How Much Did CPC Rise Beyond Inflation? Clicks vs. the IPCA

Saying that CPC rose 10% in a year says little in a country where every price went up. The more useful question is how much the click got more expensive beyond inflation. We measured it with the same accounts in both years, and the result is the Marktech CPC inflation index.

Abroad, not even the biggest reports agree. In the second quarter of 2026, Tinuiti measured Google CPC as practically flat, up 1% in one year, while Skai saw CPCs rising in the same period, with spend up 17% and clicks flat. Both can be right, because each one measures its own portfolio. Hence the usefulness of a number calculated here, in reais and net of local inflation.

• Why nobody knows how much CPC rose
• How the index is calculated
• The result: nominal vs. the IPCA
• What the number doesn't say
• Next editions

Why nobody knows how much CPC rose

Three problems pile up. First, every report reflects the client mix of whoever publishes it, which is why the numbers diverge so much. Second, almost all of them measure in dollars and in the United States: LocaliQ, for example, measured a 12.88% CPC increase in 2025, with increases in 87% of industries, all in dollars. Third, and more serious for anyone advertising in reais: none of those numbers discounts inflation. In the 12 months ended July 2026, the IPCA accumulated 4.44%. Since the IPCA is Brazil's official inflation index, any nominal CPC increase (the figure in reais, before the discount) below that means the click rose less than other prices: in real terms, it got cheaper.

How the index is calculated

The central safeguard is comparing the same accounts, under the same conditions, in both periods:

• Balanced panel: only accounts active in both periods enter, July 2025 and July 2026: a sample of 45 portfolio accounts, the same ones in both years. The name comes from statistics and the reason is practical: if the group changes from one year to the next, part of the variation comes from swapping accounts, not from the price of the click;
• Same ruler as the benchmark: keyword-based search campaigns, Google Search Network, no Performance Max, accounts in reais, at least 1,000 impressions in each of the two months. These filters leave many accounts out: what remains is a comparable sample, a fraction of the portfolio;
• Aggregate CPC: total cost divided by total clicks in each period. It is a volume-weighted average: accounts with more clicks weigh more in the number. The change between the two periods is the nominal change;
• Deflation: the real change discounts the IPCA, that is, it removes from the CPC increase the part that merely tracks general inflation, leaving the click's own price increase. The math: one plus the nominal change, divided by one plus the IPCA, minus one;
• Second reading: the median of the per-account changes (the middle value: half the accounts changed less, half changed more), in which every account weighs the same, and the share of accounts with CPC on the rise.

The result: nominal vs. the IPCA

In the panel, aggregate CPC went from R$ 1.78 in July 2025 to R$ 2.29 in July 2026: a nominal change of 28.8%. Net of the 4.44% IPCA, the real change was 23.4%. In the per-account median, the change was 6.2%, and CPC rose in 51% of the accounts in the sample.

The number that answers the question in the title is the second one, the real change. The nominal change mixes two movements: the economy's general inflation, which makes everything more expensive, and the price increase specific to the auction, which is what matters to anyone planning media.

The distance between the aggregate and the median also deserves a reading. The aggregate, weighted by volume, is dominated by the accounts that buy the most clicks; the median, in which every account weighs the same, describes the typical account in the sample. In this edition, the two together indicate that the increase was driven by the higher-volume accounts, while in the typical account the click rose only slightly above inflation.

What the number doesn't say

The index measures the size of the increase, not its cause. A real CPC increase can come from genuinely new competition, but also from the mechanisms we detail in why CPC rises without any new competitor: Google's own pricing instruments, advertisers migrating to automated bidding (when Google itself sets the bid in each auction) and changes to the page layout. Separating those causes requires other evidence, and that is left for the next editions.

On the limits: the panel is a sample of a fraction of the Marktech portfolio, cut by the filters above, not the entire Brazilian market. Comparing the same month across two years controls for seasonality (the swings that repeat every year), but does not eliminate one-off events. And a sample of 45 accounts allows reading trends, without replacing a survey of the whole market.

Next editions

The intention is to recalculate the index from time to time, always with a balanced panel and the same rules, alongside the CTR, CPC and CPM benchmark we publish with it. As the editions accumulate, the series gains history: it will become possible to see acceleration, deceleration and the effect of events like Black Friday.

If your CPC rose more than the index, it is worth separating what belongs to your industry from what belongs to your account: ad quality, the auction in your region and campaign structure usually explain a good part of the difference, and those are the points where management can act.