Context: a ZAR 45,000 mail drop across three Johannesburg suburbs
Picture a home-services operator in Johannesburg - say, an installer of home security and gate automation working the northern suburbs. They already run a digital funnel (Google Ads, Meta retargeting, a booking form wired into their CRM) and want to reach households in three adjacent suburbs their search budget never quite touches. Their plan is a physical mail drop: roughly ZAR 40,000-60,000, an illustrative estimate covering print, list, postage and creative across a few thousand households.
Most operators run that drop, get a vague trickle of calls, and never know whether it worked. The alternative is to treat the drop as a channel with the same discipline as paid search: clean list, identifiable response events, a defined response window, and a cost per response expressed in rand.
The framing principle: a mail drop is not a brand exercise, it's a geo-fenced acquisition channel. If you cannot attribute a response to it, you cannot optimise it.
How a physical drop behaves differently to a feed impression
A feed impression is fleeting and paid for per thousand. A mailer sits on a kitchen counter and is seen repeatedly by the same household over days. Keep the claim qualitative - we are not asserting response rates - but the implication matters: the mailer can carry a proposition too long for a 6-second ad and survive a distracted scroll. The trade-off is that physical mail gives you no platform dashboard: no trustworthy impression count and no automatic conversion tracking. Everything measurable has to be engineered into the piece before it goes to print.
Building the list: geo-selector, address hygiene and suppression before anything else
List quality is the binding constraint on a local direct mail strategy. Work in this order:
- Geo-selector by suburb: define exact suburb boundaries, or build by postcode-plus-street segments so you can report by segment later.
- Address correction and de-duplication: one household, one mailer - duplicates are pure waste and look careless.
- Suppression list: existing customers, staff addresses, recent complaints, and anyone who has asked not to be mailed. This list must persist between drops.
If your data team already cleans conversion data before it reaches GA4, apply the same hygiene to a postal file. A dirty postal list produces the same problem as dirty analytics: confident decisions made on wrong inputs.
POPIA and the direct-marketing boundary: what physical mail does and does not permit
South Africa's Protection of Personal Information Act (POPIA) draws a distinction between electronic direct marketing - where consent and opt-out rules bite hard - and physical mail, which sits on different footing. That does not make addressed or unaddressed mail a compliance-free zone. It means you need a defensible, documented process: where the list came from, how you handle opt-outs, and how you record suppression.
Do not rely on this article for legal advice. Check your process against the Information Regulator and DMASA guidance, and get your own legal sign-off before a large drop.
The bridge: turning a delivered mailer into a measurable digital event (QR, short link, UTM)
The whole channel lives or dies on this step. Every mailer needs a unique, attributable route back to your site:
- A short offer code printed on the piece, unique per suburb or drop wave, so a phone or in-store redemption can be traced.
- A QR code pointing to a short-link or PURL that redirects to a landing page with campaign parameters attached.
- UTM parameters on the destination so GA4 records the visit as a distinct source, not stray direct traffic.
https://go.yourdomain.co.za/northgate-home
?utm_source=direct_mail
&utm_medium=print
&utm_campaign=northgate_q3
&utm_content=mailer_a
&utm_id=drop_2026_03Follow Google's UTM conventions so naming stays consistent with your paid channels, and route the landing page through the same CRO testing regime you apply to ad traffic - see our conversion rate optimisation services for how that typically runs.
Stacking mail underneath your geo-targeted digital so the same household hears one message
A mailer works harder when it is not the only touch. Run suburb-level search and Meta campaigns over the same households in the same window, with one consistent proposition and one consistent offer code. When a resident sees the mailer and then searches your brand, the click lands on a page that matches the piece - and your digital channel gets credit for a response the physical drop seeded. That crossover is why you need matchback rather than last-click.
The matchback model: response window, incrementality and cost per response in rand
Define your response window before the drop - 30 or 60 days is common - and count every qualifying response inside it: form fills logged with the offer code, calls from the tracked number, QR visits that convert, and in-store redemptions. Then compute two figures you can act on:
| Response source | Signal captured | Attribution method |
|---|---|---|
| QR scan | UTM visit in GA4 | Session source plus landing page |
| Offer code | Code entered at redemption | CRM record tagged to drop wave |
| Phone call | Tracked number by suburb | Call log matched to campaign |
| Brand search lift | Search volume in the suburb | Pre/post comparison, directional only |
Incrementality is the harder question. For a mid-sized drop, a reasonable approach is a holdout: leave a sample of the list unmailed and compare response between mailed and held-out households over the same window. Treat the result as directional, not proof. Our custom reporting approach applies the same logic to digital: attribution accuracy over platform-reported conversions.
Per-drop unit economics: what 1,000 mailed households actually costs (estimate ranges)
Figures below are illustrative South African estimate ranges, not benchmarks. Get live quotes from printers and mail houses.
| Cost line | Basis | Estimate per 1,000 households (ZAR) |
|---|---|---|
| Single-sheet mailer, standard stock | 2,500 - 6,000 | |
| List and address cleaning | Per record, including de-dupe | 500 - 2,000 |
| Postage | Bulk mail rate, metered | 3,000 - 8,000 |
| Creative and copy | One-off amortised across the drop | 1,000 - 4,000 |
Divide total spend by attributable responses to get cost per response, then compare it to your CAC from paid digital. If the mailer's cost per response sits alongside or below your digital CAC and leads close at similar rates, the channel earns a permanent line in the plan.
Which operator this suits: catchment service business, near-store e-commerce, in-house performance marketer
Three profiles fit well. A catchment service business (security, solar, medical aesthetics, home maintenance) with defined suburbs and a high-value job suits a tight, code-tracked drop. A store or e-commerce brand with a physical location and a radius of regular buyers suits a drop tied to a redemption offer. An in-house performance marketer with clean CRM and server-side tracking can run the matchback model without external help.
When the response is flat: a diagnostic sequence before you print another run
Work through this order: was the piece delivered (check mail house confirmation)? Did anyone scan (check UTM sessions at all)? Did scans convert (check landing page and form drop-off)? Did offline redemptions appear in the CRM? If scans were high but conversions low, the problem is the landing page, not the mailer. If scans were near zero, check QR prominence and offer clarity before blaming the list. Only then consider rebuilding the offer, suburb selection or creative. See the framework in action with a real-world example from our case studies.
Sources
- Information Regulator (South Africa) - POPIA guidance, including direct marketing provisions.
- Direct Marketing Association of South Africa - industry body and code of practice.
- Google Analytics Help: Build campaign URLs with UTM parameters.
- USPS: Every Door Direct Mail - geographic list and route mechanics.
