Corporate report

HMRC鈥檚 customer segmentation

Published 5 November 2020

HMRC collects revenues, according to the laws set by Parliament, from millions of individuals and from businesses of all sizes.

To help us do this, we segment our customers into groups so we can identify their needs and risks more accurately and tailor our responses accordingly 鈥� from support to help people get their tax right, to targeted action against avoidance, evasion and criminal activity.

The way people live, work and pay their taxes continues to evolve and so the way in which we understand and segment our customers into groups also changes.

In 2019 to 2020 we improved our analysis through using more evidence-based assumptions, aligning our segmentation methodologies and revising population sizes so that estimates of receipts, yield and compliance spend figures across customer groups are always based on the best available data.

This technical paper highlights the main changes in the 2019 to 2020 estimates and the reasons behind them.

1. Customer segmentation changes

1.1 Population changes

The population figure for Individuals for 2019 to 2020 uses a different methodology to 2018 to 2019, to better align with our tax receipts coming from this segment. The 2019 to 2020 figure is a , from Official Statistics on numbers of taxpayers and registered traders. The 2019 to 2020 figure was used, adjusting to remove the 鈥榳ealthy鈥� population. The 2018 to 2019 figure was an estimate of all individuals who interact with the broader tax system.

The population figure for midsize businesses, charities and public bodies has been rounded to reflect that this is an estimate

The increase in the 2019 to 2020 鈥榳ealthy鈥� population reflects a change in 2019 to the HMRC definition of the 鈥榳ealthy鈥� population, which brings more taxpayers into scope. The previous definition was customers with:

  • incomes greater than 拢150,000 (in the last year), and/or
  • assets greater than 拢1 million (in the last year)

Customers are currently defined as 鈥榳ealthy鈥� if they have:

  • incomes greater than 拢200,000 (in any of the last 3 years), and/or
  • assets greater than 拢2 million (in any of the last 3 years)

鈥楽mall鈥� and 鈥榣arge鈥� population sizes remain consistent. The 鈥榮mall鈥� population size uses Business, Energy and Industrial Strategy Official Statistics, adjusted to remove midsize and large businesses. HMRC鈥檚 large business segment is defined as the largest 2,000 businesses by turnover and assets.

Table 1 - Population figures

Population 2018-19 2019-20
Individuals 45 million 31 million
Wealthy 500,000 700,000
Small 5.7 million 5.7 million
Midsize, charities and public bodies 207,000+ 200,000+
Large 2,000 2,000

1.2 Compliance yield customer segment changes

The 2019 to 2020 compliance yield customer segment figures are larger compared to 2018 to 2019. This is because this year, total compliance yield from across the department has been fully segmented into the customer groups.

While previous segmentation captured the majority of our compliance yield, it did not include the additional revenue from some significant categories of HMRC鈥檚 activity, such as the impact of changes to legislation or of our debt management activity.

This is included this year as our analysis in this area continues to evolve and improve. Some variability in total compliance yield by tax regime is expected as very large cases settled in an individual year can distort overall tax regime results. It is also difficult in some cases to distinguish individuals from small businesses, which is why yield figures fluctuate in these categories.

Table 2 - Compliance yield segmentation

Yield (拢bn) Small Midsize, charities and public bodies Large Individuals Wealthy Criminals Total
2019-20 (all HMRC yield) 7.6 5.0 15.8 2.8 2.2 3.6 36.9
2019-20 (without new categories, for comparison to 2018-19) 6.6 4.5 13.3 2.1 2.0 3.1 N/A
2018-19 5.6 4.0 10+ 2.4 1.8 3+ 34.1

Note: figures may not sum due to rounding.

1.3 Tax receipts customer segment changes

  • in 2019 to 2020 all tax receipts segmentations were revisited and refreshed based on most recent data
  • the Pay As You Earn (PAYE) and National Insurance contributions (NICs) receipts from 鈥榠ndividuals鈥� and 鈥榳ealthy鈥� groups were updated with the new 鈥榳ealthy鈥� population definition (which led to a higher population) and the latest data - this resulted in an increase in Wealthy PAYE and NICs receipts
  • segmentation assumptions have been reviewed, leading to a few improvements:

    • a better understanding of allocation of NICs across PAYE and Self Assessment
    • a full segmentation breakdown across individual as well as business customers for the 鈥楥orporation Tax (CT), Excise, VAT, Capital Gains Tax (CGT) and other taxes鈥� category - his has resulted in more receipts from this category being allocated to the 鈥榣arge鈥� customer segment, as well to 鈥榠ndividuals鈥� and 鈥榳ealthy鈥� groups.

Table 3 - Tax receipts segmentation

Tax receipts (拢bn) Tax head Small Midsize, charities and public bodies Large Individuals Wealthy Total
2019-20 PAYE and NICs 80 115 110 260 45 305
CT, Excise, VAT, CGT and other taxes 75 50 150 15 40 330
2018-19 PAYE and NICs 85 95 115 250 27 295
CT, Excise, VAT, CGT and other taxes 115 50 135 300

1.4 Compliance spend customer segment changes

There has been a methodology improvement to the compliance spend segmentation for 2019 to 2020 based on latest data, resulting in changes to all customer segments.

The increase in 鈥榤idsize, charities and public bodies鈥� spending and reduction in 鈥榣arge鈥� spending for 2019 to 2020 is from a change in methodology, where large public bodies and charities, previously categorised in the 鈥榣arge鈥� customer segment, have been reallocated into the 鈥榤idsize, charities and public bodies鈥� customer group.

This allows for a consistent methodology between population estimates, yield, receipts and compliance spend.

Table 4 鈥� compliance spend segmentation

Spend (拢m) Small Midsize, charities and public bodies Large Individuals Wealthy
2019-20 500 270 210 360 190
2018-19 490 210 230 360 150

Note: figures may not sum due to rounding.