Research and analysis

Evaluation of changes to how HMRC supports Self Assessment customers in the 2023 to 2024 tax year

Published 19 March 2024

Introduction

HM Revenue and Customs (HMRC) has a strong online offering, whether you want to:

• view your records, for example, to check if you owe tax

• find out how tax law applies to you, for example, through HMRC’s digital assistant, published guidance, live and recorded webinars or YouTube videos

• take action, for example, file your tax return or make a payment or claim a repayment

Customers can use online services 24/7, without having to wait on the phone or wait for a reply to a letter. This frees up HMRC’s advisers to help those who cannot use online services or have complex needs. HMRC is continuing to improve and expand its online services, increasing their capabilities and ease of use so they become the default option for customers.

In the 2023 to 2024 tax year, HMRC trialled changes to how it supports Self Assessment (SA) customers including how customers could get answers to their SA-related queries.

On 8 June 2023, HMRC announced it was trialling a seasonal SA helpline model, with the helpline closing for 3 months over the summer (from 12 June 2023 to 3 September 2023). Customers were directed to HMRC online services as an alternative where they were supported by online guidance and a digital assistant, backed up by the ability to deal with an HMRC webchat adviser, if necessary. 

On 7 December 2023, HMRC announced further changes to the SA helpline and the Agent Dedicated Line (ADL) during the SA peak filing period. This prioritised calls relating to filing returns, making payments, and querying repayments, and calls from vulnerable or digitally excluded customers. Calls on all other queries were directed to online services, again backed up by the availability of HMRC advisers on webchat. The changes were in place from 11 December 2023 until the SA filing deadline on 31 January 2024.

This publication follows the release of interim findings on the seasonal closure of the SA helpline, which were released in a letter to the Treasury Committee in October 2023. The evaluations are structured in two parts:

• Part A: Seasonal model trial

• Part B: Self Assessment peak trial

These parts are structured around the following important questions.

Question 1: Did the trials affect the ability of customers to file and pay on time?

Question 2a: Did the trials result in more customers self-serving online?

Question 2b: Did the trials result in an overall reduction in customer contact needing to be handled by HMRC advisers?

Question 2c: How many of those customers who needed to speak to an HMRC adviser were able to get through?

Question 3: Did customers who were unable to use the online services during the seasonal SA model trial get the extra help they needed?

Question 4: Did the trials result in a long-term shift from phone contact to online self-service?

Question 5: Did the trials affect customer experience and satisfaction?

Question 6: How did HMRC use resource to support SA customers during the trials?

Part A: Seasonal model trial

Introduction

This section of the evaluation covers detailed findings on the seasonal SA model trial. From 12 June 2023 until 3 September 2023, the SA helpline closed and customers who called were directed to HMRC online services as an alternative. This included support by online guidance, digital assistant and by HMRC advisers via webchat. 

Interim findings were released in a letter to the Treasury Committee in October 2023. This full evaluation provides further details on:

• filing and payments (Question 1)

• customer contact during the helpline closure (Question 2 to Question 3)

• an assessment of what happened when the helpline re-opened (Question 4)

• customer satisfaction and experience (Question 5)

• repayments (Question 6)

The evaluation is structured to answer the important questions set out in the introduction.

Question 1: Did the seasonal SA model trial affect the ability of customers to file and pay on time?

There is no evidence that the helpline closure has negatively affected filing rates or payments. Both were in line with or better than the same point in recent years.

1.75 million SA tax returns for the 2022 to 2023 tax year were filed during the SA helpline closure. This is more than during the same period in each of the previous 3 years (1.64 million in 2021 to 2022, 1.68 million in 2020 to 2021, 1.63 million in 2019 to 2020), but lower than 2018 to 2019 (1.85 million). By the end of the closure, 27% of the issued returns had been filed. This rate is higher than the previous year and in line with earlier years.

3.97 million payments with an SA reference were made during the closure. This is more than during the same period in each of the previous 4 years. SA income tax receipts in June, July and August 2023 totalled over £14.1 billion. This is more than the same period last year, when it was £12.4 billion, and higher than Office for Budget Responsibility monthly profiled forecast, £13.3 billion.

Question 2a: Did the seasonal SA model trial result in more customers self-serving online?

The seasonal model trial resulted in more customers self-serving through GOV.UK and the SA digital assistant. Other online activity continued the strong growth seen prior to the trial. The evaluation found that: 

• SA related page views on GOV.UK increased by 42% compared to the same period in 2022, compared to growth of 1% immediately before the trial

• SA digital assistant interactions where the customer did not request a webchat increased to 15,100 in the first week of the closure, compared to a weekly average of 10,800 between April and June 2023 prior to the closure, an increase of 40%

• Personal Tax Account (PTA) and Business Tax Account (BTA) SA services were accessed 3.86 million times during the helpline closure, up 26% on the same period in 2022

• the HMRC App was used 7.18 million times by SA customers during the helpline closure, up 71% on the same period in 2022

Customers were encouraged to self-serve and seek guidance using HMRC’s online services. HMRC promoted the use of GOV.UK throughout the helpline closure, which resulted in a significant increase in views for SA related guidance and self-service pages. Data for GOV.UK views is only available if a customer accepts all cookies on the corresponding page. This means that the count of page views underestimates the true value. Instead, percentage changes can be used to indicate changes over time.

During June, July and August 2023, SA related page views saw an increase of 42% on the same period in 2022. This is much higher than year-on-year growth prior to the closure. SA related page views in April and May 2023 was only 1% higher than in April and May 2022.

During the closure, usage of PTA and BTA SA service, and HMRC App by SA customers was much higher than the same period in 2022. However, this increase is consistent with the growth prior to the helpline closing. This means it is not possible to attribute the growth directly to the trial, but during the trial the trend in customers self-serving through online services continued.

The PTA and BTA SA service were accessed 3.86 million times during the closure, an increase of 0.80 million, up 26% on the same period in 2022. Year-on-year growth prior to the closure was 22%. The HMRC App was used by SA customers 7.18 million times during the closure, an increase of 2.97 million, up 71% on the same period in 2022. This was broadly consistent with strong year-on-year growth prior to the closure, which was 80%.

Across the period there were a total of 500,000 interactions with the SA digital assistant within opening hours (Monday to Friday 8am to 8pm, and Saturday 8am to 4pm). The number of interactions increased significantly with 48,000 during the first week of the closure. The weekly average prior to the closure between April and June 2023 was 23,000. Weekly volumes fell gradually towards the end of the trial. However, the proportion of customers requesting a webchat from the SA digital assistant increased from 52% before the closure to 68% during the helpline closure. This contributed to higher webchat demand than expected.

SA digital assistant interactions where the customer did not request a webchat increased to 15,100 in the first week of the closure, this was an increase of 40%. The weekly average prior to the closure between April and June 2023 was 10,800.

Overall during the helpline closure:

• 32% of the SA digital assistant interactions did not result in a webchat request (161,000)

• 68% of the SA digital assistant interactions resulted in a webchat request (338,000) 

Question 2b: Did the seasonal SA model trial result in an overall reduction in customer contact needing to be handled by HMRC advisers?

There was a reduction in customer contacts needing to be handled by HMRC advisers, despite webchat demand being higher than HMRC expected before the trial.

In the equivalent period of 2022, 864,000 SA helpline calls were routed to an adviser. During the closure, there was a reduction to around 357,000 contacts that needed to be handled by advisers. This contact was through a mixture of channels:

• 106,000 more Online Service Helpdesk (OSH) calls than the same period in 2022

• 216,000 more SA webchats than the same period in 2022

• an estimated 35,000 SA calls displaced to other helplines

There were 389,000 OSH calls during the closure where the customer chose to speak to an adviser after hearing automated messages (adviser attempts) compared to 283,000 in the same period in 2022.

Figure 1: OSH adviser attempts

Source: HMRC customer contact data

The main route to SA webchat was through the SA digital assistant. There were 338,000 requests for webchat from the SA digital assistant, much higher than the period prior to the trial. HMRC changed webchat and digital assistant systems during the summer of 2022. This means it is not possible to make robust comparisons between the trial period in 2023 and the equivalent period in 2022. Comparisons to the period immediately prior to the closure have been used instead.

Figure 2: Weekly webchat requests from SA digital assistant during opening hours

Source: HMRC customer contact data

In addition to the 338,000 requests for SA webchat from the SA digital assistant, there were 12,000 SA webchat requests from other routes including direct from PTA in-service pages. Customers could also use the OSH digital assistant if they needed technical help with online services. There were 66,000 webchat requests from the OSH digital assistant.

Question 2c: How many of those customers who needed to speak to an HMRC adviser were able to get through during the seasonal SA model trial?

Overall performance on the OSH helpline in handling calls that queued to speak to an adviser (adviser attempts handled) was below HMRC’s service standard of 85% but improved as the trial progressed to above target. Despite increased contact, performance was better than the same period in 2022. 75.3% of OSH helpline adviser attempts were handled, compared to 67.5% during the same period in 2022. Adviser attempts handled is the proportion of calls that are handled where the customer has chosen to speak to an adviser after hearing automated messages.

Performance on SA webchat was also higher than prior to the closure despite increased contact. 74% of SA webchat requests were handled, compared to 60% prior to the closure.

During the helpline closure, HMRC received 660,000 SA helpline calls into the telephony platform. Customers were played a message explaining that the helpline was closed until 4 September. Customers were told they could get instant answers to questions about SA from the digital assistant. They were also told they could find information quickly and easily in their PTA or BTA, on GOV.UK, in the HMRC App and on HMRC’s YouTube channel. Customers who required help accessing online services or needed extra support were asked to call the OSH. If the customer was calling about a bereavement they were routed to an adviser, otherwise the call ended.

Figure 3: Percentage of adviser attempts handled on OSH helpline

Source: HMRC customer contact data

Overall, 75.3% of OSH helpline adviser attempts were handled. As adviser attempts increased to a high in week 3 of the closure, the percentage of adviser attempts handled fell to a low of 50.1%. From week 4, adviser attempts reduced and performance improved. During the second half of the closure, the percentage of weekly adviser attempts handled was consistently above HMRC’s 85% service standard.

Figure 4: Percentage of webchat requests handled from SA digital assistant

Source: HMRC customer contact data

Of the 338,000 SA digital assistant interactions that resulted in a webchat request:

• 36,000 received a busy message and were not placed in a queue (11%)

• 50,000 were placed in a queue but abandoned before being handled (15%)

• 251,000 were handled (74%)

Prior to the closure, 60% of webchat requests from SA digital assistant were handled by an adviser, but this varied from week-to-week. During the first 3 weeks of the closure, 52% of webchat requests from SA digital assistant were handled. HMRC increased the resource deployed to webchat and changed the queue threshold to increase the number of customers that could join the queue. By week 4 of the closure, the percentage of webchat requests handled increased to 78% and in the final week this was 96%.

As well as SA webchat and OSH helpline, customers could also use the OSH digital assistant if they needed technical help with online services. There were 66,000 webchat requests from the OSH digital assistant but only 31% were handled compared to 60% prior to the closure. The number of customers who can join the queue (the queue threshold) is lower for OSH webchat compared to SA webchat. Because of the lower queue threshold, a large proportion of requests for OSH webchat received a busy message and were not placed in a queue (69%). Less than 1% abandoned before being handled.

Question 3: Did customers who were unable to use the online services during the seasonal SA model trial get the extra help they needed?

HMRC provided a good service to customers who needed extra support with calls handled above HMRC’s overall target of 85%. Customers calling the OSH who required extra support due to a health condition or personal circumstances were transferred to the Extra Support Team (EST). 26,000 calls were transferred to the EST during the trial, more than 3 times the amount prior to the trial. 88.2% of these calls were handled.

HMRC continued to support customers calling the SA helpline about a bereavement. During the closure, 2,500 calls were routed to an adviser and 95.9% of these were handled, significantly above HMRC’s service standard of 85%.

Question 4: Did the seasonal SA model trial result in a long-term shift from phone contact to online self-service?

It is too early to say if the trial has changed contact from phone to online self-service. While there has been limited differences in calls between September and December, the growth in use of online self-service and webchat has continued to expand rapidly. Monitoring of phone contact and online service usage throughout 2024 and the 2025 SA peak will be necessary to determine whether there is a lasting, long-term effect.

Figure 5: Weekly calls to SA helpline

Source: HMRC customer contact data

There was an initial increase in calls to the SA helpline when it first re-opened on 4 September 2023 but calls soon returned to previous year’s levels.

During September 2023, the total number of calls to the SA helpline was 12% higher than September 2022.This was driven by an increase in the first weeks of the helpline re-opening. During October the total number of calls to the SA helpline was 6% higher than 2022 but in November the total number of calls to the SA helpline was 7% lower than 2022.

SA digital assistant interactions and SA webchats handled were both much higher when the helpline re-opened compared to the same period last year. This continued the trend seen prior to the closure.

Between 4 September 2023 and 3 December 2023, compared to the same period of the previous year:

• SA digital assistant interactions during opening hours increased from 108,000 to 282,000

• SA webchats handled from all routes increased from 53,000 to 140,000

Similarly, usage of the PTA, BTA, HMRC App and GOV.UK online guidance and self-service pages remained high when the helpline reopened compared to the same period in 2022.

Between 4 September 2023 and 3 December 2023, compared to the same period of the previous year:

• PTA and BTA SA service usage had increased from 2.88 million to 3.39 million sessions, up 18%

• HMRC App usage from SA customers had increased from 4.20 million to 7.15 million sessions, up 70%

• during September, October and November 2023, views of SA-related guidance and self-service pages on GOV.UK saw an increase of 33% on the same period in 2022

Question 5: Did the seasonal SA model trial affect customer experience and satisfaction?

Online self-service scores remained high throughout the trial with higher volumes of people interacting online. There was some effect on SA webchat scores and OSH scores, which dropped during the trial, but SA telephony scores were consistent before and after the closure.

Customer exit surveys are not directly comparable across channels. However, comparisons can be made within channel across periods of time:

• online service satisfaction covering SA service and filing, pay online, BTA SA services and PTA personal tax homepage was consistent before closure, during closure and after reopening of the helpline (80.6%, 79.9%, 78.4% respectively)

• SA webchat scores dropped during the trial. However, satisfaction was still above 70% at a time when more customers who preferred to use the phone moved to webchat

• OSH telephony scores were lower early in the trial as the percentage of adviser attempts handled dropped. However, by the end of the trial, the percentage of adviser attempts handled improved, and scores returned to pre-trial levels

• satisfaction with the SA telephone helpline in the period after it reopened was higher than in the period before the closure

HMRC offers customer exit surveys on phone, webchat and online services. Customer Satisfaction is a measure of the percentage of HMRC customers that responded that they were either ‘satisfied’ or ‘very satisfied’ with the service and is referred to as CSAT. Net Easy represents the percentage of positive responses minus the percentage of negative responses to HMRC customer surveys. Customers were asked “How easy was it for you to do what you needed to do today?”. This gives a score that ranges from 100 (all positive responses) to -100 (all negative responses). HMRC report customer satisfaction and Net Easy scores for telephony, webchat and online services.

The customer exit surveys are not directly comparable across channels. Telephony exit surveys are offered at the start of a phone call and can be completed regardless of whether the customer goes on to be answered or not. Webchat exit surveys are only offered if the customer connects with an adviser.

Exit surveys for online services do not include digital assistant. Online services scores presented in this evaluation are based on SA service and filing, pay online, BTA SA services and PTA personal tax homepage. They are not all specific to a tax regime and will therefore also include non-SA customers.

HMRC cannot track the end-to-end journey of customers when they are directed to online services or online guidance from the digital assistant. This means there is currently limited ability to assess satisfaction of digital assistant robustly without bias.

It is not possible to compare consistently between the closure period 2023 and the same period in 2022 due to technical changes made to telephony systems. As a result, CSAT and Net Easy scores are compared over 3 periods:

• before closure: 1 April 2023 to 11 June 2023

• during closure: 12 June 2023 to 3 September 2023

• after reopening: 4 September 2023 to 3 December 2023

HMRC’s service standard for Net Easy is 70, and for CSAT it is 80%. CSAT and Net Easy scores for online services remained close to this standard during the closure and after reopening.

Table 1: Online services satisfaction and experience (SA service and filing, pay online, BTA SA services and PTA personal tax homepage)

	CSAT	Net Easy
Before closure	80.6%	64.2
During closure	79.9%	62.6
After closure	78.4%	58.9

Source: HMRC customer exit survey data

CSAT and Net Easy scores for the SA and OSH webchat services during the closure were lower than before the closure. All scores increased when the helpline reopened.

Table 2: Webchat satisfaction and experience

	SA CSAT	SA Net Easy	OSH CSAT	OSH Net Easy
Before closure	76.3%	55.5	46.5%	1.5
During closure	70.1%	42.2	39.4%	-9.9
After closure	73.5%	54.6	47.3%	9.6

Source: HMRC customer exit survey data

When completing a webchat exit survey, customers are asked “How would you prefer to get in touch with HMRC?”. The percentage of respondents who preferred webchat as the contact method dropped from 73% before the closure to 53% during the closure. It then increased back to 74% after reopening. Phone as the preferred contact method increased to 44% during the closure.

Table 3: SA Webchat contact preference responses

	Webchat	Phone	Other
Before closure	73%	23%	4%
During closure	53%	44%	3%
After closure	74%	22%	4%

Source: HMRC customer exit survey data

This shows that although there was a large increase in webchat usage as customers switched from phone to webchat during the closure when they were unable to phone, there was also a large increase in the proportion of customers expressing a preference for phone contact.

Table 4: Telephony satisfaction and experience

	SA CSAT	SA Net Easy	OSH CSAT	OSH Net Easy
Before closure	33.5%	-41.2	29.4%	-45.6
During closure	closed	closed	24.7%	-55.3
After closure	40.9%	-28.2	35.4%	-34.8

Source: HMRC customer exit survey data

CSAT and Net Easy scores for the OSH helpline during the closure were lower than before the closure. During the first 4 weeks of the closure, the OSH helpline saw a reduction in the percentage of adviser attempts handled and a longer average speed of answering. However, as adviser attempts reduced and the percentage handled improved, by the end customer satisfaction scores had returned to pre-trial levels.

From June to November 2023, HMRC received 720 complaints related to the helpline closure. The top 3 reasons for complaints were in relation to:

• repayments and progress chasing which included cases awaiting security and compliance checks (23%)

• late filing penalties (10%)

• SA criteria in which customers were not able to speak with someone about their criteria for being within SA and delays in closing their SA records (8%)

This represents a small proportion of the 11,000 SA related complaints received during this time.

Question 6: How did HMRC use resource to support SA customers during the SA seasonal model trial?

There was increased resolution of SA overpayments and an associated increase in repayments as a result of redeploying staff who would otherwise have handled calls on the phones.

Advisers freed up by the reduction in customer contact during the closure were redeployed to process customer post and work items where customers had overpaid tax (SA overpayments). 373,000 more SA overpayment work items were processed than during the same period last year. This increase is slightly higher than reported in the interim evaluation as it is now based on more detailed weekly figures. Clearing these work items helps to get customers’ affairs up to date, including producing repayments.

There were 677,000 repayments from the Computerised Environment for Self Assessment (CESA) system in June, July and August 2023 with a total value of £1.7 billion repaid. This was 139,000 more repayments than the same period in 2022 (+26%). Although operational processing of overpayment work items will have contributed to the repayments, other factors will also have contributed to this, and it is not possible to isolate the effect of the redeployment of advisers as a result of the trial. The volume of repayments in the 2 months prior to the helpline closure were also up on the same period of the previous year (up11%).

Before the helpline closure, there had been 124,000 calls progress chasing SA repayments this tax year. This accounted for 21% of the helpline calls and was much higher than the same period last tax year. When the helpline re-opened in September, repayment calls had fallen and were back in line with the previous year’s levels.

Figure 6: Number of calls progress chasing “What’s happening with my SA repayment?”

Source: HMRC customer contact data

Part B: Self Assessment peak trial

Introduction

In December 2023 HMRC announced that from Monday 11 December until Wednesday 31 January 2024, HMRC would only handle priority SA calls. Priority calls were those relating to returns, repayments and complex matters, as well as calls from vulnerable and digitally excluded customers. Calls on all other SA queries would be directed to HMRC’s online services including its digital assistant, backed up by the availability of HMRC advisers on webchat. The change was implemented as follows:

• when the HMRC telephony system identified a query that was not in a priority category, HMRC would use recorded and SMS messaging to direct the customer to its online services, including webchat, and end the call

• if a customer with one of these queries did reach an adviser, advisers would direct customers to HMRC’s online services, and the call would then be ended

As part of the announcement, HMRC gave examples of queries that could be resolved easily online including updating personal information, chasing on the progress of a SA registration, ending SA registration, and checking a Unique Taxpayer Reference number.

Customers who could not resolve their query online, including vulnerable and digitally excluded customers, were directed to call the Online Services Helpdesk (OSH) where they were routed directly through to an adviser. HMRC monitors all calls to identify customers who may need extra help. These customers are passed on to HMRC’s Extra Support Team (EST) who are specially trained to deal with vulnerable customers.

This section of the evaluation focuses on the period between 11 December 2023 and 31 January 2024 and will be assessing the effect of these changes on:

• return filing and payments

• online services

• phone calls

• customer satisfaction

• HMRC resourcing

Question 1: Did the SA peak trial affect the ability of customers to file and pay on time?

Customers’ filing rates and payments were in line with or better than previous tax years. Therefore, there is no evidence to suggest that customers were not able to file their tax return or make a payment as a result of the prioritisation of calls handled by the SA helpline.

6.07 million SA tax returns for the 2022 to 2023 tax year were filed between 10 December 2023 to 11 February 2024. This is in line with a comparative period for 2021 to 2022 returns, and more than previous tax years back to 2016 to 2017. By the end of the period, 91% of the issued returns had been filed, in line with previous tax years.

During the peak from 11 December 2023 to 31 January 2024, 4.84 million payments with an SA reference were made. This is more than during the same period in each of the previous 4 tax years.

Question 2a: Did the SA peak trial result in more customers self-serving online?

In the SA peak trial, there was clear evidence that customers made more use of online services than in the previous tax year, with high levels of engagement and customer satisfaction from customers.

Engagement with the digital assistant increased significantly but this also resulted in a large increase in the number of webchats. Growth in online self-service was observed both during and before the SA peak. For example, PTA and BTA SA services were accessed 13% more during the 2024 SA peak compared to 2023 SA peak.

There was a 6% increase in SA related page views on GOV.UK between December 2023 and January 2024 compared with the same period in the previous tax year. Usage of these pages is particularly high in January, which sees around a quarter of all page views each year. Data for GOV.UK views is only available if a customer accepts all cookies so total figures may underestimate the true volume of use of SA related guidance views.

Similarly, usage of the PTA, BTA and the HMRC App by SA customers increased during the 2024 SA peak. The PTA and BTA SA services were accessed 5.78 million times during the 2024 SA peak, an increase of 0.68 million (up 13%) compared to the 2023 SA peak. The HMRC App was used by SA customers across 6.03 million times during the 2024 SA peak, an increase of 2.34 million (63%) compared to the 2023 SA peak.

Usage of the SA digital assistant and webchat increased significantly during the 2024 SA peak within opening hours (Monday to Friday 8am to 8pm, and Saturday 8am to 4pm). There were a total of 328,000 SA digital assistant interactions, an increase of 55% compared to the 2023 SA peak.

Customers interacting with the digital assistant can request a webchat with an adviser. Of these digital assistant interactions:

• 48% (158,000) interactions did not result in a webchat request

• 52% (170,000) interactions resulted in a webchat request

The proportion of digital assistant interactions that requested a webchat (52%) was higher than in the 2023 SA peak (45%).

There were also 12,000 webchat requests that came through other channels, including direct from PTA in-service pages. Of these, 67% were handled.

Question 2b: Did the SA peak trial result in an overall reduction in customer contact needing to be handled by HMRC advisers

The volume of customer contact needing to be handled by HMRC advisers reduced from 1,271,000 in the 2023 peak to 1,017,000 in the 2024 peak. This continued the trend seen before the peak period. Changes made to direct non-prioritised queries to online services and end the call were a partial success, with 12% of calls redirected. Of these 61% of those called the SA helpline again within 5 working days.

Table 5: Number of adviser attempts and webchat requests by contact channel

	2023 SA peak	2024 SA peak
SA helpline	786,000	513,000
OSH	220,000	217,000
ADL	151,000	71,000
SA webchat (individuals and agents)	95,000	183,000
OSH webchat	19,000	33,000
Total attempts to speak to an adviser	1,271,000	1,017,000

Source: HMRC customer exit survey data

Calls and webchats handled by HMRC advisers

Total calls and webchats handled by an adviser on the SA, ADL and OSH helplines reduced to 664,000, compared with 747,000 in the previous SA peak (-11%). This change comprised of an increase in webchats handled by an adviser, a reduction in calls handled by an adviser on the SA and ADL helplines, and an increase in calls handled by an OSH adviser.

Figure 7 shows that total calls and webchats handled by advisers reduced by 83,000 to 664,000, compared with 747,000 in the previous SA peak. This change comprised of:

• SA calls handled by advisers fell by 158,000 (35%) while on ADL they fell by 78,000 (59%)

• OSH calls handled by advisers increased by 60,000 (54%)

• SA webchats handled by advisers increased by 90,000 (204%) while on OSH they increased by 4,000 (42%)

Figure 7: Change in demand handled from the 2023 SA peak to 2024 SA peak

Source: HMRC customer contact data

Call volumes received on the SA helpline

Between 11 December 2023 and 31 January 2024 customers made 980,000 calls to the SA helpline. This was a reduction of 148,000 calls in comparison to the 2023 SA peak, when customers made 1,128,000 calls to the SA helpline. The number of calls where the customer chose to speak to an adviser (adviser attempts) also fell.

However, during the previous year’s SA peak on 30 and 31 January 2023, HMRC’s telephony systems had an outage, which resulted in an unusually high number of phone calls as customers repeatedly attempted to speak to an adviser while the system was down.

Despite the telephony outage last year making interpretation of the figures more difficult, there is evidence that the totality of HMRC’s change initiatives has reduced the demand on the SA helpline. The volume of adviser attempts reduced compared to the previous SA peak, though it is not possible to attribute this to the changes made during the SA peak trial as reductions were also observed before the peak period (4 September 2023 to 3 December 2023).

Figure 8 shows the number of adviser attempts each week from the SA helpline reopening until the end of the peak. The figures for 2023 to 2024 were lower than the previous tax year for almost all weeks.

Figure 8: Weekly adviser attempts to the SA helpline

Source: HMRC customer contact data

SA helpline calls ended during messaging

HMRC uses automated messaging to encourage customers to self-serve via HMRC’s online services. At this point in the customer journey, there are 3 possible outcomes:

• the customer ends the call during messaging

• HMRC ends the call

• the customer continues and is placed in the queue to speak to an adviser

Where a customer ends the call, this could be because the message answered their query or because of the likely wait time to speak to an adviser.

Of the 980,000 calls to the SA helpline during the 2024 SA peak, 467,000 (48%) were ended during messaging, either by HMRC or the customer, before being placed in the queue to speak to an adviser. Table 6 shows a breakdown of how the total calls coming into the SA helpline changed from the 2023 SA peak to the 2024 SA peak.

Table 6: Change in total calls received on the SA helpline from 2023 SA peak to 2024 SA peak

	2023 SA Peak	2024 SA Peak	Change
Total calls	1,128,000	980,000	-148,000
Calls ended during messaging	342,000	467,000	125,000
Calls where the customer chose to speak to an adviser	786,000	513,000	-273,000

Source: HMRC customer contact data

Non-prioritised calls directed to HMRC’s online services during the SA peak

Of the 467,000 calls ended during messaging 113,000 were ended by HMRC as they did not relate to a priority reason and were redirected to online services. 70,000 (61%) of those ended calls resulted in a repeat call to the SA helpline within 5 working days - most of them were within 5 minutes.

Analysis of the call reasons recorded by HMRC advisers over the 2024 SA peak shows that the proportion answered that related to non-prioritised reasons (29%) was in line with 2023 SA peak (29%). Call listening suggests that HMRC advisers chose to provide support to customers with non-prioritised queries rather than directing them to go online.

Customers with non-prioritised queries were able to contact the OSH if they were unable get online. 

HMRC handled more calls on the OSH, building the capability of more customers to self-serve online. 173,000 calls were answered during the 2024 SA peak in comparison to 112,000 calls handled in the previous tax year, an increase of 60,000.

Question 2c: Were customers who needed to speak to an HMRC adviser able to get through during the SA peak trial?

Most customers who needed to speak to an adviser were able to get through, and the proportion that were able to do so was higher during this peak than in last year’s peak. However, performance remained below HMRC’s service standard of 85% for adviser attempts handled. Adviser attempts handled is the proportion of queries where the customer has chosen to speak to an adviser that get handled by an adviser.

HMRC answered 65% of calls that joined the queue to speak to an adviser, compared to 60% in the 2023 SA peak. For webchat, 67% of chats in the queue were answered, compared to 46% in the 2023 SA peak. However, there were longer wait times for calls and webchat during the 2024 SA peak compared with the previous peak.

In tables 7 and 8 telephony includes the SA, ADL and OSH helplines, and webchat includes the SA webchat (individuals and agents) and OSH webchat.

Table 7: Adviser attempts handled (%) by contact channel

	2023 SA peak	2024 SA peak
Telephony	60%	65%
Webchat	46%	67%

Source: HMRC customer contact data

Table 8: Average speed of answering (minutes) by contact channel

	2023 SA peak	2024 SA peak
Telephony	23	27
Webchat	5	12

Source: HMRC customer contact data

For the SA helpline, the adviser attempts handled was 57% in the 2024 SA peak, which was the same as the 2023 peak.

Customers had to wait longer in queues before they could speak to an adviser in the 2024 peak compared with the 2023 peak. The average time spent in the queue for calls which were handled (average speed of answering or ASA) during the 2024 SA peak was 36 minutes for the SA helpline. This was an increase of 8 minutes compared to the previous peak, when the ASA was 28 minutes.

Webchat

Table 9 shows the outcomes of the SA webchat requests. The majority (73%) of SA webchat requests were handled. There was a 92% increase in the number of webchat requests and a 204% increase in the number of webchats handled.

Table 9: Breakdown of SA webchats in the 2024 SA peak by outcome

	Count of webchats	Percentage  of requests
Handled	134,000	73%
Abandoned	22,000	12%
Busy	27,000	15%
Total webchat requests	183,000	100%

Source: HMRC customer contact data

Customers were also able to use the OSH digital assistant and webchat. The OSH digital assistant saw a 50% increase in usage from 65,000 interactions for the 2023 SA peak to 97,000 interactions for the 2024 SA peak.

Table 10 shows the outcomes of the OSH webchat requests. Of the 97,000 interactions with the OSH digital assistant, 33,000 (34%) resulted in a webchat request, an increase from 19,000 (30%) interactions for the 2023 peak.

OSH webchat requests were less likely to be handled than SA webchat requests. The number of customers who can join the queue (the queue threshold) was lower for OSH webchat compared to SA webchat. Because of the lower queue threshold, a large proportion of requests for OSH webchat received a busy message and were not placed in a queue (64%). While 1% requests had customers abandon before being handled.

Table 10: Breakdown of OSH webchats in the 2024 SA peak by outcome

	Count of webchats	Percentage of requests
Handled	12,000	36%
Abandoned	200	1%
Busy	21,000	64%
Total webchat requests	33,000	100%

Source: HMRC customer contact data

Question 4: Did the SA peak trial result in a long-term shift from phone contact to online self-service?

It is too early to say if there has been a long-term shift from phone contact to online self-service. Analysis of phone contact and online service usage throughout 2024 and the 2025 SA peak will be necessary to determine whether there is a lasting, long-term effect.

Question 5: Did the SA peak trial affect customer experience and satisfaction?

Customer satisfaction was lower for the SA helpline during the 2024 SA peak than both before the peak and when compared to 2023 SA peak. Although comparing customer satisfaction across channels is not possible, customer satisfaction was high for online services and for webchat. This suggests that directing customers to these services may have provided some customers with a better service than that provided via HMRC helplines.

To measure customers’ perception of HMRC’s service, HMRC offer customer exit surveys on phone, webchat and online services. Customer Satisfaction is a measure of the percentage of HMRC customers that responded that they were either ‘satisfied’ or ‘very satisfied’ with the service and is referred to as CSAT. Net Easy represents the percentage of positive responses minus the percentage of negative responses to HMRC customer surveys. Customers were asked “How easy was it for you to do what you needed to do today?”. This gives a score that ranges from 100 (all positive responses) to -100 (all negative responses). HMRC report customer satisfaction and Net Easy scores for telephony, webchat and online services.

The customer exit surveys are not directly comparable across channels. Telephony exit surveys are offered at the start of a phone call and can be completed regardless of whether the customer goes on to be answered or not. Webchat exit surveys are only offered if the customer connects with an adviser. How a customer can access feedback surveys varies between other online services.

Customer satisfaction with online services

Customers can leave feedback on a range of HMRC’s online services. Table 11 presents CSAT and Net Easy scores for a selection of important online services for SA customers. The satisfaction scores below cover users of PTA personal tax pages, Pay Online, SA return filing, and BTA SA services. HMRC online exit surveys are not always specific to a tax regime and will therefore include non-SA customers.

CSAT and Net Easy scores for online services were broadly consistent between the 2024 peak and 2023 peak and remained high.

Table 11: Online services satisfaction and experience (SA service and filing, pay online, BTA SA services and PTA personal tax homepage)

	2023 SA Peak	2024 SA Peak
CSAT	80.2%	79.7%
Net Easy	61.0	60.4

Source: HMRC customer exit survey data

Customer experience on the telephony helplines

Telephony customer satisfaction levels were lower during the 2024 SA peak when compared to the same time last SA peak for both the SA helpline and the ADL. For the SA helpline, CSAT and Net Easy were also lower in the 2024 SA peak than in the period between the reopening of the helpline and before the peak (4 September 2023 to 3 December 2023).

The ADL serves a wide range of customers and as a result satisfaction scores do not just cover SA customers nor customers raising just SA queries. Variations in customer satisfaction on the ADL may also be related to resource changes made on that helpline earlier in the year.

Table 12:  CSAT for the main telephony helplines

	2023 SA peak	pre-2024 SA peak	2024 SA peak
SA	42.7%	40.9%	35.2%
ADL	53.4%	31.8%	33.5%
OSH	24.1%	35.4%	34.3%

Source: HMRC customer exit survey data

Table 13: Net Easy for the main telephony helplines

	2023 SA peak	pre-2024 SA peak	2024 SA peak
SA	-23.6	-28.2	-36.2
ADL	3.1	-35.2	-34.9
OSH	-56.7	-34.8	-38.5

Source: HMRC customer exit survey data

Customer experience on webchat

For both SA and OSH webchat, CSAT was lower during the 2024 SA peak than during the 2023 peak. SA also saw a drop in Net Easy, which remained at a similar level for OSH. For SA webchat in particular, customer satisfaction levels were high.

Table 14: CSAT for SA and OSH webchat

	2023 SA peak	2024 SA peak
SA	82.3%	76.8%
OSH	52.5%	44.3%

Source: HMRC customer exit survey data

Table 15: Net Easy for SA and OSH webchat

	2023 SA peak	2024 SA peak
SA	65.3	57.9
OSH	4.3	4.9

Source: HMRC customer exit survey data

Question 6: How did HMRC use resource to support SA customers during the SA peak trial?

The size of the SA population is increasing and HMRC has less resource to deal with the increasing number of customers in the tax system. To manage this, HMRC is directing customers to its online services to resolve their queries.

Across all contact channels, HMRC had fewer overall staff to deploy to the SA helpline and the OSH over the 2024 SA peak compared to the 2023 SA peak. More staff were deployed to OSH (telephony and webchat) and SA webchat, but fewer staff were deployed to SA telephony and post.

HMRC prioritised providing support to those customers trying to use online services by increasing resourcing to the SA webchat, OSH helpline and OSH webchat compared to 2023 SA peak. This decision was taken based on the lessons learnt from the closure of the SA helpline over the summer.

HMRC also diverted resource from SA post onto the SA helpline, to provide support to customers who still needed to speak to an adviser on the phone. However, HMRC still deployed fewer staff onto the SA helpline during the 2024 SA peak than the previous SA peak.

Glossary

Term	Definition
Agent Dedicated Line (ADL)	The helpline for customer to call for queries about Self Assessment or PAYE when they are formally authorised to act as an agent or tax adviser for a client.
Average speed of answering (ASA)	The average time waited by a customer in the queue before their query is handled by a HMRC adviser. Can apply to both phone calls and webchats.
Adviser attempts	The number of calls where the customer chose to speak to an adviser after hearing automated messages.
Adviser attempts handled	The proportion of queries where the customer has chosen to speak to an adviser that get handled by an adviser. Can apply to both phone calls and webchats.
Business Tax Account (BTA)	An online account for use by individuals, sole traders, partnerships and limited companies for information on taxes including Self Assessment, VAT, PAYE and Corporation Tax.
Closure period, pre closure period, post closure period	The 12 weeks during the SA helpline closure, and periods of similar length that were before the closure or after the reopening.
Customer satisfaction (CSAT)	The percentage of HMRC customers that responded they were either ‘satisfied’ or ‘very satisfied’ with the service when responding to the exit survey on HMRC’s phonelines, webchat or online services. In other publications, the CSAT for online services may be referred to as digital services.
Digital assistant	An automated online tool which customers can use to find information about technical support with HMRC’s online services. It is one way to access HMRC’s webchat.
Extra Support Team (EST)	The team who support customers that call if a health condition or personal circumstances make it difficult when they contact HMRC.
HMRC App	An app for mobile devices to get information about tax, National Insurance, tax credits and benefits. For example, it can be used to check an individual’s Unique Taxpayer Reference or to make a Self Assessment payment.
Net Easy	Represents the percentage of positive responses minus the percentage of negative responses to HMRC customer surveys. Customers were asked “How easy was it for you to do what you needed to do today?”. This gives a score that ranges from 100 (all positive responses) to -100 (all negative responses). It is included in the exit survey on HMRC’s phonelines, webchat or online services. In other publications, the Net Easy score for online services may be referred to as digital services.
Online Service Helpdesk (OSH)	The helpline for customer to call for support with HMRC’s online services. There is also an OSH webchat.
Personal Tax Account (PTA)	An online account for use by individuals to check their records and manage their details with HMRC.
Self Assessment helpline (SA helpline)	The helpline for customers to call for advice on Self Assessment. There is also a SA webchat.
Webchat	An online service where customers can speak to a HMRC adviser. Most commonly accessed through HMRC’s digital assistant. This evaluation considers the SA webchat and the OSH webchat.
2023 SA peak, 2024 SA peak	The periods in December and January which lead up to the SA filing deadline on 31 January each year. For the 2024 SA peak, this was from 11 December 2023, the date on which the changes were made to the SA and ADL helplines. For the 2023 SA peak, an equivalent number of days was used as a comparator.

Evaluation evidence sources

The data included in this evaluation comes from HM Revenue and Customs’ customer contact data and administrative data.

The customer contact data contains several different sources across multiple contact channels and contains information for each instance an individual engages with HMRC’s services. These contact channels include telephony, post, digital assistant, webchat, the Personal Tax and Business Tax Accounts, the HMRC App and interactions with GOV.UK web pages.

Data on GOV.UK page views is made available to all government departments and agencies by Government Digital Service (GDS). In this evaluation’s approach, SA GOV.UK pages are based on page titles or links that contain the phrase ‘Self-Assessment’. This list of pages is not exhaustive of all guidance in relation to SA or customer contact. This data is only available if a customer accepts all cookies so total figures may underestimate the true volume of use of SA related guidance views.

HMRC’s administrative data contains information on SA returns filing and payments.

Related links

Letter from Jim Harra, HM Revenue and Customs, to the Treasury Select Committee in October 2023 – Interim findings on the seasonal closure of the SA helpline.

HMRC to trial seasonal Self Assessment helpline – Announcement on GOV.UK of the seasonal closure of the Self Assessment helpline between 12 June and 4 September 2023.

Self Assessment helpline to focus on priority queries – Announcement on GOV.UK of changes to Self Assessment helpline between 11 December 2023 and 31 January 2023.