Your Ad Costs Didn’t Spike 45%. They Compounded 22% — And Won’t Stop

The number everyone is repeating, and where it actually comes from

The “45% since 2024” figure is everywhere in performance-marketing Slack channels and agency decks this September. It has the shape of a fact — specific, alarming, easy to repeat. It also does not appear in a single primary source. It traces back to a handful of vendor and content-mill blogs that cite no platform data, no research firm, and no methodology, repeating a number that looks copy-pasted from one post to the next.

Trace it further and it doesn’t dissolve into “a handful of blogs” — it resolves to one. Adtitudemedia.com’s post on rising CAC and platform dependency, written by the agency’s own co-founder, states flatly that “Meta and Google auction prices have risen roughly 45% compared to 2024” — no citation, no study, no named methodology. That same uncited paragraph is also the traceable origin of another number circulating in Indian D2C circles: the claim that blended CAC rose from ₹380 to ₹502. Two separate “statistics,” repeated across dozens of decks and Slack threads, both sourced to one unverified sentence on one agency’s blog.

That matters because if you walk into a budget meeting with a number you can’t defend, the conversation stops being about strategy and starts being about whether you did your homework. So before building a response, it’s worth replacing the number with one that survives scrutiny.

What Meta’s own filings actually say

Meta’s Q2 2026 results report average price-per-ad up 12% year-over-year, with ad impressions delivered up 14% YoY in the same quarter. Both numbers rising together is the detail most summaries drop: Meta isn’t just charging more per ad, it’s also selling more inventory. You’re paying more per unit into a bigger pool of competing bidders.

The same figures were carried in Meta’s official release as syndicated by PR Newswire — “average price per ad increased by 12% year-over-year” and “ad impressions… increased by 14% year-over-year” — confirming these are the company’s own disclosed numbers, not a third-party estimate.

Pull the quarter-by-quarter trend and a pattern shows up that a single “vs. 2024” number hides. Growth held at +9% in Q2 2025 and +10% in Q3 2025, dipped to +6% in Q4 2025, then stepped up to +12% for two straight quarters — Q1 and Q2 2026. That Q4 dip is real and worth remembering: prices don’t rise every quarter without exception. But the step-up to +12% and its holding there for two quarters is what analysts who track Meta’s disclosures are reading as a new baseline, not a blip that reverts.

Compound the two-year figures — 1.09 (Q2 2025) × 1.12 (Q2 2026) — and you get roughly 22% higher price-per-ad in Q2 2026 versus Q2 2024. That’s the defensible number. It’s lower than 45%, and it’s a worse problem, because 22% compounding quietly over eight quarters is exactly the kind of drift that doesn’t trigger an alarm until the founder asks why the channel stopped working.

Google is a different story — don’t lump the two platforms together

The same viral claims usually bundle Google in with Meta, and the data doesn’t support that. Alphabet’s Q2 2026 Search & other advertising revenue rose 17% YoY to $63.3 billion, and total advertising revenue reached $81.6B, up 14.5% YoY. But that growth is coming mostly from more people clicking paid ads, not from Google charging more per click. AI Overviews and other search-results changes are pushing more of the total click volume into paid slots, which grows revenue without necessarily inflating your CPC the way Meta’s auction is inflating price-per-ad.

The split behind that revenue growth is now confirmed, not just directional: Google Search paid clicks rose roughly 13% YoY in Q2 2026 while cost-per-click rose only about 3% YoY. What’s solid is the shape of the story: Google’s 2026 cost pressure on you is coming through more competition for the same slots, not a platform-wide price hike. If your paid-search costs are climbing faster than that, the driver is more likely category crowding than Google itself.

Independent agency trackers land in the same range, which is the point — nobody measuring this from the outside gets anywhere close to 45%. Tinuiti’s Q2 2026 benchmark report puts Meta CPM up 10% YoY (Facebook +13%, Instagram roughly flat) and Google Search CPC up just 1% YoY, with Google Shopping CPC actually down 1%. Skai’s Q2 2026 report shows paid social CPM up 9% and paid search CPC up 14%. Line them up against Meta’s own 12% and Google’s own 3%: five independent measurements, all sitting inside a 0–20% band. Not one comes close to 45% — it isn’t just unsourced, it’s off by two to four times what any real measurement shows.

Confirmed split, Q2 2026: Google Search paid clicks +13% YoY, CPC +3% YoY — growth is volume-driven, not price-driven.

What this actually costs you, in rupees

Platform-level price-per-ad inflation is not the same thing as your account’s CAC. Your own creative fatigue, audience saturation, and targeting precision move your numbers independently of the auction. A brand claiming a 40% CAC jump cannot honestly attribute all of it to “the auction” — some of that is buyable back through better creative and audience work, since impressions delivered are also up double digits, meaning more inventory is available, not just pricier inventory.

Still, the platform-level floor sets a real cost baseline. Here’s what the verified 22% two-year rise looks like against a real Mumbai account:

This holds targeting and creative constant and applies only the verified platform-level price rise — it’s the floor, not the ceiling. If your account’s actual CAC has moved further than this, the gap between ₹550 and your real number is your own account’s story: fatigue, saturation, or category competition, and it’s fixable with account work rather than budget pleading.

India’s D2C market itself is part of why the auction keeps tightening: it’s commonly benchmarked at roughly $108B in 2026 with about 24% CAGR, which means more brands bidding into the same Meta and Google inventory every quarter — a structural, demand-side reason prices climb even without a platform algorithm change. The sourced India-wide number backs up the pressure without backing up the panic: the dentsu–e4m Digital Advertising Report 2026 puts India’s digital ad spend up 19% YoY to ₹71,621 crore in 2025. That’s genuine, citable growth — and neither that report nor any other major Indian industry study publishes an auction-price inflation figure anywhere near 45%.

For reference, India remains one of the cheapest major markets for this inventory in absolute terms — Meta CPMs are commonly benchmarked around ₹50-400 against a global blended CPM near $13-14, and Google Ads CPC for e-commerce/D2C keywords typically runs ₹15-80 depending on keyword type — branded terms near ₹25, category terms ₹45-80. These are industry-benchmark aggregations, not numbers either platform discloses directly, so use them as a sanity check on your own bids, not as a substitute for your actual account data.

The framework: build your own 2024-baseline tracker

The fix for “I don’t have the receipts” is a tracker you fill in once and update quarterly. It answers one question your founder will actually ask: is your drift in line with the platform, ahead of it, or behind it. Pull your Q2 2024 CPA and your current-quarter CPA for one platform, mark whether your creative has been refreshed in the last 90 days, and the tool below reads out where you stand against the verified 22% baseline.

Decision rule: hold, renegotiate, or shift mix

Once your tracker is filled in, the decision isn’t a vibe, it’s arithmetic. Drift at or below 22% with your ROAS still above your floor means hold current spend and mix. Drift above 22% with stale creative means fix the account before touching the budget — a visual refresh usually buys back more margin than a bid-strategy change. Drift above 22% with fresh creative already in rotation means category crowding, not a fixable account issue — bring the number to leadership and ask for a real budget increase or a mix shift toward owned/retention channels. Negative drift — your CPA has actually fallen versus Q2 2024 — means you’re outperforming the platform trend: document what’s working before scaling into it. The calculator above runs this logic on your own numbers.

The 45% number was never going to survive a citation check, and eventually someone in the room was going to ask for one. The 22% number will survive it, because it’s built from two figures Meta actually disclosed. That’s the trade worth making: a smaller, real number over a bigger, borrowed one.

The next step is mechanical, not strategic. Before your next budget conversation, pull your account’s actual Q2 2024 CPA per platform and run it against the 2024-Baseline Tracker above. You’ll know within twenty minutes whether your problem is the auction or your account — and which one you can actually fix this quarter.