Google Ads cost in South Africa is media + management + landing readiness, not a CPC screenshot.

Ask that question to three different people and you get a CPC figure, a retainer figure, or an unhelpful "it depends", and all three can be technically true while still leaving you unable to actually set a budget. A number you can plan around needs three layers stacked on top of each other: media paid straight to Google, the people or process running the account, and the conversion system that decides whether a click ever turns into a conversation.

Media spend moves with the auction. Management is usually a flat monthly fee, sometimes with a percentage of spend added. Landing pages, tracking, creative and sales follow-up all sit outside the auction itself, and yet they are exactly what decides whether that auction was ever worth entering.

Nexus prices paid media management as Launch from R5,000/month (Meta or Google), Scale at R10,000/month (Meta + Google), and Dominate at R18,000/month, plus a once-off R1,800 setup fee and 10% of monthly ad spend. The media budget itself is billed directly by Google to you, not to us.

Cost per click in South Africa swings widely by industry, match type, quality score and competition. A high-intent professional-services search term often costs more per click than a broad awareness query, and can still work out cheaper per customer once qualification and close rate are factored in properly.

Cost per lead is a better number than CPC, but only when "lead" actually means something to the sales team reading the report. A form fill from a researcher three provinces outside your service area is not the same unit as a booked consultation from someone with budget and the authority to say yes.

Build the model in reverse: average contribution per new customer, multiplied by close rate from qualified lead, gives you an allowable acquisition cost. That single number sets the ceiling for what media plus management can spend to produce that outcome.

The cheapest click is not the cheapest customer. Optimise for qualified conversations your team can win.

Nexus paid media principle

A boutique conveyancing attorney sets a R15,000/month Google Ads budget, split loosely across "attorney," "conveyancing" and "property lawyer" terms. Three months in, cost per lead looks reasonable, but almost none of the leads have converted into instructed matters. A closer look at the search terms report shows the bulk of spend going to broad match "attorney" queries, people searching for divorce lawyers, labour lawyers and criminal defence attorneys, none of whom needed conveyancing at all.

The fix cost nothing extra: tightening match types, adding a long list of negative keywords for unrelated legal specialities, and shifting spend toward exact and phrase match on conveyancing-specific terms. The same R15,000 budget, redirected, produced roughly a third of the previous click volume and several times the qualified enquiry rate, because the auction had been quietly subsidising searches that were never going to convert in the first place.

This is the most common, most avoidable Google Ads mistake among SA SMEs: treating broad match as a volume lever rather than recognising it as an invitation for the algorithm to guess broadly, sometimes wildly, at what counts as relevant.

Set a learning budget large enough to actually produce a decision, not so large that an unproven hypothesis becomes an expensive accident. Too little volume and you cannot tell whether the problem is the offer, the audience or the page. As a rough sanity check, aim for enough spend to generate at least 20 to 30 clicks per ad group per week, below that, the auction has not gathered enough signal for the algorithm to optimise, or for you to trust what it is telling you.

Protect your geo and language settings carefully. Broad national targeting for a business that only physically serves one metro is a fast way to waste budget and pollute your own learning, every click from a suburb you cannot service is a click that should never have entered the auction in the first place.

Keep brand search separate from non-brand from day one. Brand campaigns tend to look efficient because the person searching your name already intends to contact you, do not let that efficiency create a false sense that "Google Ads works" while the harder, more valuable job of capturing non-brand demand is quietly failing next to it.

• Primary offer and CTA clear on mobile above the fold
• Conversion tracking verified (form, call, WhatsApp, purchase)
• Landing page matches ad intent and promise
• Negative keyword and search-term review rhythm exists
• Sales response SLA defined for new leads
• Qualified-lead definition agreed with sales
• Monthly decision rule: scale, hold, fix or cut

1. Define the commercial unitQualified lead, booked meeting or purchase, with a written definition.
2. Instrument conversionsGA4/Google Ads conversion actions tested on desktop and mobile.
3. Build a tight structureIntent themes, geos you can serve, brand vs non-brand separation.
4. Ship a matched landing experienceSame promise, proof near CTA, fast mobile contact path.
5. Review weekly, decide monthlySearch terms and lead quality weekly; budget allocation monthly.

Running Google Ads in-house can work well when someone on the team genuinely has the time, skill and authority to cut waste on an ongoing basis. It falls apart when the account quietly becomes a weekend hobby, no negative keyword reviews, no landing tests, no lead-quality feedback loop, while the platform happily keeps spending regardless.

Agency management earns its fee when it buys better account structure, faster creative iteration and reporting you can actually trust, not when it produces a dashboard nobody reads critically. Ask what actually ships each month and how decisions get made. If the honest answer is "we monitor it," that is not management. That is observation with an invoice attached.

Nexus Launch (R5,000/month) suits a single-platform start. Scale (R10,000/month) fits Meta + Google running together with real testing depth. Dominate (R18,000/month) adds fuller-funnel coordination and tighter CRM or offline tracking, useful once qualified-lead reporting has to reconcile against actual closed deals, not just a rising count of form submissions.

Auction competition is not flat across the year. Categories tied to January new-year resolutions, tax season, back-to-school or festive-season retail see cost per click spike as more advertisers compete for the same limited attention window, and a budget that looked generous in a quiet month can suddenly under-deliver in a competitive one.

Plan around your own category’s actual demand curve rather than a flat monthly figure carried forward unchanged. A business that knows its busiest enquiry period lines up with a seasonal spike should shift budget toward that window in advance, rather than reacting to rising CPCs once the competition has already arrived.

If your website cannot convert, fix that before you raise a single rand of media spend. If tracking is broken, fix that before you judge cost per lead on numbers you cannot trust. If sales response is slow, fix that before you blame Google for a problem that starts three steps after the click.

Read this alongside our paid-vs-SEO and marketing budget guides to see how Google fits the wider channel mix, then set a capped test with a genuine kill criterion written down before you launch it.

Ready to scope management around a real media budget? Take the Growth Plan quiz or contact Nexus with your offer, geo and monthly media range, and we will tell you plainly what is realistic.