John Howrey
Design Leader

Crystal Clear

Bio, Resume, and Links

Product Design Case Studies
Watson Talent Admin
Prospecting Hub
Watson for Education

Communication Design Case Studies
Joanna Czech
Impact Arts
Drama Nerd Designs
The Musical Theatre Map

Theatre Posters and Art

Conceptual Work and Sketchbooks

Watson Talent
Self-Service Admin

We created the Watson Candidate Assistant Self-Service Admin tool because we wanted to build a road to success and a tool that actually helped people.

We knew that the programs available for recruiting, hiring, and training were lacking—and that finding a solution would take more than just a few years of development.

“This is the best chatbot manager I’ve ever seen…and I’ve been looking around.”
When it comes to recruiting and hiring, there's always been a lot of paperwork involved—but it's not just about filling out forms anymore (or at least it shouldn't be). It's about helping people find their dream jobs and make the most of their talents. That's why we created Watson Talent: the first fully self-service administration tool for Watson Candidate Assistant.

Our customers were having to wait months for installation, and we knew we had to step up our game if we wanted to keep them happy. So we proposed a new way of doing things: one that put our customers in charge of their own success stories by providing a more app-like experience than an enterprise approach. By using IBM's 9 experiences framework, we developed a new getting started experience that walks through each step in plain language and illustrates every detail along the way. We cut down installation time by over 90% while also refining design details.

It started in a meeting with the Offering Manager for Watson Candidate Assistant. She was telling me some story about her product, Watson Candidate Assistant. I had just started with Talent Management Solutions and was getting to know the products. The executive leadership had told me that they’d expect a focus from me on this growth area.

I don't remember the specifics of her story but it revolved around the fact that the development team was handling the individual implementation of their product for each and every customer.

I was floored! I started to ask questions. "You mean that we're training Watson for everyone?...

...Is that scaleable?"


I know that she'd thought of all this before. In general, though, our products tend to be SaaS in name only. Everything is so customizable and so unique and special and unicorny that while it is technically possible for your to get your own software up and running, it requires so much internal knowledge and insight that it's really not possible without great effort and assistance from IBM—and at a great cost to IBM.

The offering manager agreed with me that it was a problem but, I guess, she felt stuck. The business was demanding more and more of the 3-person dev team and we could have them building an admin too. We'll get to it later.

My point became, unless we make a self-service admin tool, there won't be a later. Each new customer is actually a nail in the coffin. We can't sustain a business where is costs us more to implement the product than we get in payment.

And looking at the market, we're not alone in this need. Today, many of the IBM products take weeks if not months to get up and running. Beyond up-and-running for the customer, that’s our time-to-profit.

I said "what if we change what up and running means?" What if we can hand our users the keys? 

Everything speaks. From the minute they hear the product name, they are experiencing something. And we need that entire thing to be amazing.

Sometimes, at IBM, I would get a blank stare. But this time, her eyes were full of excitement.

So we started building an understanding of what the keys would look like. 

I knew that this couldn't be just a design and OM journey. And now would it actually be about just this one product. As part of the Watson Talent Ecosystem, we needed to create a cohesive experience. Set up all of your products at once. Oh, btw, you should buy all the Watson Talent products.

So, we started creating a multi-headed hydra—a truly cross-functional team that hadn't been attempted yet in Talent. Beyond Design, Development, and Offering Management we started to include support, services, and sales. Marketing. Customer Success. Contract Extensions. All the people that are creating experience.

We broke it down into component parts. Obviously, we could't get every product running on new sprint immediately. We also needed to align on The Who. The why, and the what because we could start sorting out the How.

We held workshops. And design sessions. Bringing many voices into the mix and aligning around users and needs and opportunities. We could even do some light scoping.

We developed some users stories and, most importantly, a set of criteria and experience philosophies to guide the work. These have stayed at the heart of our decision throughout. What's great about a design foundation like this is that we know that we can always return to the basics. And we have a clear cut way of quantifying the success of the design.

We iterated a lot and early. I brought early sketches with me — just one simple user story — to a User Group Meeting and grabbed people in hallways to get their opinions. Overall, our concepts were resonated. And resonating with customers strongly enough to start to impress more internally.

We organized the designers into teams with a dual focus. They were building the specific components as needed by the different products—after all, the engineers were still going to be siloed and we needed to be collaborating with them on a day-to-day process of delivery. But we also need to work together on common design elements like shared patterns. I instituted a new sense of collaboration and critique on the design team. Over the past year, we've gone from single designers designing in a cave to a more studio-like approach with designers talking to each other and sharing in regular critiques—even though our team is split by the Atlantic Ocean.

Another big change is the focus on storytelling. This team had been talking through their design as a series of notes about buttons and form fields. For this work, we talk in terms of story. We start with a user need and never stray from that user’s perspective.

We reached an MVP stage before I left IBM, with a full getting started experience being live for Watson Candidate Assistant. It is proving to the business that this is a viable plan — and we're proving it with response from customers. We're also seeing that the WCA developers have more time to build more stuff. They are spending 75% less time working on customers and converting all of that time to new product feature. (There's still a little data enablement that they help with as our common data approach comes to life.)

I’m making this summary sounds really lovely. I won’t lie: it was incredibly difficult. We never got designers officially deployed to this project. There have been very negative opinions about it. People have blocked efforts to get alignment saying that our products aren’t supposed to be self-service and won’t be. 

We are so used to making admin experiences that work for our developers. This is really changing that up, I know. The process is also changing things up.

What kept us going was this: since the launch of the MVP Self-Service Admin, there’s not been one complaint, one service or support call, or anything. Not one call.

What's next is the part where the experience comes together under one umbrella. We have the concepts and the alignment on a user-centric vision. Now we're just waiting for the resources—and they're freeing up before the end of this quarter.

This all coincided with the early draft of IBM Carbon and we leveraged that framework to make sure we were putting our best foot forward. It also introduced our portfolio to the way of the future and converted members of the dev team to be UI framework believers.

“I wish more things had that type of workflow for being able to set something up. It was amazingly simple.”

“I don’t need to know web page design to get this done, looks good regardless.”
Design and Product
Strategy Creative Direction
Information Architecture
User Research