IRS live chat apps have room for improvement, watchdog finds
The IRS is providing a good service to taxpayers with its live chat applications, per a new watchdog report, but there could be problems if the agency doesn’t get a handle on performance measures.
The Treasury Inspector General for Tax Administration found that the IRS hasn’t analyzed its chat apps, nor does it have reliable metrics around their use. The agency piloted live chat functionality in 2017 and launched automated chatbots in 2021.
In conducting its audit, TIGTA reviewed statistical reports that didn’t add up. One showed a live assistor handling up to 603 live chats simultaneously — a stat the IRS admitted was incorrect due to an internal cap for the number of concurrent chats set at three.
The watchdog also found miscalculations of the apps’ “handle time metric,” which is supposed to measure how long it takes a live assistor to manage one live chat. IRS IT officials speculated that those miscalculations could be attributed to live assistors not exiting out of chats, which are automatically closed after four hours.
TIGTA flagged additional discrepancies in chat application reports, finding that the number of resolution codes often did not match the true number of live chats.
“For example, there were a total of 635,684 resolution codes shown in the Calendar Years (CY) 2023 and 2024 statistical reports, but these reports only reflected 613,056 live chats,” the report said.
The poor data collection speaks to a broader point made by the TIGTA: Evaluations made based on shoddy statistics could draw false conclusions about how the chat apps perform.
Automated chatbots also exhibited some worrying traits, according to the report. There were 29 responses and 44 keywords or questions that the technology either provided insufficient information on or wasn’t able to recognize at all. Per TIGTA, the IRS “began taking corrective actions to address the deficiencies” as soon as the watchdog raised the issue.
Overall, the watchdog noted that the live chat functions do provide an “effective and efficient service to help taxpayers meet their tax obligations.” But there are concerns that some of the live assistors have taken on too heavy a load. In a review of 40 live assistors, TIGTA found that 60% were working multiple live chats at the same time.
“Working multiple chats concurrently can decrease customer service and increase the risk of inappropriately disclosing taxpayer information to the wrong taxpayer,” the report stated. Overworked assistors could lead to inaccurate information and inaccurate returns, it noted.
TIGTA delivered nine recommendations to the tax agency, including one aimed at establishing a process to ensure accurate data from the live chat applications, and another to institute real performance standards so the IRS can truly evaluate the effectiveness and efficiency of the tools.
The tax agency agreed with all nine suggestions and said it has already pursued “corrective actions” for a handful of them.
