WorkIBM Digital Business Assistant

Imagine that you have an assistant that helps you focus on your daily work and task at hand. The assistant is out there staying on top of your business-critical workloads. If something comes up, the assistant proactively notifies you of the situation, recommends actionable steps and drives you towards successful task completion.

Following video provides a quick intro to the project.

Design Challenge

Our biggest challenge was to design a product that was true to our target users. We were aiming to design something so simple that even a user with very little technical skill could start using the app right away on their laptop or mobile device. This proved to be a challenge for our larger team, as most of the stakeholders would tend to saturate the app with a lot of features, which were, of course, valid, but took us further away from the user target we were designing for. As a design team, it was our job to advocate for our users and keep the product team on task with meeting user needs.

My role

As the UX designer responsible for innovation and end-user experience, my role was to drive cutting-edge solutions to deliver a delightful experience for business users. Within my design team, I cooperated with Design Researchers to uncover the real needs of users, after that, I worked with three other UX Designers to establish a cross-platform experience consistency. Eventually, I worked with Visual Designers to make sure the experience also looks ∓ feels excellent.

Within my wider squad team, my role was to ensure the delivery of a delightful MVP on time. While the engineering team focused on building their codebase, I went through a series of sketches, wireframes and user testings to have a solid design base. Following this, I made sure the whole team is in sync by avoiding the waterfall process and showing my work regularly instead. We would have a daily scrum and a weekly 3-in-box meeting, where we would go over the designs and progress of the code. A lot of those meetings meant convincing the engineering team to do a pixel perfect implementation, while many times we had to negotiate and postpone some features for MVP+1. If we had to delay something, my task was to make sure that at least a tiny part of the already minimized experience is included in MVP.

Research

This project started from scratch, without any legacy software, and that is unique within IBM. The scope was rather ambitious, and before we organized a design thinking workshop, we dived into research to better understand the problem space.

First, the entire design team helped our design researcher with intense research on who the users are. After understanding the users, we created personas. We then reviewed our competitors and created a good mental image on which direction to go.

Persona 1 - Amy

We found out that most of the business people had very little time to deal with technology and they would hate setting something up. They would expect the business software to work seamlessly - like any other consumer app they use on their phone, otherwise, they wouldn't use it. We named this kind of a user - Amy. She became a retail manager who was responsible for multiple retail stores.

Amy's main pain points were:

Persona 2 - Otto

While most of the business users don’t like to tinker with technology, we discovered there is usually a member on the team who loves it. We called him Otto. Otto would explore new tools and technology that could improve his work life. He had no IT background, nor any coding skills, he was just excited about new tech.

He would say: “If it takes 2 hours to format the spreadsheets manually, but 30 minutes for writing the macro, then I’ll do it now and run a macro in the future to save time.”

We found that Otto would advocate for new software within his team and he would be willing to set it up too in his free time.

Persona 3 - Maureen

We understood that there might be a need to have a developer involved in setting up the Assistant for larger companies. We took the persona from another team which used the same persona. Later on, we decided to spend more time on Persona 1 and Persona 2, to define the experience at first, before diving into the developer world.

Workshop

The design team was spread across four different countries and three different time zones. Many of the engineers were located in Paris, France and many of the project managers were in Austin, TX. To us, it was crucial to set a common goal. We held a strategy face-2-face workshop with the whole team to align them on the problem space, as well as brainstorm the best possible solutions. Our team followed Enterprise Design Thinking by IBM, which is an iterative process focused on observing, reflecting, and making. You can learn more about it here.

Because we conducted user research in advance, we were able to drive a shared vision for the product that both met the user needs and that had a feasible direction and end goal.

Workshop Outcome - Hills

We narrowed down the scope and came up with two hills. Hills are a key to Enterprise Design Thinking by IBM - they help the whole team march in the same direction and you are quickly able to reflect if you succeed or not.

After the workshop, we created a Spotify tribe model, in which I lead a design for an End User squad. In my squad, there were 2 Product Managers, 1 iOS Developer, 1 Android Developer and 2 Front-End Engineers responsible for the Responsive Website. We had a daily scrum to make sure we all work on the same hill and at the end of each sprint, we playback our progress to the whole tribe. Down the road, my close collaboration with Visual Designers, other UX Designers, the Engineering team and Product Management team ensured a successful pixel-perfect implementation of the designs.

Wireframing and Prototyping

Based on my expertise and a strong interest in consumer apps I was assigned to Hill 1 and got to lead the end user experience. I gathered the personas, pain points, user needs and hills; after that, I started sketching the initial designs for the Assistant. I began prototyping different solutions that ranged from a full experience app that would be used on a daily basis to a minimal app that would run only in the background.

Initial Sketch of the full-blown work tool

The first design was a full-blown app, that would intelligently work with all your tasks, files, projects, emails, calendar events, social media and news. It would be so powerful that it would become the primary tool for business users. The team liked the direction and the fact that the tool resolved all the boxes of pain points, but our management realized that it could compete with the IBM Business Process Manager. This resulted in me being asked to iterate on it and come up with something smaller.

The first Iteration of a smaller mobile app

The first iteration focused on an Assistant notifying business users on their mobile devices. They would be able to download an app and choose from a wide range of business skills. This one became the preference of the whole team. The users liked the design and the idea, so I started a 1-year journey with it.

Number of further iterations

Based on user and stakeholder feedback I went through many iterations to find the best User Experience. I tried anything from a WhatsApp style to a chatbot conversation and later on to a feed.

TOP 3 Design Decisions

1. Mobile First Experience

The project management team wanted to tackle individual business users and mid-sized companies. From our research, we knew that those users spend a lot of time using their mobile phones for work. Therefore, I advocated doing a complete Mobile First Experience, which would lead to a mobile app on both the Apple App Store and Google Play Store. That would ensure a consumer-like experience with a smooth, easy interaction!

"We will make a responsive website only!

    - the initial reaction of the Lead Developer

The team had very little knowledge of native apps and had a hard time seeing the value. I spent time with them explaining why it made sense to build a native app instead of relying on a responsive website. The team understood the need and hired iOS and Android developers to work on native implementation.

2. Twitter-feed-like-experience

This decision goes in hand with the previous mobile first decision. I iterated on a number of interactions ranging from a chatbot to a dynamic Twitter-like card experience. Based on a variety of skills, that could be activated for the Assistant, and the fact that anyone could create skills, I managed to convince the team to create a feed experience with infinite scroll - for both Web and Mobile.

3. Prebuilt Skills

The team assumed that Otto would always be the necessary middleman to create some skills before Amy would be able to use the Assistant. However, that would make the app useless and even Otto would have no idea of what can be created. Therefore, I strongly advocated having a set of IBM prebuilt and business partner skills, which would be publicly available so that anyone could start to use the app immediately.

Project Demo

The following video shows what we delivered for an MVP. What you see in the video was built natively for iOS and Android - though Material Design inspired the Android app.

Future of the Project

What initially started as a concept and as the “new breath” in business management software became a hot topic at IBM. The project was publicly announced in March 2017 at the IBM Interconnect conference in Las Vegas and got a lot of attention (read more here). After the conference, we successfully started the open beta of the web app, together with both native iOS & Android app.

Talk of Jim Casey, Lead Product Manager, at IBM Interconnect 2017

"We are the first digital assistant configurable by you to work with the tools and data you use to get your job done."

    - Jim Casey, Lead Product Manager

A couple of months later, many teams wanted to cooperate with our team, and the essence of this project was eventually absorbed into the award-winning IBM App Connect. After a year of hard work, it is good to know that their team will continue using our designs, ideas and approaches to address the needs of our business users.

My 2 cents

To be honest ... this project was my baby. I enjoyed working on it because it was a new product with a vision to disrupt how business users work. It had to be damn simple, without 100+ of customizations for every kind of business and use case. There was ONE use case and that is it. This project turned out to be one of the most challenging and at the same time rewarding moments at IBM, as not everybody was onboard of doing something simple, so it was a designer's task to open everyone’s eyes.

This project was the reason why I relocated to IBM Design HQ in Austin, TX, USA.

UPDATE 2018: I decided to share a blog with fellow designers about: How not to go insane when your epic project is killed 💔 as a story about this project.

External Links

 

Explore next project: IBM RAAS