Google recently launched it’s NFC-based payment solution, Wallet. For a company like Google, strong in engineering but less so in design, Wallet was a particularly risky product to build. You can’t half-bake the design of a consumer payments product. It has to instill absolute confidence among its users. Workflows have to make complete sense with a minimum of learning. You can’t do it the same old Google way.
In the face of those risks, the product launched to reviews like this from Greg Kumparak of TechCrunch:
Google Wallet is great, magical, impressive, and all sorts of other positive adjectives
How did Google do it? How did we get good design done in an engineering-driven environment? This article is a record of the strategies I used in my time as UX manager over the Commerce group at Google (which covered a range of products including Wallet).
1. We fished for champions
To get good design work done in a non-design-driven company, you simply must have a champion with decision-making power. If you’re a design leader and you don’t have a champion, you have to get one. Make this your single most important goal. It’s a rare design leader who can charm a hostile audience and bend them to their view of the world. Focus on people who have characteristics that lend themselves to being influenced by the contributions that UXers can make.
On Wallet, Google Commerce VP Stephanie Tilenius was an obvious person to focus on from a pure power perspective, but she also had other characteristics that made her a logical focal point for relationship development:
- She had a passionate desire to force Google to break out of the pattern of mediocre product launches.
- She had the intelligence to understand the value of good design in creating high quality customer experiences.
- She was new to Google… the engineers weren’t completely comfortable with her non-engineering-y leadership style. She didn’t appear to have many allies in the organization. She needed UX as much as we needed her IMO.
How you actually build a relationship with an organizational leader is the subject of another post, or book, but the short version in this case is that, in any communication with Stephanie, I emphasized messages that touched on the characteristics above. “Here’s how we’re going to avoid a Buzz-style launch”, “as you can see, this design makes anything else Google has produced appear average”, “here’s how we can use design review to improve product quality”, “you really need your first product launch at Google to make a big splash, and this design can do that”, etc. This was not about deception. It’s a repositioning of design work and design process to connect with the needs of an individual who is empowered to help you and your team get good work done. I genuinely wanted Stephanie to succeed and in turn for us all to succeed. Cultivating a tight relationship with her was absolutely central to our success.
2. We stopped talking like designers and started talking like locals
To get good design work done in a non-design-driven organization, you have to use the language of the organization. Using the language of design may be useful if you’re talking with fellow designers, or if you’re an agency pitching services to execs. But when you’re in-house, you need to talk in the language of the house. At Google, the house prizes the launch above all else. When communicating with engineers and PMs, I tend to avoid discussing anything that isn’t clearly connected to the launch.
On Wallet, I rarely talked to an engineering manager without expressing my anxiety/excitement/desire to get the product shipped. “Look, I don’t care about blablabla, I just want to get the best product shipped as soon as possible.” And this was true. It’s what I genuinely wanted. But I needed to make sure that my engineering counterparts knew that that’s what I wanted. Because we had this common goal, I had great partnerships with eng leaders Wall, Rob, and AZ.
3. We generated both excitement and anxiety about the vision
To get good design work done in a non-design-driven company, you have to exploit your story-telling skills to paint the picture of what is and what could be. Designers often forget that we have the power and skills to craft a product vision in a form more tangible to humans than a PRD. By prototyping and presenting the user experience vision, and getting engineers excited about the prospect of building such a vision, we create the conditions for our vision to become reality. On the flip side, if a user experience is less than adequate, we can use our story-telling powers to raise anxiety about – and action on – the inadequacies.
In the case of Wallet (and Shopper/Offers), we created – as most project teams do – a presentation of user interfaces outlining the core use cases. We mocked onboarding flows, transaction flows, and movements between applications. We highlighted the clunkiness of app transitions. We highlighted the unnecessary inconsistencies resulting from lack of coordination across disparate project teams. We also highlighted how frighteningly simple and straightforward the Wallet experience was. The presentation told the complete story – warts and gems – of the mobile commerce experience. I delivered it weekly to Commerce leadership. Critically, we printed the presentation and pinned it on the wall in Stephanie’s office, updating it as new designs came to hand. She was able to see the good, bad, ugly of what we were building along with our post-it notes and commentary. We got her and the rest of the team simultaneously excited and upset about our direction. This was a huge help in getting the team focused on solving the most problematic wrinkles in the experience while protecting the good stuff.
4. We didn’t short change visual design
To get good design work done in a non-design-driven company, you have to acknowledge that many Silicon Valley people think that design is only about how products look, not how they work. “You guys make our stuff look pretty”. You can use this to your advantage. To entice executives and engineers/PMs to view your team as a credible design group, present visual design work that blows them away. Wireframes are nice, but high fidelity mock ups are much better.
On Wallet, we were short on visual design talent. Jonathan Yu had his hands full with interaction design work. The talented visual designer Sunkwan Kim was spending more time on Emerald Sea and less with us. We had zero visual design support. I had some contacts at a New York-based design agency. I called them to see if they were available. Coincidentally, the same agency was supporting our marketing team. I put the enthusiastic Chris Nesladek on point for directing their work. As you should expect with an agency, the quality of their work was variable, but the good stuff was excellent. They were able to work with Chris’s direction to develop a tight visual design system across all of our mobile applications. The visuals got our stakeholders excited. More importantly – right or wrong – their visual design work helped raise the perceived quality and importance of our collective design work.
5. We aired bad blood, fast
To get good design work done in a non-design-driven company, you have to resolve conflicts fast. Conflicts between eng, PM, design will arise. They always do. In an engineering-driven company, UXers should prepare themselves to lose most of the battles. Certainly, in a rational company like Google, decisions are made based on merit. But decision makers – unconsciously or not – tend to agree with like-minded people. And that’s more likely to be an engineer or PM, than a UXer. The best you can do is to get all of the affected parties together to air their complaints when conflicts occur. In many cases, when you do so, everyone will realize how stupid the arguments were and will all go back to work. But there will be legitimate gripes. Have everyone air their thoughts. Go around the table. Write down the grievances and start to whiteboard ideas on how to reach a compromise. Show that you want to make allowances. People will reciprocate.
Wallet had some detailed arguments about the fundamental direction of the product. Was it going to be a consumer product or was it a system utility? These decisions would fundamentally influence how the product would be positioned from both marketing and design standpoints. The consumer angle eventually won out, but not without some disappointed people. The good news was that those people got the opportunity to voice their disagreement, were able to make their arguments, and did so to the group leadership.
6. We got the right people on the right tasks
It goes without saying that to get good design work done, you need to have solid talent. But talent isn’t enough. You also need to manage and allocate that talent effectively. Huge projects like the Wallet/Shopper/Offers combo cannot be done by one designer. It takes a team. For that team to be successful, you should have some knowledge of each team member’s strengths, and you need to be opportunistic in assigning those strengths appropriately to get the work done well. This is where the role of a design manager becomes more like a baseball manager. You need to know when to sit your starter, when to bring in your pinch hitter, when to warm up the bullpen, and when to ask the GM to recruit a slugger … if you have a GM. Wallet was fortunate in that we had the right set of personalities at the right time IMO:
- When I started in Commerce, Jonathan Yu and Sian Townsend were the UX team members on Wallet. Jonathan was a very solid, mature, all-round designer. Strong in interaction design, he produced many iterations of the core workflows while the Wallet team was determining the product positioning. Importantly, he had an optimistic attitude and was able to collaborate well with engineers to establish the core interaction and conceptual model for the product. Jonathan was the perfect person to establish a “beachhead” for UX with the Wallet team. Sian Townsend was a technically brilliant researcher who was able to boil her findings down to useful chunks for her audience. She had a keen sensitivity to people’s willingness to accept research, rolled with the punches and got work done in an agile spirit. Yet she was never afraid to raise usability or fundamental product issues that were overlooked. She was a voice of reason.
- Chris Nesladek had been coordinating with Frank Harris on Google Offers but as I spent more time with him, it became clear that he had the most expertise in mobile (coming from Android) and, more importantly, he was the only team member who pushed the broader organization to think about the cross-product implications of their work. Chris was a prolific, detail-oriented designer and had a passionate approach to asserting his vision. He was the ideal person to coordinate a unified design system across Wallet/Shopper/Offers. I recruited him to take on this role, and while it took some time for engineers to warm up to him, he was very effective in forcing discussion and resolution on the most critical user experience issues. He was the change agent that every challenging, multi-team project needs.
- I recruited Alex Cook to the Wallet team knowing that he would be the closer: the designer who could build a team to get the project over the finish line. Alex tends to take big hairy design problems, somehow comes up with rational solutions, then works closely with engineers or directly with code to close out all of the details to get it launched. Perhaps his greatest strength though is in building teams. And that’s what he did. As Jonathan and I transitioned out, Alex came in, recruited and oversaw the team of designers that got Wallet to launch.
So these were a few of my take-aways from the Wallet experience. Not necessarily revolutionary information here – this stuff isn’t rocket science – but interesting to see it applied in a real world context. I didn’t stick around to see Wallet through to launch. That is perhaps one of my biggest regrets re: leaving Google. Still, it’s fantastic to see such a relatively well polished consumer product delivered by a team at Google. It’s a rare thing. Against the odds, we got good design work done. All of us.
This is an edited version of a post that appeared on Graham’s personal blog.