A two-partner firm opens Google Ads, sees "payroll services small business" in Keyword Planner and gets excited. About 9,900 searches a month nationwide, from business owners, the exact clients a firm wants for recurring monthly work. Then they see the price column. The top-of-page bid tops out near $295. For one click.
That number isn't a typo, and it isn't a sign that payroll clients are worth $295 a click to an accounting firm. It's a sign of who else is in the auction. This chapter is about bidding like a firm that knows which auctions it can win: pay for the top spot where people hire a CPA, and walk away from fights priced for somebody else's business model.
Payroll Clicks Run Up to $295. The Search for a CPA Tops Out Near $23.
Keyword Planner prices every search as a pair: a guess for a slow day and a guess for a busy one. We priced the accounting searches nationwide for this guide, and the spread is wild. Payroll sits alone at the top, with "payroll services small business" quoted at $63.53 to $294.70 and plain "payroll services" at $38.50 to $217.60. IRS help, small business accounting, fractional CFO work and bookkeeping fill the middle, between about $11 and $74. At the bottom are the searches where someone is hiring a preparer this week: "cpa near me" at $4.11 to $22.53, "tax preparation services" at $3.59 to $18.45 and "tax preparer near me" at $2.68 to $13.26.
Our example market, Boston, showed the same order, with payroll on top, as Chapter 5 first flagged. We didn't verify exactly who drives payroll prices that high, so we won't pretend to. What we can say is that when we captured a live Boston search for "payroll services small business," the ads on the page came from payroll software companies, not accounting firms. Whatever a payroll software subscription is worth to them, that's the company a firm keeps when it bids on payroll.
Meanwhile "cpa near me" draws 74,000 searches a month nationally, at a small fraction of the price. That's where a new account's money works hardest.
A Payroll Bid and a Tax-Prep Bid Should Never Share a Wallet
In Isolation, BrandRocket's name for the exact-match campaign, you set bids by hand, and you set them at the ad group level. Each service group from Chapter 4 gets its own bid. That's the whole point of the structure: individual tax, small business accounting, bookkeeping and payroll are different auctions with different prices, and one bid for all of them guarantees you'll overpay somewhere and get outbid somewhere else.
Start firm. A new account has no track record, so Google has no Quality Score history to reward. In the first weeks, bid near the high end of Keyword Planner's range on the groups that matter most, enough to show up at the top and start collecting data. Then ease back as the account proves itself. Google rewards a keyword whose ad and page match the search, so the rent on that top slot drops as those grades climb, and the bid can follow it down without giving up the spot.
Here's what that looks like on a real search. Keyword Planner quotes "cpa near me" at $4.11 to $22.53 nationally. An individual tax group built around it might open at $20 a click in January, high enough to land above the free listings while Google sizes up a stranger. Give it two full weeks, then read the scoreboard in the next section. If the firm is sitting on top nine times in ten, shave a dollar and read it again the following week. When the top share starts slipping under three in four, the last cut went one dollar too far; put it back. Bids here move like a thermostat, not a light switch.
If a firm decides payroll is worth testing, it gets its own group, its own bid and ideally its own budget ceiling. A payroll group sharing a campaign budget with tax preparation can burn through the day's money by lunchtime on a few $200 clicks, and the tax-prep ads never show.
Every Top Spot You Skip on "CPA Near Me" Rings Another Firm's Phone
Setting bids by hand without a scoreboard is doing a client's books without a bank statement. Google keeps the statement in two columns. Top impression share is the percentage of eligible searches where your firm landed in the ad block above the free listings. The absolute version counts only the times your firm held the number one position outright.
On "cpa near me" and the other searches that turn into engagement letters, keep that first number somewhere between three in four and nine in ten. Not every single time. The last few points of a busy auction are the priciest ones, like an extended return e-filed at 11:58 p.m. on October 15. But drop much below three in four and the firm down the street is picking up calls your ads helped create.
Falling short comes with an explanation attached. Google splits the missed searches into two piles, and each pile has its own cure:
- Lost to rank. Your ad didn't score high enough to make the cut. Before touching the bid, pull up that keyword's Quality Score and the three grades behind it from Chapter 4. A sharper ad or a page about the exact service often buys back the spot at a lower price.
- Lost to budget. The day's money was gone before the searches were. Add money, or trim the group to the phrases that actually sign clients so the budget still has something left at 4 p.m.
Auction Insights Names Who's Outbidding You. Some Accounting Auctions Are Worth Losing.
Google's auction insights report compares your performance with the other advertisers competing in the same auctions: how often they appear, how often they show above you, how often they take the top spot. One limit to know: the report doesn't show anything when your impression share is under 10%.
For an accounting firm, auction insights answers a practical question: who am I actually bidding against? On "cpa near me" it might be a few local practices and a national online accounting service. On payroll it might be software companies with very different economics. On tax resolution it might be national relief firms with large budgets.
Run payroll through the same chain Chapter 3 used, just to see the shape of it. Chapter 3's example paid $12 a click, turned 4.85% of clicks into inquiries and signed 1 in 3, for about $742 a client. Keep those two rates and swap in a $150 payroll click, comfortably inside that search's range. Each inquiry now costs about $3,100, and each signed client about $9,300. That's an illustration, not a forecast; your conversion rate on payroll could be better or worse. But it shows why the payroll auction is a different sport. A firm with a large, sticky payroll book might earn that back. A two-partner practice that runs payroll for eleven clients as a courtesy probably can't.
Then comes the hard call most firms never make: leaving an auction that has outgrown them. If a service's clicks push the cost per client above what a client in that service is worth, you have three moves. Fix quality if the problem is relevance. Lower the bid and accept fewer, cheaper wins. Or pause the group and move the money to the auctions you win at a price that makes sense. Leaving isn't losing. Staying in an auction priced for someone else's business is.
Broad Match Doesn't Know What a CPA Lead Is Worth. Your Hand Bids Already Do.
Exploration, the broad-match campaign Chapter 6 set up, lets Google set every price. Its strategy is Maximize Conversions, and the price cap it works under is copied straight off Isolation's receipts: the average your exact-match groups spent to get one inquiry from a would-be client.
If Isolation has been bringing in inquiries at $180 each, Exploration starts with a $180 target. That gives Google's automated bidding a real benchmark: find more people like the ones Isolation found, at about the price Isolation paid.
Two warnings. First, give it time. Automated bidding needs a learning period, and Exploration's first few weeks will look uneven. Don't tighten the target in week one because a few days ran expensive. Second, don't set the target lower than reality to squeeze the price. A target that's too tight starves the campaign of traffic, and it learns nothing.
A Bid Can't Fix a Call Nobody Answers
The best bid in the account is worthless if the phone rings out. Bidding decisions should be checked against what actually happens after the click, and for accounting firms, that's mostly phone calls.
Review call quality regularly. If you use Google's AI call scoring from Chapter 1, or a call-tracking tool that records calls, listen to a sample from each ad group. If you don't record calls, look at call lengths: a group that produces lots of 20-second calls is producing wrong numbers and hang-ups, not clients. Then feed what you learn back into the bids. The groups that bring real clients earn higher bids and more budget. The groups that bring noise get cut.
Bids should also follow the firm's calendar. In the first week of April, a practice that's already booked solid is paying full price for calls it will have to turn away. Pull the individual tax bids down, or pause that group until the deadline passes, and keep the small business and bookkeeping groups running, because those clients sign up year-round.
And check the other side of the phone. Responsiveness was the most common subject in the 622 client reviews we read for this guide. In Chapter 3's math, closing 1 in 4 leads instead of 1 in 3 raised the example cost per client from about $742 to about $990. No bid adjustment recovers that. Answering does.
Next: Headlines About Answering the Phone
Chapter 8 turns to the ads themselves: how to write headlines in your clients' own words, why reviews make better ad copy than marketing slogans, and why a firm that picks up on the second ring might have the strongest headline in town.
Earlier chapters: Chapter 6 on exact and broad match, and Chapter 3 on what a client is worth. The accounting guide's home page lists them all.
Deciding which auctions to fight and which to leave is a weekly part of our Google Ads management for accounting firms.
