Section III: Customer Support

The Report to Congress said, “Customer support would be critical to the success of any Direct

File option and is also a major cost driver.” ¹⁸ Direct File piloted not only a new product, but

also a new means of delivering customer support.

Live chat was selected as the primary customer support channel because it allowed for

customer support to be integrated directly into the product and didn’t require taxpayers to

leave Direct File to get assistance through another channel (phone). Taxpayers with more

complicated issues that could not be resolved through chat could have their issues escalated,

and a team member would make an outbound call to the taxpayer to try and resolve the issue.

The chat-first approach to customer support also aligned with the pilot strategy of starting

small, enabling Direct File to test, learn, and get it right before expanding to all taxpayers. Live

chat was integrated within the Direct File product in a way that allowed customer support to

gradually expand in concert with the overall number users in each phase of the pilot. Figure 13

shows the daily chat volume over the course of the pilot.

18 IRS. IRS Report to Congress: InflaƟon ReducƟon Act §10301(1)(B) IRS-run Direct e-File Tax Return System. Page 17. htps://www.irs.gov/pub/irs-pdf/p5788.pdf

28

Figure 13. Chat volume per day ¹⁹

Another benefit of live chat was the opportunity to experiment with a channel that has become

increasingly common in the private sector. Other business units in IRS, such as Automated

Collection System Support (ACSS), have begun using live chat in recent years. The Direct File

pilot was the largest use of live chat support within the IRS to date, and it allowed the agency

to continue learning how live chat meets taxpayer preferences and expectations and how

taxpayers would react when live chat was the only option for receiving assistance. This has

the potential to impact IRS taxpayer service overall as the agency looks to provide taxpayers

with more choices in how they can interact with the IRS.

From March 4 ^th through April 20 ^th, Direct File staffing averaged 41 CSRs per day. This average

was determined by using the total CSR hours available to chat in the system, divided by 7.5

hours of production time over a total of 47 workdays. The actual quantity of CSRs varied

throughout the 15-hour workday (7:00am to 10:00pm Eastern), from a low of 7 CSRs to a high

of 62 CSRs. Figure 14 shows how demand for live chat corresponds to Direct File usage.

19 Note that purple lines represent the weekends. Customer support was made available on the weekends as we moved further into filing season with the excepƟon of March 31 (Easter Sunday).

29

Figure 14. Chat volume by time of day ²⁰

The CSRs handled 38,600 chats with an Average Handle Time (AHT) of nine minutes. Based

on the staffing available to support the pilot, the IRS had the capacity to handle approximately

450,000 Direct File chats per month. Given that approximately 10% of pilot participants

engaged customer support, this means the IRS could have assisted 4.5 million filers with the

400 CSRs who were trained for the pilot. Average Wait Time (AWT) was generally less than a

minute, as shown in Figure 15.

Figure 15. Average Handle and Wait Times

20 11:00pm and 12:00am reflect chats that were requested aŌer hours due to queues not being turned off. Rather than leƫng these chats languish, HQ staff would someƟmes work to resolve the chats in queue.

30

Opportunity: Authenticated chat

Taxpayer feedback about live chat as a communication channel was positive overall. However,

we identified opportunities to improve the chat technology features available to representatives

and better equip them to handle more complex and difficult cases.

Allowing representatives to see the history of Direct File chat interactions with an individual

taxpayer is one such feature. In some cases, busy taxpayers had to leave a chat and come

back later to finish receiving assistance. In other cases, taxpayers lost connectivity on their

mobile phones and dropped before a chat was over. Returning taxpayers had to start from the

beginning when chatting with a new representative, which could be a frustrating experience for

both taxpayers and the representatives.

Allowing representatives to access a taxpayer’s IRS Online Account information would also

help CSRs to more quickly troubleshoot and resolve issues. For example, some of the most

common errors that result in rejected returns are entering an incorrect prior year AGI or self select PIN. Even later in the filing season, when Direct File began connecting to the Online

Account to suggest prior year AGI within the product, taxpayers contacted customer support

after receiving a rejection. Direct File representatives sent taxpayers to self-service resources

to check their Online Accounts and understand why the error might have been occurring. But

allowing representatives direct access to information like prior year AGI and the date when last

year’s return was filed would help representatives quickly pinpoint the issue and instruct

taxpayers what data they should enter.

Finally, allowing representatives to see exactly what taxpayers see on their screens would

help CSRs more quickly understand and resolve issues. Using co-browsing or screensharing

features, CSRs could guide taxpayers through the product to correct errors, or they could

confirm and document technical bugs to report to the product team.

All of these features can be enabled by allowing taxpayers to seamlessly “authenticate” or log

into their chat session at the same time that they log into Direct File.

The Direct File team piloted new ways to provide a better customer experience for taxpayers

based in service and human-centered design methodologies. Service design is the activity of

planning and arranging an organization’s resources (people, resources and processes) in

order to directly improve the employee’s and the customer’s experience. Human-centered

design is a methodology that incorporates feedback from the people for whom you are

designing throughout the design process.

The team knew they wouldn’t be able to anticipate every question a taxpayer would ask while

using Direct File, so they began with an initial set of Knowledge Articles for CSRs to use,

knowing that it would be constantly iterated upon throughout the pilot. This content was

31

drafted with plain language by content designers, vetted by IRS counsel, and then translated

into Spanish.

The Direct File customer support team engaged with teams across the IRS in order to

implement the service. Collaborators included TS, Accounts Management, Contact Center

Support Division, Online Services, the Servicewide Electronic Research Program (SERP)

team, Human Capital Office, and the Office of Chief Counsel.

The Direct File CSRs were located at four geographically-distributed sites. CSRs had a year or

more of experience supporting the TS Accounts Management program, which provided

foundational IRS knowledge to build upon. All of the 400 representatives, leads, and managers

engaged in an in-person training ahead of being assigned to Direct File. The in-person training

was focused on the Direct File taxpayer experience, tutorials to learn the live chat platform,

and the general policies and procedures of the pilot program. Additionally, CSRs, managers,

and leads engaged in on-the-job training and refreshers to help reorient them ahead of being

staffed on Direct File. CSRs who were not scheduled to staff Direct File would instead support

other Accounts Management work to ensure Direct File only used the level of resources

needed each day.

Opportunity: Improved training

Based on taxpayer and representative feedback throughout the pilot, should Direct File

continue in future years, there are several opportunities to improve training:

• We now have a robust body of knowledge resources that were refined throughout the

pilot based on real taxpayer questions and feedback. In the future, we could more

confidently design realistic practice scenarios around these resources, with a focus on

the most common and the most complex issues, such as technical troubleshooting

and rejections. This would provide CSRs with more time to practice their skills using

real scenarios.

• Knowing that questions about navigating the Direct File product were a significant

portion of taxpayer chats, we could dedicate more training time to increasing familiarity

with the product, in addition to exploring and training representatives on potential new

chat features like co-browsing or screensharing with taxpayers.

• Communicating via chat is a skill that is different from interacting with taxpayers via

phone. Future training can spend more time reinforcing best practices for

communicating complex information via chat in a way that’s natural, conversational,

and user-friendly.

The exact length of training should be driven by the needs of any redesigned curriculum.

32

Customer support was integrated within the operations of the service itself. The “one team”

approach promoted a holistic experience for taxpayers by ensuring that the product and

customer support teams were in close communication, with daily (or even more frequent)

feedback loops improving the quality of both the product’s user experience and of customer

support’s answers. The teams set up multiple methods for gathering feedback from taxpayers

and CSRs to inform improvements to all aspects of the Direct File experience. CSRs were

able to escalate to the product team particularly challenging taxpayer inquiries that may

require hands-on support, a fix within Direct File, or a follow-up call-back from a Direct File

representative.

Critical improvements to Direct File often stemmed from taxpayer or CSR feedback, including

enabling taxpayers to reset their draft tax return, more informative error messages, and clearer

language and descriptions throughout the product. Approximately 200 taxpayer cases were

escalated to the product team during the pilot. Chat transcripts were reviewed and synthesized

by the product team to identify and prioritize

These feedback loops also existed during training,

with feedback from CSRs relayed to the product

Direct File. This culture of rapid iteration led one

CSR to remark in training feedback, “Having the opportunity to see our suggestions come to

fruition so quickly? This is so much fun. Normally things move at the speed of government, but

this really feels like we are an extremely important part of shaping the future of the IRS and

how we interact with the country as a whole.”

In the Touchpoints survey, taxpayers were asked, Figure 16. Please rate your experience

“Please rate your experience contacting Direct contacƟng Direct File customer support for

File customer support for assistance.” 90% of assistance.

taxpayers who had used customer support

responded that their experience was Excellent or Above Average, with only 2% responding that their experience was Below Average or Very Poor

(see Figure 16).

33

CSRs noted the opportunity to improve the training and use of chat to communicate more

naturally and conversationally with taxpayers, avoiding information overload to the taxpayer.

Were Direct File to continue in future years, the live chat experience could be improved by:

• Expanding training, especially to:

o Allow for more familiarity with the Direct File tool

o Practice more complex topics such as technical troubleshooting

o Reinforce best practices with using prepared responses in a way that is natural

and user-friendly for the taxpayer

• Implement new chat technology features that would better equip CSRs to handle more

complex cases, such authenticating chat sessions so CSRs can see the history of

interactions with an individual taxpayer and co-browsing/screen-sharing during

authenticated chat sessions.

34