
Context: Four Branches, Two Targeting Systems That Never Agree
A home-services group with branches in Sandton, Centurion, Benoni and Krugersdorp finds its Google Ads bidding on suburbs it cannot service, while its Google Business Profile (GBP) ranks for areas it does not target. Ads location settings and GBP service areas run on different assumptions: Ads targets where users are located or show interest, GBP defines where you serve. When they disagree, budget leaks and local visibility suffers.
Until you reconcile these systems, every rand spent on local search risks chasing a postcode you cannot profitably serve.
A Pretoria dental group with three clinics within 20 km cannibalises its own Ads and map pack rankings because catchments overlap. Location bid modifiers can separate them, but only after you define each branch's true service radius.
Why 'Gauteng' Is a Targeting Mistake Before It Is a Strategy Mistake
Gauteng is 15 million people across three metros and dozens of suburbs. Treating it as one market blurs budget and floods ads into areas you cannot service quickly. Within Johannesburg alone, Sandton, Soweto, Randburg and Midrand have different competitive dynamics. A single 'Gauteng' campaign cannot distinguish a Benoni homeowner needing plumbing from a Sandton office needing after-hours IT support.
Treat these as separate markets: Johannesburg (Sandton, Soweto, Randburg, Midrand), Pretoria (Centurion, Pretoria East), Ekurhuleni (Benoni, Boksburg) and the West Rand (Krugersdorp). Google's location targeting documentation explains the settings; mapping them to your business is yours.
Presence or Interest vs Presence Only
The default 'presence or interest' setting shows ads to people in, or showing interest in, your targeted locations - a Soweto user searching 'plumber' could see your Sandton ad. For local services, switch to 'presence only' to restrict ads to people physically in your target area. For fixed-location businesses like a dental clinic, presence only is usually correct; for delivery or remote services, presence or interest may capture out-of-area demand.
Validate that location targeting matches operational capacity. Serving a lead you cannot reach in reasonable time erodes margin and ruins reviews.
Radius vs Polygon: Setting a Realistic Catchment
Radius targeting draws a circle around an address, set in kilometres in South Africa. A 15 km radius around a Benoni branch might include parts of Boksburg and Brakpan but exclude Springs. If your average travel time is 30 minutes, 15 km is more realistic than 25 km in peak traffic. Use the radius tool to visualise branch overlaps; for irregular areas, combine radius with postal code targeting.
Test radii in a draft campaign and use the 'Locations' report to see where ads actually served. Adjust on cost per qualified lead, not clicks.
Location Bid Modifiers and Multi-Branch Cannibalisation
The Suburb Page Threshold: Only Build Pages Your GBP Genuinely Covers
Creating dozens of suburb pages for areas with no physical presence or service area is a common mistake; Google may ignore them or treat them as low-quality. A sensible threshold: only build a page if you have a branch within 15 km or have set that suburb as a GBP service area. A Benoni branch might warrant Benoni, Boksburg and Brakpan pages, but not Sandton. Each page needs unique content - local landmarks, branch-specific staff, area FAQs.
Quality over quantity: five well-optimised suburb pages that convert beat fifty thin pages Google ignores.
Call Tracking, Direction Requests and Store Visits
Local intent often converts offline. Implement call tracking (Google forwarding numbers) and track direction requests and store visits where eligible. Call reporting availability in South Africa follows Google's documentation; store visits reporting has unpublished eligibility thresholds, so not all advertisers see this data. Use a third-party rank tracker to monitor map pack impression share for key suburb queries, and compare cost per qualified lead (CQL) by branch - a call lead is often worth more than a form fill. Blended CPA hides branch-level problems.
Define qualified lead criteria (e.g., call duration over 2 minutes, form with valid address), track by branch via UTM parameters and call tracking numbers, and review monthly to shift budget from weak branches to high-yield ones. For a structured approach, see Prebo Digital's custom reporting methodology.
Where POPIA and Consent Mode Change the Local Measurement Story
South Africa's Protection of Personal Information Act (POPIA) took full effect on 1 July 2021 and is enforced by the Information Regulator. When collecting location data or call recordings, obtain consent. Google's Consent Mode respects user consent choices while maintaining measurement; Consent Mode v2 is essential for accurate local conversion tracking. Without it, you lose visibility into conversions from users who decline tracking, skewing branch-level attribution.
A Comparison Table: Google Ads Local Targeting vs Local SEO Effort
| Aspect | Google Ads Local Targeting | Local SEO (GBP + Suburb Pages) |
|---|---|---|
| Time to impact | Immediate (once live) | Weeks to months |
| Cost model | Pay-per-click (ZAR per click) | Fixed investment (time/resources) |
| Control | High (precise radius, bid modifiers) | Moderate (influenced by GBP and content) |
| Measurement | Clicks, conversions, CQL per branch | Map pack impressions, direction requests, calls |
| Best for | Urgent, high-intent leads; testing new areas | Long-term presence; building local authority |



