Every sales-automation vendor has a slide that says leads answered in five minutes convert some enormous multiple better than leads answered in an hour. We are not going to quote that number, because we cannot source it for Zimbabwe and neither can they. Instead, this guide builds a small, transparent model for a Harare letting agency in which every assumption is visible, and shows how to find the break-even for your own business. The structure transfers to workshops, clinics and wholesalers; only the values change.
The setting, with sources
- Where the enquiries come from. Propertybook lists more than 8,500 properties and partners with over 100 agencies; property.co.zw showed 806 rental listings at the time of writing. Portal enquiries arrive by email alert and, increasingly, straight to the agent’s WhatsApp.
- What a letting is worth. Propertybook’s 2025 guide puts monthly rentals at roughly USD 450 to USD 4,500 across locations, with Harare commanding a premium over the national range. We use a mid-market USD 800 flat in the model.
- Where the tenant is. POTRAZ’s Q4 2025 report has WhatsApp at 20.69% of all mobile data. A good share of prospective tenants are on WhatsApp-only bundles (Econet sells weekly ones from USD 0.50), which means a listing link in a reply may not open for them; photos and viewing slots have to be sent in the chat.
- What a reply costs. Under Meta’s per-message pricing, a reply inside the 24-hour window the tenant opened is free.
The model
One agency, one month. The bold values are assumptions you should replace.
| Input | Value | Note |
|---|---|---|
| Enquiries per month | 150 | From portals and WhatsApp combined |
| Share arriving outside 08:00–17:00 weekdays | 45% | Tenants search at night and on weekends |
| Median time to first human reply today | 3 hours (next morning for evening enquiries) | Ask your agents; the CRM will tell you after month one |
| Viewing booked per 100 enquiries, today | 18 | Your current rate. Count it for a month. |
| Lettings per 100 viewings | 35 | Unchanged by automation; depends on stock and price |
| Agency fee per letting | USD 800 | e.g. one month’s rent on an USD 800 flat |
Today’s output: 150 enquiries → 27 viewings → about 9.5 lettings → ≈ USD 7,560 in fees.
What automation changes, and what it does not
The automation acknowledges every enquiry in under a minute with the listing reference, asks two questions (budget, move-in date), sends three photos in the chat, and offers viewing slots from the agent’s calendar. The agent still does the viewing and the negotiation.
It does not change lettings per viewing. It changes one number: viewings booked per 100 enquiries, because fewer tenants have moved on before anyone replied. How much it changes it is unknowable in advance for your agency, so instead of asserting a figure we ask a different question.
The break-even question
How many extra viewings per 100 enquiries does the automation need to produce to pay for itself?
Let C be the total monthly cost of the automation (our run fee plus platform costs; the pricing page lists the drivers and the scoping document gives the number). Each extra viewing is worth 0.35 × USD 800 = USD 280 in expected fees.
Break-even extra viewings per month = C ÷ 280. Per 100 enquiries = (C ÷ 280) ÷ 1.5.
| If C (USD/month) is | Extra viewings/month to break even | Extra viewings per 100 enquiries | Viewing rate needed (from 18) |
|---|---|---|---|
| 150 | 0.54 | 0.36 | 18.4 |
| 300 | 1.07 | 0.71 | 18.7 |
| 600 | 2.14 | 1.43 | 19.4 |
| 1,200 | 4.29 | 2.86 | 20.9 |
Read the last column. At a monthly cost of USD 600, the automation needs to lift the viewing rate from 18 to 19.4 per 100 enquiries, roughly one and a half extra viewings from every hundred people who asked, to pay for itself. Everything above that is margin. Whether it lifts it by 1.4 or by 8 is what the pilot measures, and the CRM report answers it in month one because it now records time-to-first-reply and viewing rate per source.
Why this is the honest version
A claim like “391% higher conversion” cannot be tested in your business; a break-even can. It also exposes the case where automation is not worth it: an agency with 30 enquiries a month and a USD 1,200 cost needs to lift its viewing rate by more than 14 points, which is implausible, and we would say so.
The same model for other businesses
- Workshop: enquiries → quotes → jobs; value per job; the automation lifts quote-to-job by chasing at 24h/72h.
- Clinic: bookings → attended; value per attended visit; the automation lifts attendance by reminders and refills cancellations.
- Wholesaler: enquiries → orders; average order margin; the automation lifts enquiry-to-order by answering in seconds instead of next morning.
In each case: current rate, value per unit, cost of the automation, break-even lift. Four numbers, all yours.
Two Zimbabwe details that decide the design
- Bundles. Photos and slots in the chat, never a link as the only path.
- Numbers. Every enquiry stored as +263 7x xxx xxxx so the tenant who enquired from the portal on Tuesday and WhatsApped on Thursday is one person with one thread, not two leads for two agents.
Bring your four numbers to the scoping worksheet; it captures the volume and the channel, and the scoping document carries the break-even for your build.