The most common hesitation I hear from South African business owners considering an AI receptionist is: "I'm not sure it's worth the cost." Which is a completely reasonable thing to question. So let's actually work through the maths — honestly, with real numbers from real business types.
The ROI calculation for an AI receptionist comes down to three questions:
- How much revenue are you currently losing to missed calls?
- How much does the AI receptionist cost?
- What additional value does the automation (WhatsApp, reminders, reviews) add?
Question 1: How much are you losing to missed calls?
This is the starting point, and the answer is almost always higher than business owners expect. Here's a simple calculation framework:
Step 1: How many calls does your business receive per day? (Check your phone's recent calls log for a typical week)
Step 2: What percentage go unanswered? (National SA SME average is approximately 62%, but this varies significantly by business type)
Step 3: Of those unanswered calls, what percentage would have converted to a booking or job? (Conservative estimate: 40%)
Step 4: What's your average job or appointment value?
Example for a Johannesburg electrician:
- 25 calls per day
- 40% unanswered = 10 missed calls per day
- 40% would have converted = 4 potential jobs per day not captured
- Average job value: R4,000
- Daily lost revenue: R16,000
- Monthly (22 working days): R352,000
Even if these numbers are off by a factor of 5 — even if only 20% of the conservative estimate applies — that's still R70,400 per month in recoverable revenue.
Question 2: What does Agent Help cost?
For a trades business like our electrician example, the Professional plan at R5,999/month is the appropriate starting point — 500 included minutes, WhatsApp post-call notifications, calendar booking, and weekly analytics.
Total annual cost: R71,988.
Question 3: What does the ROI calculation look like?
Using conservative assumptions:
| Metric | Before Agent Help | After Agent Help |
|---|---|---|
| Calls answered rate | 38% | 100% |
| Potential jobs captured/day | 1.5 | ~3.5 |
| Monthly revenue recovered | Baseline | +R88,000 |
| No-show rate | 18% | ~9% |
| Monthly AI cost | R0 | R5,999 |
| Net monthly gain | — | ~R82,000 |
This is a conservative model. We're not counting the WhatsApp follow-up conversion improvement, the Google review growth, or the time saved on receptionist admin. We're purely looking at call recovery.
ROI case study: dental practice in Cape Town
A dental practice with 3 dentists and one receptionist was seeing an average of 12–15 no-shows per week across all three chairs. At R1,400 per appointment, that was approximately R18,200 in missed revenue weekly.
After implementing Agent Help's Professional plan and the WhatsApp reminder sequence:
- No-shows dropped from 15/week to 6/week within the first month
- Weekly no-show revenue recovery: ~R12,600
- Monthly recovery: ~R50,400
- Monthly Agent Help cost: R5,999
- Net monthly gain from no-show reduction alone: R44,401
That's before counting any of the additional appointments booked by the AI answering after-hours calls.
When does an AI receptionist pay for itself?
On average, Agent Help clients recover their full monthly subscription cost within the first week through previously missed calls alone. The longest it has taken a new client to see positive ROI is 18 days. Most see it within the first 3–5 days.
The reason it pays back so fast is structural: the cost of the AI is fixed and relatively low. The revenue it recovers is highly variable but tends to be much larger. Even capturing a single additional appointment or job per day at R1,000 average value produces R22,000/month in recovered revenue against a R2,999–R5,999/month cost.
Calculate your specific ROI
Use the free calculator on our website, or call us and we'll walk through the numbers with you.
