Key takeaways
- Phone reports show only part of reception work; online requests, walk-ins, tasks, letters, prescriptions and messages often take as much time again.
- Online demand has grown fast: GP practices received 9,617,628 online requests in March 2026, 69% more than in August 2025[1].
- A two-week tally of every task type, with a rough time per task, is enough to see the true workload and where it peaks.
- Convert counts into staff hours by hour of day, then compare with the rota; the gaps explain most overtime and missed breaks.
- Repeat the count twice a year and after any big change, using exactly the same method.
Why phone data alone misleads
Most practices can produce a call report in minutes: calls offered, answered, abandoned and waiting times by hour. It is tempting to plan reception staffing from that report alone. But a receptionist's day also includes processing online requests, greeting arrivals, handling walk-in queries, scanning and workflowing letters, actioning clinician tasks, sorting prescription requests, replying to messages and chasing results. None of that appears in the call report, so a team that looks overstaffed on the phone data can be overwhelmed in practice.
The mix has shifted quickly. Since online consultation tools had to stay open throughout core hours, online requests have grown sharply: practices received 9,617,628 online requests in March 2026, 69% more than the 5,690,082 in August 2025[1], equal to 153 requests per 1,000 patients[1]. Each one has to be read, routed and often answered by the admin team.
Source: Care Quality Commission, 2026[1]
Show the data
| Item | Value |
|---|---|
| August 2025 | 5,690,082 |
| March 2026 | 9,617,628 |
The workload streams to count
| Stream | What to count | Where the number comes from |
|---|---|---|
| Phone calls | Calls answered and outbound calls made | Telephony reports |
| Online requests | Requests received, by type, and how many reception handled | Online consultation tool reports |
| Front desk | Arrivals, walk-in queries, specimens handed in | Tally sheet at the desk |
| Clinical tasks | Tasks sent to admin by clinicians | Clinical system task reports |
| Documents | Letters and results scanned, coded or workflowed | Document management reports |
| Prescriptions | Repeat requests processed by admin | Clinical system or pharmacy portal |
| Messages and email | Patient messages, practice inbox emails | Messaging and inbox reports |
| Other | Forms, insurance reports, registrations, complaints | Tally sheet |
Where a system can produce a report, use it. Where it cannot, such as walk-ins and ad hoc tasks, use a simple tally for a fixed period. Do not try to count everything forever; two weeks of good data is far more useful than six months of patchy data.
Running a two-week count
Choose two normal weeks
Avoid bank holiday weeks and the peak of the flu campaign for your baseline. Count those separately later if you want to plan for them.
Agree the categories
Use no more than ten task types, defined in one line each, so everyone counts the same way.
Pull system reports
Download hourly or daily figures for calls, online requests, tasks, documents and prescriptions for the same two weeks.
Tally what systems miss
Put a tally sheet by each workstation and at the desk for walk-ins, specimens, forms and other tasks, split into hourly blocks.
Time a sample
For each task type, time ten examples across different staff and take the middle value. A rough figure is fine; the aim is proportion, not precision.
Note who was working
Record how many people were actually on each task in each hour, not just how many were rostered.
Explain to the team why you are counting and that it is about the workload, not individual speed. People record more honestly when they know the result could lead to more staff or a better rota.
From counts to staff hours
Multiply each task count by its typical time, add the streams together for each hour of the day and compare the total with the staff hours on the rota for that hour. Leave room for breaks, interruptions and handovers; nobody works at full stretch for a whole shift. The result is a simple chart of needed hours against available hours, which shows where the rota does not match the work.
Example (fictional): A practice counts two weeks of work. Between 8am and 10am, phone calls need about 7 staff hours a day, online requests 2 hours, the desk 1.5 hours and urgent tasks 1 hour: 11.5 hours of work in a two-hour window, or almost six people. The rota has four people on reception and one on the online list. Between 2pm and 4pm, the count shows less phone work but a large block of letters and tasks, while three people are rostered on the phones. Moving one person's start time and shifting document work to the afternoon closes most of the gap without new hires.
Keep the calculation in a spreadsheet you can reuse. When volumes change, such as after a new online form or a list size increase, you only need to update the counts.
What else the count reveals
- Avoidable contact: calls asking whether a request has been received, or chasing results, show where patients lack information.
- Rework: tasks returned between clinicians and admin more than once point to unclear protocols.
- Hidden specialists: one person doing all the coding or all the prescriptions is a single point of failure.
- Mismatched hours: document work done during peak call times, or phones lightly staffed when online requests spike.
- Growth: comparing counts year on year shows whether workload is rising faster than staffing.
The wider system is also changing what reception handles. NHS England's annual report estimates that the NHS App helped save 4.1 million hours of staff time in 2025/26[2]. Some of that shows up at reception as fewer calls for routine tasks, but online routes also generate new admin work, which is why measuring all streams matters.
Seasonal and special counts
Your two-week baseline describes a normal fortnight. Most of the pressure, though, comes from weeks that are not normal. Once the baseline exists, run shorter counts for the periods that break the rota, so you can plan them rather than survive them.
- The first working day after each bank holiday, when the long weekend's demand arrives at once.
- Flu and other vaccination campaign weeks, which add booking calls, walk-in questions and clinic admin.
- The weeks after a change to the appointment system or online form, when patients are learning the new route.
- Year-end weeks for recalls and reviews, when searches and invitations add to the admin load.
- The week after a local news story about access, which often brings a spike in calls and complaints.
For each one, compare the needed hours with the baseline. If the first day after a bank holiday needs half as many hours again, roster for it every time and block leave on those days.
Measures to report every month
- Calls answered and abandoned by hour, and the longest wait each day.
- Online requests received by day and the time from receipt to first response.
- Tasks and documents outstanding at the end of each day.
- Overtime hours and missed breaks reported by staff.
- Staff hours available against estimated hours needed for the busiest two hours of the day.
Telephony features such as call queuing with position messages and call-back[3] change how calls are spread through the day, so re-run the count after any change to the phone system. If you use an AI Employee such as Ava for routine practice questions, record what it handles and what it passes to the team in the same log, so the workload picture stays complete.
Sharing the results
A workload count only changes things if people see it. Present the result to the reception team first: they did the counting and will spot anything that looks wrong. Then take a one-page summary to the partners with three items: where needed hours exceed available hours, what rota changes would close the gap at no extra cost, and what extra hours, if any, are still needed.
Show the result as a simple picture of needed and available hours across the day. Partners who see the morning gap drawn out tend to agree to change start times quickly, and staff who see their invisible work counted feel recognised, which matters for retention as much as the rota itself.
Pitfalls to avoid
- Counting only what is easy to count, so the hardest work stays invisible.
- Measuring individual speed and creating a culture of rushing.
- Using a week with unusual events as the baseline.
- Changing the categories between counts so results cannot be compared.
- Collecting data and never sharing the result with the team.
Workload pressure in general practice is widely reported. In a BMA survey of more than 1,300 practices, 74% saw workload rise after online access changes[4]. A careful count is how a practice turns that general pressure into a specific, fixable picture of its own.
Frequently asked questions
How long should a reception workload count run?
Two normal weeks is usually enough for a baseline. Repeat it twice a year and after significant changes, such as a new phone system, online form or list size increase.
Should we measure how long each person takes?
Time a sample of tasks across the team to get typical times, but avoid measuring individuals. The aim is to understand the total workload and when it happens, not to rank staff.
What if the count shows we need more staff?
First check whether moving shifts or tasks between hours closes the gap. If it does not, the count gives partners clear evidence of how many extra hours are needed and when.
References
Care Quality Commission (2026). State of Care 2025/26: capacity in primary care (opens in a new tab) ↑
NHS England (via GOV.UK) (2026). NHS England annual report and accounts 2025 to 2026 (opens in a new tab) ↑
NHS England and DHSC (2023). Delivery plan for recovering access to primary care (opens in a new tab) ↑
British Medical Association (2026). Pressures in general practice data analysis (opens in a new tab) ↑