Key takeaways
- For most solicitors the real choice is how to combine people and automation, not which one to pick.
- People are best at welcome, judgement, distressed callers and existing clients; automation is best at overflow, out-of-hours and consistency.
- A full-time receptionist's cost includes pay, employer National Insurance, pension, cover and recruitment, not just the hourly rate.
- Pick a hybrid model by looking at when calls are missed today, not at the technology.
The choice is rarely either or
Most firms that look at automated call answering are not trying to replace a receptionist they value. They are trying to solve a specific problem: calls missed at lunchtime, voicemails after 5pm, a receptionist who cannot answer three lines at once, or a single-person reception desk that closes whenever that person is on holiday or ill. Framed that way, the question becomes which calls a person should take and which an automated service can take well.
This guide sets out where each option works best, what each really costs, and three hybrid models that suit different kinds of firm.
What a human receptionist costs
Median gross hourly pay for UK receptionists was £13.51 in 2026, and median annual pay for a full-time receptionist was £25,865[1]. Legal secretaries, who often cover reception in small firms, had a median of £14.95 an hour[1]. The floor is rising too: the National Living Wage for workers aged 21 and over is £12.71 an hour from April 2026, rising to £13.21 from April 2027[2].
Source: Office for National Statistics, 2026[1]
Show the data
| Item | Value |
|---|---|
| Receptionists | £13.51 |
| Call and contact centre staff | £14.86 |
| Legal secretaries | £14.95 |
Pay is only the start. Employers pay Class 1 secondary National Insurance at 15% on earnings above the £5,000-a-year secondary threshold in 2026 to 2027, although eligible employers can reduce their bill by up to £10,500 through the Employment Allowance[3]. Add workplace pension contributions, holiday and sickness cover, recruitment and training, and the desk, phone and systems each person needs. None of this means a receptionist is poor value; a good one is often the reason clients come back. It does mean comparisons should use the full cost.
Where a person is best
- Welcoming clients in the office, checking ID documents in person and making people feel looked after.
- Callers who are distressed, confused or vulnerable, where tone and patience matter more than speed.
- Existing clients who know the receptionist and expect to be recognised.
- Calls that need judgement: an angry client, a sensitive message, a caller who will not say why they are ringing.
- Local knowledge and relationships with agents, courts and other firms.
- Handling post, couriers, deliveries and the many small office tasks no phone service covers.
Where automation is best
- Overflow: calls that arrive while reception is already on the phone.
- Out of hours: evenings, early mornings and weekends, when callers would otherwise reach voicemail.
- Consistency: asking every new enquirer the same intake questions and producing the same written note.
- Peaks: Monday mornings, stamp duty deadlines or after a local news story.
- Simple information: opening hours, directions, documents to bring, where to find published prices.
- Cover during holidays and sickness without temporary staff.
Side-by-side comparison
| Factor | Human receptionist | AI receptionist |
|---|---|---|
| Availability | Working hours, one call at a time | Every hour, many calls at once |
| Empathy and judgement | Strong | Limited; must hand over sensitive calls |
| Consistency of intake | Varies with workload and person | Same questions every time, as scripted |
| Notes and records | Depends on discipline | Written note of every call |
| In-office tasks | Yes | No |
| Cost structure | Salary and employment costs, largely fixed | Service charges, usually linked to use or capacity |
| Holiday and sickness | Needs cover | No cover needed; plan for outages |
| Change and training | Training time; knowledge stays with the person | Script changes; knowledge sits in the configuration |
| Professional limits | Trained staff know not to advise | Must be designed and tested not to advise |
What callers want
Callers are comfortable with the phone but cautious about automation. In the 2026 consumer tracker, 79% of consumers were happy to meet face to face for an initial meeting with a lawyer and 66% by telephone[4]. Research for the Legal Services Board found the main barriers to using AI to access legal services were lack of trust in the technology (51%) and lack of human oversight (49%)[5].
That points to a simple rule for any automated line: be honest that it is automated, keep it to the jobs it does well, and make reaching a person quick. Callers who feel trapped in an automated loop will hang up and call the next firm on the list.
Three hybrid models
Overflow only. The receptionist answers as now; calls that are not answered within a set number of rings, or that arrive while the line is busy, go to the automated service. This suits firms with a strong receptionist who simply cannot take every call.
After hours. People answer during office hours; the automated service takes calls before opening, after closing and at weekends, captures enquiries and books first consultations. This suits firms whose enquiries often come in the evening, such as family, employment and private client practices.
Automation first, people for the important calls. The automated service answers every call, deals with simple information and new enquiry intake, and passes urgent calls, existing clients and anyone who asks straight to a person. This suits busier firms with several teams, where routing calls correctly is itself a job.
Whichever model you choose, write down the rules in one place: which calls go where, how quickly a person must call back, and who checks the automated notes. A one-page call handling plan, shared with the whole team, prevents the most common failure of hybrid set-ups, where everyone assumes someone else is following up the messages.
What changes for the receptionist
In a hybrid model, the receptionist's job usually gets better rather than smaller. Fewer interruptions mean more time for the work that needs a person: greeting visitors properly, calling back existing clients who left messages, chasing documents and keeping the diary in order. Many firms also ask reception to own the automated service day to day: reading the overnight notes, spotting enquiries that were routed to the wrong team and suggesting script changes.
Say this before anything changes, and involve the receptionist in choosing the model and writing the scripts. They know which callers ring repeatedly, which questions come up every day and which calls should never go anywhere but to a person. That knowledge makes the automated service better, and it makes clear to the team that the aim is support, not replacement.
Choosing a model
Measure missed calls
For two weeks, record how many calls go unanswered or to voicemail, at what times and from whom.
Sort the calls
Separate new enquiries, existing clients, third parties and others. Note which needed a person.
Find the gaps
If most missed calls are at lunchtime, overflow may be enough; if they are after 5pm, consider after-hours cover.
Talk to the team
Ask reception which calls they would happily hand over and which they want to keep.
Pilot and review
Start with the smallest model that fixes the gap, review every call for two weeks and adjust.
Example: a high street firm with one receptionist
Example (fictional): a high street firm with eight fee earners and one receptionist measures its calls for two weeks. It finds that around one call in six goes unanswered, mostly between 12.30pm and 2pm and after 5pm, and that most of those are new enquiries. The receptionist's lunch break is covered by whoever is free, which often means nobody.
The firm chooses an overflow and after-hours model. The receptionist keeps every call she can answer and all in-office work; the automated service takes overflow and evening calls, books consultations and sends notes. The receptionist reviews the notes first thing each morning and calls back anyone flagged as an existing client. Within a month, the partners can see new enquiries from evening calls that previously went to voicemail.
Questions to settle before you decide
- Which calls must always reach a person, and how fast?
- What will the automated service say about itself at the start of each call?
- Who reviews call notes, and when?
- How will callers who prefer not to use automation reach someone?
- What happens if the service is unavailable?
For the detail of what automated answering can and cannot do, read AI receptionists for law firms. The compliance side is covered in AI phone answering and the SRA, and the difference between answering calls and completing intake work in AI receptionist vs AI employee. For other ways to free support staff time, see AI for practice management and admin tasks.
Frequently asked questions
Is an AI receptionist cheaper than a human receptionist?
It depends on volumes and the model you choose. Compare the full cost of employing someone, including employer National Insurance, pension, cover and recruitment, with the service charges, and remember that most firms combine the two.
Can an AI receptionist replace our receptionist?
Rarely in a firm with an office. People are better at welcome, judgement, distressed callers and existing clients. Automation works best for overflow, after-hours calls and consistent intake.
Which hybrid model suits a small firm?
Often overflow plus after-hours cover: the receptionist answers what they can, and the automated service takes unanswered and evening calls and sends notes for follow-up.
Do clients mind talking to an automated receptionist?
Many are comfortable if it is honest about being automated, handles simple tasks well and lets them reach a person quickly. Trust and human oversight are the main concerns.
References
Office for National Statistics (2026). Earnings and hours worked, occupation by four-digit SOC: ASHE Table 14 (2026 provisional) (opens in a new tab) ↑
GOV.UK (2026). National Minimum Wage and National Living Wage rates (opens in a new tab) ↑
GOV.UK (HMRC) (2026). Rates and thresholds for employers 2026 to 2027 (opens in a new tab) ↑
Legal Services Consumer Panel (2026). How consumers are using legal services (Tracker Survey 2026) (opens in a new tab) ↑
Legal Services Board (2024). Legal Needs Survey 2023: Summary report (opens in a new tab) ↑