We believed that a donation to charity would best fit to our persona.
Not quite the normal way of filling out a feedback survey, right? That’s what happens during a design thinking workshop!
The weekend and the (partial) implementation
A few weeks passed and Anja from our team brought up the topic that we had two nicely developed ideas as an outcome of the design thinking workshop – but no implementations. I think many in the team thought it would be nice to at least investigate the implementation a little further.
That weekend, I realized that the Google AIY voice kit for the Raspberry PI and a pixl.js development board with a display include all bells and whistles to get a basic protoype technically implemented. And pillows… are sold everywhere. I went across the street where I live and bought a cheap pillow for EUR 2.50. I opened it up, lasered a stencil, sprayed some text onto it and attached all hardware components via stitching or glueing. It was making at it’s best and really fun!
On the software side, I essentially prototyped many parts of the system but did not connect the pieces. Millow, the feedback pillow, starts talking when you press the center blue blinking button. It will then listen for what you have to say and return the text it understood for furhter processing and interacting with the Qualtrics API. On the pixl.js, I was able to use one of the integrated buttons to start the display with QR code that can be scanned by a smartphone. I even programmed the included NFC chip to link to the same URL, so a customer can scan the QR code or touch the tag to open the web browser. As part of the research, I also wrote a python script that would connect to the pixl.js via a BLE connection and change the code (e.g. later this would change the generated QR code).
Thoughts on giving feedback and implications on customer experience
A pillow in a retail space to leave feedback? Yes, I know. That was the idea of our design thinking workshop and I had tons of fun realising a basic implementation of it. But really, a feedback pillow? Probably not, although it would drive me into a store if someone really did that…
But the idea of providing easy access to feedback systems and making it a fun, instant experience is a good one – I think. We also did a bit of research and we’re quite sure that many people want to leave feedback, but often there is simply no way to leave feedback. Also, the use of voice in combination with automatic analysis would provide a pretty easy, non-conventional way of leaving feedback.
So maybe such a feedback system will have not the form of a pillow. Maybe we could integrate it into retail racks? Every few meters, close to the products or areas you are when you want to give feedback. Maybe it will not use lot’s of electronics in the end, maybe we simply use QR codes and NFC tags to start the feedback process and use the customer’s smartphones for the user interface. Would it still make sense to use voice in such situations? Would you rather type?