Metrics finance

Benchmarking your practice against peers

How to benchmark a GP surgery, dental practice or clinic using public data: the main UK sources, choosing a fair peer group and turning gaps into action.

Ava research teamPublished 6 min read

On this page
  1. Why benchmark at all
  2. The main public data sources
  3. What to benchmark first
  4. Choosing a fair peer group
  5. Quality ratings as a benchmark
  6. Comparing fairly
  7. Example: a fictional practice benchmarks itself
  8. Sharing benchmarks with the team
  9. Pitfalls that mislead
  10. Dental practices and private clinics
  11. Frequently asked questions
  12. References

Key takeaways

  • Free public data covers most of what a practice needs to benchmark: patient experience, appointments, workforce, quality ratings and dental activity.
  • Choose peers with similar list size, age profile, deprivation and contract type, not just neighbours.
  • Compare rates and shares, not raw counts, and check sample sizes and recording before drawing conclusions.
  • Benchmarking is for finding questions; the answers come from looking inside your own practice.

Why benchmark at all

Inside a busy practice it is hard to tell whether a problem is local or universal. Are our phones worse than everyone else's, or is this simply what winter looks like? Is our staffing generous or thin? Benchmarking answers those questions by setting your figures beside those of comparable practices. Done well, it stops a team from either panicking about problems everyone shares or ignoring problems that are theirs alone.

Variation in the system is large, which is what makes benchmarking useful. ICB areas in England range from 38 to 56 GPs and from 13 to 45 nurses per 100,000 patients[1], against an average of 46 fully qualified GPs per 100,000 patients in 2025/26[2]. For dentistry, England had 43 NHS dentists per 100,000 people in 2025/26, ranging from 34 to 69 across ICB areas[3]. Two practices a few miles apart can be working in very different conditions.

The main public data sources

Free and low-cost benchmarking sources for UK practices
SourceWhat you can compareHow oftenWatch out for
GP Patient SurveyEase of contact, experience, overall rating, by practiceAnnuallySmall samples for small practices; check confidence intervals
Appointments in General PracticeAppointment rates, mode, role, booking interval, missed appointmentsMonthlyRecording and category mapping vary between practices
QOF achievement dataAchievement, prevalence and personalised care adjustments by indicatorAnnuallyHigh prevalence or adjustments can flatter or depress scores
General practice workforce dataClinicians and staff per patient by roleRegularlyNetwork-employed staff may sit outside practice figures
CQC ratings and reportsOverall and key question ratingsWhen inspectedRatings can be several years old
Friends and Family TestPositive and negative feedback sharesMonthlyCollection methods differ, so response rates vary widely
NHS dental statisticsCourses of treatment, units of activity, patients seenAnnuallyContract values and local commissioning differ
Accountants' benchmarksIncome and cost ratios for clientsAnnuallySamples are each firm's own clients

Start with the GP Patient Survey for patient experience, because it is large and comparable. The 2026 survey received 654,714 responses, 566,443 of them online, with a national response rate of 22.1%[4]. Nationally, 76.7% of patients rated their practice as good overall[5], a useful reference line for any practice-level comparison.

What to benchmark first

It is tempting to download everything. Resist it. Pick a handful of measures that matter to your patients and your finances, benchmark those well, and add others later. For most GP practices, five measures cover the essentials.

  1. Ease of getting through by phone and overall experience, from the GP Patient Survey.
  2. Appointments per thousand patients and the share delivered on the day of booking.
  3. Clinical staff per thousand patients by role.
  4. QOF achievement on the indicators with the most points.
  5. Missed appointments per thousand patients.

For dental practices, swap in patients seen in the last 24 months as a share of the active list, chair utilisation, failed appointment rate, staff cost as a share of income and new patient conversion. For private clinics, use enquiry response time, conversion, utilisation and repeat booking rate.

Choosing a fair peer group

The most common benchmarking mistake is comparing with the wrong practices. A neighbouring practice may share a high street but serve a completely different population. Build a peer group of five to ten practices that resemble yours on the factors that drive demand and cost.

  1. Match list size

    Pick practices within roughly a third of your list size either way; very small and very large practices run differently.

  2. Match age profile

    Compare the share of patients over 65 and under 5, which shape demand more than any other factor.

  3. Match deprivation

    Use the deprivation decile of the practice population to avoid comparing very different needs.

  4. Match setting and contract

    Urban, rural or dispensing; GMS, PMS or APMS; NHS, mixed or private for dental practices.

  5. Keep the group stable

    Use the same peers year after year so trends mean something.

If you cannot find enough close matches locally, widen the search across the ICB or a similar area elsewhere in England. A well-matched practice a hundred miles away is a better comparator than a poorly matched one down the road.

Quality ratings as a benchmark

CQC ratings are the most visible benchmark for patients. Of GP practices assessed under CQC's current framework, 87% are rated good and 3% outstanding, while 9% require improvement and 1% are inadequate[6]. A good rating is therefore the norm rather than a distinction; what sets practices apart is the detail in the report and how recently it was issued.

CQC ratings of GP practices assessed under the current framework, England
CQC ratings of GP practices assessed under the current framework, EnglandStacked bar chart. Outstanding: 3%; Good: 87%; Requires improvement: 9%; Inadequate: 1%.Outstanding: 3% (3%)Good: 87% (87%)Requires improvement: 9% (9%)Inadequate: 1% (1%)CQC ratings of GP practices assessed under the current framework, EnglandStacked bar chart. Outstanding: 3%; Good: 87%; Requires improvement: 9%; Inadequate: 1%.Outstanding: 3% (3%)Good: 87% (87%)Requires improvement: 9% (9%)Inadequate: 1% (1%)
  • Outstanding 3%
  • Good 87%
  • Requires improvement 9%
  • Inadequate 1%

Source: Care Quality Commission, 2026[6]

Show the data
ItemValue
Outstanding3%
Good87%
Requires improvement9%
Inadequate1%

Read your peers' reports, not just their ratings. Notes on access, governance and staff support often show practical ideas worth borrowing, and they show what inspectors look for in practices like yours.

Comparing fairly

  • Use rates and shares: appointments per thousand patients, staff per thousand patients, percentages of patients reporting good experience.
  • Check sample sizes: survey results for a small practice can move several points from chance alone.
  • Check recording: an apparently low appointment rate may reflect unbooked triage contacts rather than less activity.
  • Use the same period for every practice in the comparison.
  • Look at trends over two or three years as well as the latest figure.
  • Note context: a practice that absorbed patients from a closure will look different for a year or two.

Workload per clinician deserves special care. The average number of patients per full-time-equivalent GP in England was 2,143 in September 2026[7], but practices that rely heavily on nurses, pharmacists and other roles can have far more patients per GP without worse access. Compare total clinical staffing per patient as well as GPs alone.

Example: a fictional practice benchmarks itself

Example: Priory Gardens Surgery, a fictional practice with 11,000 patients in a mid-deprivation town, picks seven peers with similar list sizes and age profiles. Its patient survey score for ease of phone contact is well below the peer median, its appointment rate per thousand patients is close to the median and its nurse staffing per patient is the lowest in the group.

Example: the manager visits the two peers with the best phone scores. Both use call-back queues and both have moved routine booking online, so mornings carry mostly urgent calls. Priory Gardens adopts both changes and recruits an extra practice nurse. The following year its phone score moves to the peer median. The benchmark did not tell the practice what to do; it told the manager where to look and whom to ask.

Sharing benchmarks with the team

Benchmarks land better when the team helps interpret them. Show the comparison at a practice meeting, explain how the peer group was chosen, and ask staff what they think explains the gaps. Reception teams often know exactly why phone scores differ from a neighbouring practice; nurses know why a recall programme works better elsewhere.

Avoid league tables that name colleagues or turn benchmarks into targets on their own. A gap is a starting point for learning, and the most useful outcome of benchmarking is often a visit to a peer practice that does something well, rather than a new target on a dashboard.

Pitfalls that mislead

  • Cherry-picking peers that make the practice look good, or bad, to support a decision already made.
  • Treating the national average as a target when your population is very different.
  • Reading a single year's survey result as a trend.
  • Comparing ratings issued years apart as if they were current.
  • Benchmarking costs without matching what each practice includes, such as network-employed staff or premises reimbursement.

Write down the peer group, the sources and the date of each figure on the benchmarking sheet itself. Anyone picking it up later, including the next practice manager, can then repeat the exercise on the same basis and see real change.

Dental practices and private clinics

Dental practices can benchmark activity using NHS dental statistics and costs using the dental earnings and expenses estimates and accountants' benchmarks. Private clinics have fewer public sources, so they rely more on trade association surveys, accountants' client benchmarks and their own trends. In every case, the principle is the same: compare like with like, use ratios, and treat any gap as a question rather than a verdict.

  1. Pick the five measures that matter most to your business.
  2. Find a public or trade source for each, or decide to track it internally.
  3. Choose a stable peer group.
  4. Review the comparison once a year and after any major change.
  5. Turn each large gap into one question for the team and one practice to visit or call.

Frequently asked questions

Where can a GP practice find benchmarking data?

The GP Patient Survey, Appointments in General Practice, QOF achievement data, general practice workforce statistics, CQC reports and Friends and Family Test results are all public and cover every practice in England.

How do we choose comparable practices?

Match on list size, age profile, deprivation, setting and contract type. Keep the same peer group over time so trends are meaningful.

What should we do with a benchmarking gap?

Treat it as a question. Check recording and context first, then look at what the best-performing peers do differently and test one change at a time.

References

  1. Care Quality Commission (2026). State of Care 2025/26: primary and community services (opens in a new tab) ↑

  2. Care Quality Commission (2026). State of Care 2025/26: capacity in primary care (opens in a new tab) ↑

  3. Care Quality Commission (2026). State of Care 2025/26: dental (opens in a new tab) ↑

  4. Ipsos (GP Patient Survey provider for NHS England) (2026). 2026 GP Patient Survey results released (opens in a new tab) ↑

  5. Ipsos (GP Patient Survey provider for NHS England) (2026). 2026 GP Patient Survey results released (opens in a new tab) ↑

  6. Care Quality Commission (2026). State of Care 2025/26: primary and community services (opens in a new tab) ↑

  7. British Medical Association (2026). Pressures in general practice data analysis (opens in a new tab) ↑

Book a demo

See Ava handle your calls.

Book a short demo. We’ll walk through your calls and workflow, and agree what Ava should take care of.

Book a demo