Loading...

When Customers Call

Heuristic evaluation research

Project category:
Software experience design
My role(s):
Manager, researcher, designer, presenter
Project duration:
3 months
Project tags:
customer experience fintech management UX design UX research service design

Defining the problem

One of the more challenging projects I encountered during my time as a design manager in financial tech was an effort to seek solutions for a customer care team that was resource-strapped and looking for ways to help customers more effectively help themselves.

The customer care team at Fannie Mae has several touch points with customers. Some of them are contact centers set up to handle technical issues, while others assist customers at financial institutions — loan officers and underwriters, for example. Finally, there are those tasked with addressing inquiries from the general public.

At the height of the housing crisis of 2008, Fannie Mae and Freddie Mac carried an inventory of 242,000 REOs (bank-owned properties) between them. During this time period, Fannie Mae's customer care contact centers were being constantly inundated with calls and emails. It was also during this time that Fannie Mae had last negotiated their service rates with the vendor who managed their call centers. These services were provided on a cost-per-contact basis.

In the ensuing years, an internal team also worked to set up a website to provide self-service to members of the public. These contacts were generally divided into a few key categories:

  1. those wishing to inquire about Fannie Mae's foreclosed properties,
  2. those seeking to affiliate professionally with Fannie Mae in order to sell homes on their behalf, and
  3. those who owned homes at risk of being foreclosed, or tenants who lived in those homes.
An analysis of inbound inquires revealed this breakdown by customer type.
An analysis of inbound inquiries revealed this breakdown by customer type.

During the years following the peak of the 2008 housing market downturn, the market gradually normalized and Fannie Mae's inventory diminished steadily. By 2018, the number of REO properties being managed was less than 25,000 — a ten-year low.

Trending data showed ten years of steadily reducing mortgage delinquencies, which directly impacted the inbound volume to Fannie Mae’s contact center.
Trending data showed ten years of steadily reducing mortgage delinquencies, which directly impacted the inbound volume to Fannie Mae’s contact center.

A complicating factor we needed to consider was a newly-renegotiated contract with the contact center vendor. As the housing market had recovered, the inbound call volume had tailed off considerably. As such, the per-contact pricing that had been agreed to during an extremely busy time was no longer the profitable partnership that it had once been. As a result, a renegotiated rate meant that the next contract year would see a 300% increase in the cost per contact.

With the housing crisis abating and the market normalizing, the number of inquiries to the contact center should also have been trending downward at the same rate. Instead, the opposite was true.

The customer care team asked us to help them figure out why this might be happening, and then bring some ideas to the table that could address the problem.

Collecting the evidence

Our research started in Salt Lake City, Utah — the home of our public-facing contact center where agents answered incoming calls to 1-800-2-FANNIE, as well as replying to emails and online chat sessions.

The research plan called for a mix of contextual inquiry and interviews. We observed several of the customer care agents in their work environment, as well as listened in to a number of the calls they received over multiple days. We spent time following up on these calls with the agents to assess the types of inquiries that required the most energy to resolve, as well as the types that seemed to result in the most escalations.

In addition, we performed a statistical analysis of incoming calls, emails and chat sessions.

All of this work revealed some key initial findings:

  • 40% if incoming calls were about REO properties.
  • 10% of REO-related calls required escalation to tier-two support or a supervisor.
  • 15% of inquiries were simply to ask about the status of an offer on a Fannie Mae property.
  • 25% of calls/emails/chats pertained to contract & addendum issues.

With these metrics as success markers, we turned our attention to Fannie Mae's primary source of information for the general public: the HomePath website.

Establishing KPIs was an important goal of the project, and brought focus to the scientific side of design research.
Establishing KPIs was an important goal of the project, and brought focus to the scientific side of design research.

Investigating for impact

Working together with the Customer Care team, we pursued a hypothesis that addressing the usability and information architecture of the site would likely result in an improved experience for customers; and would also promote discoverability of key REO-related information, thereby reducing customers' reliance on the contact centers.

With a focus on sharpening the website's capabilities in service of our goals, we selected a set of heuristic analysis markers that would be used during a usability audit of the site. Each heuristic was also subdivided further into child categories — factors contributing to its overall score.

The heuristics used for the audit comprised:

  • Consistency
  • Standards
  • Recognition rather than recall
  • Technical clarity
  • Flexibility & efficiency of use
  • User support
  • Forgiveness
  • System status
  • Accommodation

After identifying the primary workflows for both homebuyers and real estate agents, the team worked to evaluate and record their findings, isolating those areas where the website performed the worst. While the homebuyer experience fared slightly better than the agent experience, a much clearer picture of usability issues had materialized. We could much more effectively label the issues — and even quantify them to an extent — and set that all-important performance baseline in order to inform how to measure success of future improvements.

A heuristic scoring framework helped to inform the priority of design work.

Picking our priorities

Concurrent with the heuristic analysis, we used site traffic patterns along with the contact center inquiry data to further develop a strategy that would help us select appropriate areas of emphasis to focus on when recommending improvements and optimizations. Our goal was to focus on both streamlining workflows, as well as re-organizing some of the site content to help curtail the need to reach out to our contact centers in the first place.

By presenting our recommendations as already-categorized priority lists, we offered a strong design point-of-view and added emphasis to those areas we saw as being the most potentially impactful for customers.
By presenting our recommendations as already-categorized priority lists, we offered a strong design point-of-view and added emphasis to those areas we saw as being the most potentially impactful for customers.

With our heuristic evaluation data in hand, we selected four primary areas of focus.

  1. We would reconstruct the home page in order to more effectively help our customers self-select their path according to their role.
  2. We would work to improve the search results, with more impactful and better-organized contextual information.
  3. We would come up with a more forgiving “wizard” process to assist potential homebuyers with making offers on Fannie Mae-owned properties.
  4. We would optimize the FAQ experience to guide customers in finding their own solutions more easily.

Building to a solution

The final step was to transition into action mode and produce some recommended design updates that would directly address the priorities we had previously called out. A series of wireframes offered some suggestions for how to begin to re-tool the HomePath website and offer customers a significantly upgraded experience.

Homepath web site design
Homepath web site design
Homepath web site design

In the end, the customer care team was well-positioned to work with the HomePath team in a joint effort to bring about some key changes to their website, as well as implement new procedures for contact center staff to begin training customers on how to use the already-available information on the public website as a way to find answers to their questions.

All of this was done with an eye on the primary goal of reversing the trends for each of the key data points. Also, with an established heuristic evaluation process having been established, the team was well-equipped to perform the same analysis in the future after changes were eventually instituted.

Highlighting project successes

Integrated with customer care problem solve

We engaged with the customer care team in a way that allowed us to integrate into their standard 90-day problem solve cycle with a very specific goal in mind. This helped to foster the relationship of the customer experience design team, finding news ways of working with business partners and internal stakeholders.

Homepath web site design

Identified usability gaps

Heuristic evaluations became the key to isolating our strategic design priorities. A data-informed, scientific approach lent credibility to our results and actively demonstrated the team’s ability to do more than just suggest cosmetic changes to designs. We established ourselves as trustworthy partners who used rigorous methodology to identify problems and develop solutions.

Usability gaps icon

Presented recommendations as instructions

Since HomePath had plans to engage with an outside vendor to perform the technical work required to make our suggested changes, we were able to provide wireframes and flow diagrams that were annotated and well-documented. We saw this as an excellent opportunity to demonstrate our design documentation methodology and offer creative direction at the same time — even after our role in the project had ended.

Annotated designs icon
Visit the site

More Case Studies

Copyright ©, Jeremy Piontek