A chiropractor opens Google's Keyword Planner to figure out what "chiropractor near me" will cost. The tool answers with a range: a top of page bid of $5.44 at the low end and $22.81 at the high end. The middle of that is about $14. Sixty new-patient clicks a month at $14 is roughly $850, so that becomes the budget.
That is the most common way a small business sets its first Google Ads budget, and it is advice we gave ourselves. In August, in Google Has No Minimum Spend. Your Market Does., we told readers to open Keyword Planner and "take the middle of that range as your working cost per click."
Then we checked. That chiropractor is a real account we manage, and over the 90 days to October 20 its clicks on "chiropractor near me" cost $23.77 each. Not $14. More than the top of the range. At that price, the $850 budget buys about 36 clicks, not 60.
One keyword is an anecdote, so we ran the same comparison across every account we manage that had enough search traffic to measure. The middle of Keyword Planner's range came in low on most of them. Here is what the numbers actually are, what we found when we put them next to real bills, and how to set a budget that survives contact with the real price.
Keyword Planner's Range Is a Spread of Other Advertisers' Bids, Not Your Price
Start with what Google itself says the numbers mean, because the labels invite a reading the definitions do not support.
The two bid columns are called "Top of page bid (low range)" and "Top of page bid (high range)." Google's help page defines the low end as "an approximation of the 20th percentile" of what "advertisers have historically paid for a keyword's top of page bid," and the high end as an approximation of the 80th percentile, for your chosen location. Those figures come from the last 30 days, and they "take into account the bids that resulted in an ad being shown among the ads at the top of the first page of search results." Keywords with limited bid history simply get no range.
So the range is a summary of other advertisers' bids for a top position over the past month. It describes the market. It is not a quote for your account, and Google never says it is.
What you are actually charged for a click is a separate thing. Google's own description is that "you pay only one cent more than the minimum amount required to keep your ad's position." That minimum depends on who else is in that particular auction, your ad's quality, and dozens of signals about that particular search. It can sit comfortably inside the range. It can also sit well outside it.
The search volume column has its own fine print. "Avg. monthly searches" counts a keyword "and its close variants," shown for exact match only, averaged over 12 months by default, and Google says plainly that "your search volume statistics are rounded." And in the section on forecasts, Google adds the warning that matters most to a local business: "You can expect results to be less accurate if you're targeting a small geographic location."
None of that is hidden. It is just written in a help article most owners never open, beneath a tool that presents its numbers with total confidence.
We Priced 182 Keywords Against Real Bills. Most Cost More Than the Middle of Keyword Planner's Bid Range.
Here is the test. We took every search keyword in the accounts we manage that had at least 10 clicks between July 23 and October 20, 2026, and recorded what each one actually cost per click. Then we asked Keyword Planner for its numbers on the same keywords, in the same locations those campaigns target. That gave us 182 keywords with a Keyword Planner bid range, spread across 16 accounts, carrying about 15,700 clicks and roughly $111,000 of real spend. The businesses range from a chiropractic clinic and a debate academy to an industrial equipment dealer and an education publisher.
Where the real cost per click landed:
- Below the low end of the range: 6% of keywords
- Inside the range: 47%
- Above the high end of the range: 47%
Weighted by clicks rather than keywords, 55% of all the clicks we paid for cost more than the top of Keyword Planner's range. The median keyword cost 1.5 times the middle of its range.
We tried to break that result, because a number like this deserves suspicion. Looking only at exact match keywords, where the comparison is cleanest, the median was 1.3 times the middle. Removing the one account that pulled the average up the most, it was still 1.3 times. Rerunning the whole thing on the last 30 days only, to match Keyword Planner's own 30-day window, it was 1.5 times. In 11 of the 16 accounts, the typical keyword cost more than the middle of its range. The result moved a little each time. It never flipped.
The keywords that came in under the low end are worth a look, because they are almost all the same kind of keyword. Nine of the 11 were the business's own name or its own event. Your own name is the cheapest click you will ever buy, because nobody else is as relevant to that search as you are. Keyword Planner does not know that, and prices your name like everyone else's.
We cannot see inside Google's calculation, so we will not pretend to know exactly why real prices run above the range. What we can tell you is what it does to a plan. If the median keyword costs 1.5 times the middle of its range, a budget built on the middle buys about a third fewer clicks than you planned. Your first month looks like a failure when it was really a pricing error.
A Third of the Keywords That Got Clicks Showed No Bid Range in Keyword Planner
The 182 keywords above are the ones Keyword Planner could price. They were not all of them.
Of the 279 keywords in our accounts that earned at least 10 clicks, 97 came back with no bid range at all. That is 35% of the keywords that were working, and together they took about 4,700 clicks. Many of them were also listed at 0 or 10 searches a month.
One example: an exact match keyword for a medical equipment maker, "functional rehabilitation equipment." Keyword Planner lists it at 10 searches a month, with no bid range. Over the same 90 days, it drew about 370 impressions a month, and Google reported that those impressions covered 83% of the searches it was eligible for. Ten searches a month, according to the planner. Hundreds, according to the account.
The pattern in the blank rows was consistent: they were the specific searches. Particular equipment, particular services, phrases with a product type and a use case in them. Those are often exactly the searches a small business wants most, because a specific search usually comes from someone closer to buying.
So treat a blank in Keyword Planner as what it is: Google telling you it does not have enough data to say. It does not mean nobody searches for this, and it certainly does not mean the click will be cheap. If the keyword describes exactly what you sell, test it with a small, fixed budget and let the account tell you the real numbers.
Keyword Planner Rounds Every Search Volume to One of a Few Dozen Numbers
Look closely at the search volumes Keyword Planner gave us across 279 keywords and you notice something odd: they only ever took one of 30 values, plus zero. 10, 20, 30, 40, 50, 70, 90, 110, 140, 170, 210, 260, 320, 390, 480, 590, 720, 880, 1,000, 1,300, 1,600, 1,900, 2,400, 2,900, 3,600, and on up the ladder.
That is the rounding Google mentions, and it is coarser than most people realize. The steps get wider as the numbers get bigger, so a keyword shown at 1,000 could be several hundred searches away from one shown at 1,300, while one shown at 140 might really be closer to 170.
The rounding was not the only issue. For exact match keywords, we could estimate how many searches each one was actually eligible for, because Google reports both the impressions and the share of eligible searches those impressions represent. The median keyword was eligible for about 1.6 times the volume Keyword Planner listed, and 43% of them for more than double. Some came in under. The ladder is a rough sketch, not a measurement.
That does not make volume useless. It makes it relative. Keyword Planner is good at telling you that one phrase is searched far more than another. It is poor at telling you how many clicks you will get next month. Use it to rank keywords against each other, not to forecast traffic, and certainly not to multiply against a bid to produce a budget.
In a Small Market, Keyword Planner's Bid Range Gets Too Wide to Budget From
Even when the range is honest, it can be too wide to be useful.
Across our 182 keywords, the high end of the range was a median of 4.4 times the low end. The chiropractor's range ran from $5.44 to $22.81. A search for portable toilet trucks for sale ran from $0.47 to $1.61. A conference keyword for an education publisher ran from $3.33 to $94.11, which is a 28-fold spread. When the top of the range is four times the bottom, "the middle" is not an estimate. It is a coin toss with extra steps.
This is the real-world version of Google's warning about small geographic areas. A local business is targeting a small slice of searches in a small slice of the map. There are fewer auctions, fewer advertisers, and less history for Keyword Planner to summarize, so the summary gets loose. The national ecommerce brand budgeting off a huge pool of searches can lean on the range far more than a dentist targeting three ZIP codes.
The Top of Keyword Planner's Range Was Closer to What Our Clients Paid
So what should an owner actually do with the tool? Our data points at a simple adjustment. The median keyword in our accounts cost about 95% of the high end of its range. The top of the range was not a ceiling, since plenty of keywords went over it, but it was a far better working estimate than the middle.
Here is the method we would use to set a first budget:
- Price from the high end. For each keyword you plan to run, take the top of page bid (high range) as your working cost per click. For a keyword with no range, write "unknown," not zero.
- Decide how many clicks you need to learn anything. A budget that buys five clicks a month teaches you nothing. Work out the number of clicks it takes to expect a handful of leads at a realistic conversion rate.
- Multiply, and run it as a fixed test. High-end cost per click times the clicks you need is your first month's budget. Hold it steady for two to four weeks. Do not adjust it every few days.
- Replace the estimate with your own number. After the test, pull your actual average cost per click by keyword from the account. That is your price now. Keyword Planner's range has done its job and can go back in the drawer.
Run the chiropractor's numbers through it. Budgeting from the middle gave $848 for a target of 60 clicks, and the real price delivered about 36. Budgeting from the high end gives $1,369, which at the real price buys about 58. One plan sets you up to conclude Google Ads does not work. The other gets you the data you were trying to buy.
If that high-end number feels like more than you can spend, that is useful to know before you start, not after. The answer is to narrow the plan: fewer keywords, a tighter area, or the single service that makes you the most money. Not to quietly pretend the clicks cost less. We go deeper on sizing a first-month budget in Your First Ad Budget Isn't an Investment. It's Tuition., and on choosing a bid strategy once your own numbers arrive in You're Bidding for Clicks. You Wanted Customers.
Keyword Planner Is Good at Comparing Keywords, Not Pricing Them
None of this makes Keyword Planner a bad tool. It is a free look at how a market searches, and it is genuinely useful when you ask it the questions it can answer.
- Finding ideas. Type in your service and it will show you the phrases people actually use, including ones you would never have guessed.
- Comparing keywords with each other. Rounded volumes still tell you which of two phrases is searched far more, and the ranges tell you which one is the more contested auction.
- Seeing the seasons. The monthly trend shows when demand for your service rises and falls, which is worth knowing before you set a budget for the year.
- Catching bad searches before they cost you. The suggestion list is full of searches you do not want, like "free," "jobs," and do-it-yourself queries. Add those as negatives before you spend a dollar. We walk through that habit in You're Paying Google for Searches You'd Never Choose.
The one thing it should not be is your price list. Use it to decide what to test, price that test from the top of the range, and let your own account tell you what a click really costs. We have updated our August advice to say the same.
If you would rather have someone price your market from real account data instead of a planner's estimate, that is part of the everyday work of our Google Ads management team. And if you are setting this up yourself, start from the high end. Your first month will thank you.
