What you'll learn in this article
- The average CPC ranges I actually see, sector by sector, across the accounts I manage
- Why the published industry benchmark was wrong for six of my nine accounts
- The three variables that move keyword price more than the industry label does
- The sequence I use to run a price check before committing budget to a new account
- What a CPC far above or far below benchmark tells me about account structure
Every quarter I pull average CPC by sector across the accounts I manage and compare it against the benchmark tables that circulate online. The gap is consistently large enough that I stopped using published tables as anything more than a rough sanity check. When a client asks me what the google ads keyword price is in their vertical, the honest answer is a range with three conditions attached and the conditions matter more than the range.
What follows are the numbers I use as my own reference points, where they come from, and more useful than the numbers themselves the reasoning I apply when an account lands outside them.
How I built these benchmarks, so you know what they're worth
Nine accounts, all Search, all Italian or single-country European targeting, spend between €3,000 and €40,000 a month. I pulled trailing 90-day average CPC at campaign level, then excluded brand campaigns entirely brand terms sit at €0.15–€0.60 almost everywhere and drag every blended average toward meaninglessness. Every figure below is non-brand generic traffic only.
What the price actually is, mechanically
Before comparing anything it helps to be precise about what you're comparing. The number in your account is actual CPC, not your bid. Google's documentation on actual cost-per-click explains that you're charged the minimum needed to clear the Ad Rank thresholds and beat the advertiser immediately below you, which is why actual CPC is usually below max CPC and why two advertisers with identical bids on the same term pay different prices. That single mechanism is the reason a sector average tells you so little on its own.
The three variables that outweigh the sector
Across the nine accounts, the industry label explained far less of the variance than these did: geographic concentration (a city-level campaign in a dense market prices differently from national coverage), match type mix, and ad quality. Two accounts in the same vertical ran 2.3x apart on CPC, and the entire difference sat in those three variables.
The numbers I use as reference, by sector
These are ranges, not points. The low end is what a well-structured account with good quality signals achieves; the high end is what I see on accounts that are working but not optimised.
Legal and professional services: €4.50–€14
The most expensive vertical I work in, and the most variable. Specialist terms in competitive cities reach the top of that range easily. What surprised me is that the spread within a single account is often 5x the same campaign will have terms at €5 and terms at €25. An average here is nearly useless; I plan at the term level.
B2B software and services: €3.20–€9
Lower than legal, but with a longer conversion lag that makes the price feel higher than it is. The accounts I manage in this space cluster around €5–€6 on generic solution terms. Competitor terms run 30–50% cheaper but convert far worse, which is a trade I evaluate separately when I look at bidding on competitor terms.
Home services and trades: €2.80–€8.50
Emergency-intent terms anything with "urgent", "24h", "now" sit at the top of that band consistently. Non-urgent variants of the identical service can be a third of the price. This is the clearest example I have of intent, not industry, setting the price.
Ecommerce and retail: €0.35–€1.60
By far the cheapest clicks and by far the highest volume. Product-level terms sit at the bottom, category terms at the top. Anyone benchmarking an ecommerce account against a services benchmark will conclude their account is broken when it isn't.
Health and wellness services: €1.80–€6
Wide because it mixes regulated and unregulated subcategories. Aesthetic and dental terms price toward the top; general wellness sits low.
Education and training: €1.40–€5.50
Strongly seasonal. The same term can move 40% between enrolment season and off-season, which means a single-quarter benchmark for this vertical is misleading unless you note which quarter it came from.
How I run a google adwords keyword price check on a new account
When I take over an account or scope a new one, this is the order I work in. It takes about an hour and it has never failed to change my initial budget estimate.
Step 1 Get the actual number before the estimated one
If there's history in the account, trailing 90-day non-brand CPC by campaign is the truth. Planner estimates are a forecast built on other advertisers' behaviour; the account's own data is a measurement. I only fall back on estimates when there is genuinely no history.
Step 2 Split by intent tier, not by campaign
I bucket terms into three tiers: transactional, comparative, informational. Every account I've done this on shows a clean price gradient across the tiers, usually 3–4x from bottom to top. That gradient is more actionable than any industry average, because it tells you where the budget is going rather than what it costs on paper.
Step 3 Check the geographic layer
A national campaign averages together a dense expensive market and a thin cheap one. I segment by location before drawing conclusions. On one home services account, removing two metro areas from the average dropped the figure by 34% — the "sector price" was mostly two cities.
Step 4 Read the price against quality, not in isolation
Because actual CPC is discounted by ad quality, a CPC sitting above benchmark is often a quality problem wearing a pricing costume. Before assuming a vertical is expensive, I check whether the account is paying a quality premium the relationship between keyword quality signals and what you pay is where I've recovered the most budget on inherited accounts.
Step 5 Convert price into a decision
A keyword cost figure means nothing until it meets a conversion rate and a value. €12 per click is cheap for a term that converts at 8% into a €4,000 contract and ruinous for one converting at 1% into a €90 order. I never quote a price range to a client without pairing it with the volume needed to make it work, which is the same logic I apply across keyword bidding decisions generally.
What I infer when the price is off benchmark
CPC far above the range with normal conversion rate: a quality or structure issue, not a market issue. The auction is charging you a premium your competitors aren't paying. I check ad relevance and landing page alignment before I touch bids.
CPC far above the range with poor conversion rate: a matching issue. You're winning auctions you shouldn't be in. The price is a symptom; the search terms report holds the cause.
CPC far below the range with low impression share: you're not really competing. Cheap clicks with 15% impression share means you're picking up leftovers at the bottom of the page. That's not a bargain, it's absence.
CPC far below the range with healthy impression share: strong quality signals. Rare and worth understanding, because whatever is producing it is usually replicable in the account's other campaigns.
A price that moves sharply without any change on your side: a new entrant or a competitor changing strategy. I check auction insights before assuming anything internal broke. Roughly a third of the sudden CPC jumps I've investigated turned out to be someone else's budget decision.
A wide intra-account spread is normal; a narrow one is suspicious. If every term in a campaign prices within 20% of the others, the keyword set is probably too homogeneous to be covering the funnel.
What I stopped doing
Quoting a single industry average to clients. It sets an expectation that the account will then fail to meet for reasons that have nothing to do with the sector. I quote a range with the three conditions attached.
Comparing my accounts to published US benchmark tables. Different market, different currency, different competitive density. The directional ranking of sectors transfers; the absolute numbers do not.
Treating brand and non-brand in the same average. Brand terms are so cheap they hide everything. Every benchmark I keep is non-brand only.
Using the estimate as a plan. Estimates are a starting point for accounts with no history and nothing more. I've seen forecast and actual diverge by more than half in both directions.
Assuming an expensive keyword is a bad keyword. The most expensive terms in my legal accounts are also the ones producing the profitable cases. Price is an input to a value calculation, not a verdict.
The practical takeaway
The google ads keyword price in your vertical is a range whose width is set by intent, geography and quality and those three move the number more than the industry label. Use the sector benchmarks above to know roughly where you should land, then use the intent-tier split to know why you're landing where you are.
If you take one thing from the nine accounts behind these figures: a keyword cost that sits outside benchmark is diagnostic information, not a verdict on the market. Six times out of ten it's telling you something fixable about the account rather than something fixed about the auction.