Optimizing Call Centers

Services

Software

Time / Team

• User Research
• UX Strategy
• Stakeholder Management
• Design
• Prototyping
• Usability Testing

• Fireworks
• Axure
• Photoshop
• OmniGraffle
• Confluence
• Camtasia

• 2009-2012
• 1 Product Manager
• 1 BA
• 1 Designer
• 1 Engineer
• 6 Developers

My Tasks

• Redesign an ecommerce call-center application to be more user friendly, vastly more efficient, easier to learn, and cheaper to support
• “Be the catalyst for change on this team” - My Manager
• Improve the morale of a product team left to languish for many years
• Restore the reputation of our company (Art Technology Group) in the eyes of industry analysts and customers including AT&T, Restoration Hardware, J.Jill, Walmart, Costco
• Resolve concerns of potential buyers (Oracle)

My Actions

Researched the “Commerce Service Center” application, the users, and the history of the project from the team leads.

Relationships

The original platform

• Looks circa 1985
• Painful to look at
• Users hated it

Orange means ??

Copying doesn’t need help

Contextual data, labels, icons, buttons, all jumbled

Tabs occlude key features

Customers inside Commerce tab, Commerce inside Customers tab

63 icons!

Crucial feature

Duplicate feature

7 levels of nested boxes

Unnecessary feature

Bold is over-used

Useless feature

Phantom lines

Blue means ??

Green means ??

The feature set was vast and the learning curve was 9 months.

The visual architecture, UI and workflows were a mess.

The application sat on top of call-tracking and productivity software used by call centers so there were myriad integrations to consider.

The application had been built through acquisition and the original architects were no longer at the company, so no one knew how it all worked, which made any change a daunting prospect.

Everyone wanted change in theory, but in practice no one had the skills or the confidence to know how to proceed.

So…

To get past the team’s fear of failure and change aversion, I suggested we go on site visits to cultivate customer relationships and gain deeper understanding of their pain points, ideas, processes, etc.

Watching call center workers handling angry customers all day, every day, we developed deep empathy for people and renewed commitment to making their lives better.

Now that we’d all witnessed how much users suffered, we all felt an intense commitment to help them.  Plus now that we’d seen the patterns behind the problems first hand, the solutions were obvious to everyone.

We learned that customers had been allowed to install and configure everything on their own, with no support or standardization so their numerous hacks and workarounds made every click take 90 seconds to load - when scaled up to millions of calls per month (all while customers were screaming at them) the cost was staggering.

Next Steps:We couldn’t force them to scrap all their customizations right away, especially since their mistakes were caused by our not having supported them during the installation process. So we needed some low hanging fruit to prove the concept and build trust / traction - all without touching their customizations.

I designed a flattened navigation hierarchy, with all the contextual searchers universally available at the top level, which would save users from having to move up and down the hierarchy looking for the search engine they needed, or change context from one type of call to another.

This model could be applied to Callers, Tickets, Products, Catalogs, Pricing, Brands, Orders, Articles, Returns - everything.

I reviewed the model with lead developers and it would be a simple code change for us and wouldn’t disrupt any of the client’s customizations.

Jen Wilkenson

Search

Context

Options

The team, and management, were trepidatious about the new concept, and still concerned about making a mistake in such a high-stakes project. So I redesigned the entire app using this new concept so they could see it in context - which they loved, but they were still apprehensive about moving forward with anything this new.

Client / catalog / price switchers

63 icons → only 5

Call mechanisms

Primary tools

Low contrast avoids eye strain

All clutter = gone

Nesting = Gone

Consistent text link colors

Clear visual language

Adaptive context alongside primary buttons

Submit Order

Create Schedule

Submit + Schedule

Consistent primary  and secondary buttons

Clearly they needed absolute certainty this would work and the only way forward was to build a prototype so I could prove the efficacy of this solution beyond all reasonable doubt.

So I built the prototype and created a usability test to prove that while the concerns of our team were valid, once the customers started using the modified tools to handle actual calls, they would conclusively love the application within 40 minutes.

The test consisted of multiple phases and methodologies to get their initial impressions, then dive a little deeper, then start using the tools in a context they were completely familiar with, to see how their impressions might evolve and to avoid superficial interpretations that come from simplistic stand-alone methodologies.

First Impressions

Usage Tests

Scoring

They hated it, but we knew they would because they were deeply entrenched in the familiar (which they also hated). This also convinced our detractors that we were taking their concerns seriously and that we were not biasing the results.

I then let the users take control of the prototype to handle incoming calls from customers with unusually complex issues (which required them to use the new features repeatedly).  This showed everyone how quickly they would accept the changes.

I also designed the scoring system for the test to allow the rest of the team to do the scoring and see for themselves how right this path really was.

Results

• reduction in learning curve (estimated by clients)
• reduction in call-times (estimated by clients)
• million in recovered revenue per quarter (estimated by clients)

• to win over change-averse clients
• 36 furious clients became passionate advocates
• 200,000+ users now enjoy their jobs

Plus Oracle’s buyout of ATG hinged on this work so indirectly this project was worth 1 billion dollars

Accolades

Growth MindsetWhat I learned: