A Kiwi goes to Dibi 2017


TOBIAS AHLIN / Lead UX Designer at Minecraft

With so many amazing conferences to pick from this year, it wasn't an easy decision to choose just one, but I’m pleased to say Dibi 2017 did not disappoint. Here’s a journey through my time at Dibi and my thoughts and takeaways.

It was my first time in Edinburgh, had to take a photo.

At a glance (the conference in a nutshell 🥜)

The conference was held at the Edinburgh International Conference Centre, which is pretty central in the city, easy to find and get to, and I stayed about a 5 minute walk away. They had a good mix of speakers, catering was pretty good and it wasn't too busy, so there was opportunity to grab the speakers for a chat in breaks, which is always a plus.

Check out their website for more details: dibiconference.com


First up was Joshua Davis, this guy is a bit of a legend, his work was incredible and his attitude was refreshing. He creates what he likes, how he likes. He designs in code to create amazing dynamic digital artwork, most recently getting into projection installations and working with some heavy hitters in the music industry; Bruno Mars, Taylor Swift, Nine Inch Nails, Pharrell Williams, Snoop Dogg & Phantogram to name a few.

A cool little tip he noted: To inflame his creativity he would aim to produce a few concepts but impose a few rules around them, some limitations, such as, they can only be black and white, only use shapes etc.

Here is some of Joshua’s work

Phantogram — Fall In Love

He codes these programs on top of other programs that change and react to the music input. The shapes shift and patterns change as the tones change.

Taylor Swift — Shake It Off

Quotes that stood out

“Take risks, you’ll fail, who cares, try something else!”
“Shut the f😱 k up and make something!”
“Can I kill people with my crazy artwork?”
“If you can get people to sleep in your artwork…that’s comfortable.”

Next up

Mike Kus spoke mostly around the web design industry as a whole, how to stand out and the threat of builder systems like WIX & Squarespace.

“Be the black sheep”

A lot of the web is starting to look the same, so many designers just follow the leader with their work, at the detriment of their clients and those end users.

He spoke about his process for extracting the wants and needs of his clients, establishing real empathy. He often has multiple meetings, interviews with the stakeholders and creates word maps in ideation, involving the clients in ideation sessions to really draw out ideas and fully understand their needs and the needs of their users.

“Style can convey ideas and concepts in micro seconds”

Mike spoke about the skill of helping stakeholders see the value in not just following template systems or what the “other guy” is doing but really striving to convey their message through good design, as many of his clients did not even know what they wanted.

He also spoke about going the extra mile to produce unique work, hand crafting custom fonts for some clients, creating his own textures and photography and using visuals to convey an idea and message.

Don’t be afraid to do something new or unique, produce what’s best for the message, purpose or result, not just what’s popular on Dribbble.

Next up

Molly Nix, a Senior Product Designer from Uber. Molly spoke about empathy, risk and inclusive design at Uber. It was great to hear how her team work with their product on an international scale, really learning and establishing localised patterns across countries and cultures, much the same as we are doing at Schibsted.

She spoke about how it’s vital to have real empathy and understanding from your users, no matter where they are. When designing, you really have to get out from behind your desk and use the product you’re working on — everyone at Uber must experience being a driver and spend time using the app to establish empathy and context, which is a fantastic way to do this.

“We’re lucky working at Uber, push a button and a user turns up”

We do the same thing at Schibsted; we all interact with our apps and experience actually using the system instead of just conducting user interviews or user tests, or solely relying on our researchers to advise.

Molly had an example of a problem that can come about from not fully validating work internationally and only locally. Below is a screenshot of the drivers app in uber, highlighted is the earnings tab, the issue here was in San Fransisco this tested well and was a sound, validated solution — people understood this was a bar graph which related to earnings.

However, when the team took this to India they found an entirely different environment where internet connectivity was a huge pain point for their drivers and almost all test participants thought this was an icon for internet connection bars.

“We’re adding humanity to the UX process”

Molly had a lot of amazing work she showed around the products she works on at Uber — self driving cars and tackling bigger social issues through design and technology, such as safety and congestion by utilising AI & environment mapping technology.

Below is a slide showing how Uber are trying to get more users using uberPOOL to do exactly this, one of the big UX issues is establishing expectations around uberPOOL vs uberX, uberXL, Executive etc.

She spoke about how they are encouraged to fail fast, learn and evolve.

Another cool take away for me was an ideation technique Molly touched on which involved improve, called “Body storming”.

“For building happy and healthy design teams”

Here are some examples of body storming in action:

“The central tenant of improve is “Yes, and..”

The Uber Eats rollout was an interesting topic; Molly spoke about how it was initially within the Uber app as a tab option, you could only select one meal and it would be with you in 10-mins within a contained area. This grew to add more and more options as the usage grew, until finally it became a stand alone app.

“We’re encouraged to fail fast”

It changed every few months based on constant learning, understanding and iteration, even doing 5 full releases inside one and a half years.

Next up

Tobias Ashlin, currently a Lead UX Designer at Minecraft, ex Spotify & GitHub. Tobias spoke about his past experiences with some great examples and stories to back up his topics. What stood out for me was his comments around designers validating/testing their work in hugely biased ways to prove their designs work, and thus contaminating their own results.

“Let’s prove we’re right”

He also touched on characteristics of good judgement, which aim to help designers shape their behaviour for the good of the users and their products…

…striving to be fair and objective, weighing proofs for and against, not giving simple easy way out answers to complex or large questions, being open to being wrong as new evidence is presented and not taking feedback personally.

Next up

Vibha Bamba is a Senior Product Designer at Airbnb and spoke about ratings, context and expectations. This was another really interesting talk, similar to the localisation take aways from Uber, and was super relevant to my team at Schibsted.

Airbnb’s 5 star system: 1, 2, 3 treated as negative — 4 & 5 are positive.

Airbnb’s rating system seemed straight forward and simple; 5 stars is a very common design pattern but collection of relevant feedback for hosts so that they can actually can take action on issues, was a real problem. Vibha showed us a few user interviews with hosts who were extremely frustrated and confused, some were pretty angry about unhelpful feedback, but also left guests with weird reviews to view on their profiles. The reviews often did not make any sense and looked wrong, completely contradicting one another.

For example: 2 star ratings with positive reviews? 4 star ratings with negative reviews?

A 2 star rating with positive review?

y tho…

After interviewing many of the users who left these ratings and reviews, it turned out that many of these users wanted to look “nice” or “positive” on public facing reviews, so would leave positive reviews or “nice” reviews but then give lower star ratings. The reviews would often be fairly unhelpful as actionable feedback for the hosts, and obviously look really odd. These issues also seemed to come out of guests not exactly being sure of what to review, or how to frame the feedback, they didn't want to seem overly mean or harsh so just put something polite and expressed negative feelings in a more global rating.

Hosts would see a 2 star review and be given no help as to why it was 2 stars; how can I change my offering and better my service if I’m not sure why they did not like my place?

Another interesting and common example Vibha gave was of a host who had a location just outside of the city centre — her price was lower to reflect this and she offered a clean but fairly basic room. This host was getting extremely frustrated by the unusual reviews and ratings she was getting — some 2 stars and some 4–5 stars, contradicting reviews / ratings and it was costing her bookings.

One user booked the room months ago and it turns out was having a bit more of a holiday on this trip and was not happy about the location and a few other little things.

The positive review was left by a user who only needed the room for one night and booked it short notice, so was happy with the location and quality.

After doing a lot of research around these cases they found the core issue was expectations. Also users did not know what to give feedback on, so they would leave feedback that was just unhelpful, things the host could do nothing about (why don’t you have a garden or a pool?), usually because they had no direction as to what to review but wanted to write something.

So Vibhas team came up with a new system that allowed guests to give more useful feedback with tags, hints and ratings around specific contexts (cleanliness, location etc.) but also expanding on issues where appropriate.

For example, shown here:

Cleanliness — “What were the main issues?” “Towels and linens” — now guests have a way to explain what they really think more accurately and hosts have actionable feedback.

Obviously this is an evolving design problem for Airbnb, like most things are, but as a case study it really nicely highlights the importance of research, investigation and having empathy for each of your users, getting to the ‘why’ behind the problem.

Next up

Chris Hammond, was talking about IBM Design Thinking; the challenges with shifting the focus of such a huge company from software and engineering, to instead being focused on user’s needs using Design Thinking.

Kicking off his talk, Chris had some people come onto the stage to prove a few points and illustrate Design Thinking ; he asked them to design a vase…

“Design a vase”

everyone drew basically the same thing. He then said, “I’m going reframe the request, as we focused on what has happened before: we focused on the result and not the problem.” He said, “Design a better way to enjoy flowers”. Then the volunteers started really concentrating and came up with all kinds of different solutions. It was a cool ideation example and a great way to kick off the talk.

“Design a better way to enjoy flowers”
“Reframing the problem”

He then took us through how Design Thinking was iterated on, how it came about, how it was implemented. He spoke about buy in — how fundamental it was to have the CEO on board, and with her support they could onboard a very engineering focused company onto a new way of thinking.

“Risk is all about learning fast”
“Curious = humility with ambition”
“Grit = Perseverance with passion”
“Empathy = First for ourselves, then for our users”

Chris also had an awesome case study he shared which was a pro bono initiative IBM did for a healthcare system that needed a lot of assistance which was cool to see.

Next up

David Bailey & Nikos Tsouknidas, from the BBC. David spoke about his career in design, where he started The Designers Republic, before working for himself at an agency called Kiosk and how he eventually came to be the Creative Director of the BBC UX&D.

He spoke about creating a design culture at the BBC — how the teams move around from product to product to keep them from going native and getting tunnel vision. So they move their designers from team to team to keep everyone fresh and interested: iPlayer > BBC Sport > BBC News > Weather etc.

“Stop going native”

Global Experience Language

David then moved onto GEL which is the BBC’s Design DNA, which was awesome to hear about as, at Schibsted, we are currently creating our Global DNA.

A little look at the BBC’s GEL
Relationship structure

Nikos spoke about designing with developers and the relationship between the GEL guidelines and the living DNA environment. The flow between the GEL website <> Component Library <> GitHub.

Next up

Madeleine Kroper & Stef Huber, from BCG Digital Ventures — who had a great talk about the differences between Start-up vs Corporate and the cross overs from each that successful teams adopt.

They had some amazing examples of experiences they had with many different clients when it came to company culture, changing strategy and explaining different values to some stereotype personalities — an example below called the “Silverback Problem”.

Madeline & Stef spoke about shifting thinking at some corporates, with old waterfall techniques building bloated products, and how they moved them to working in a more iterative and agile way. One of my favourite quotes from this section is:

“If you’re not assumed, it’s not an MVP.”


Dibi was a great conference. It was awesome to hear about some of the experiences, struggles & successes of other design teams around the world who are working on some interesting problems.