What Is a Good CPA? How to Set Your Real Target in 2026

A good CPA is any cost per acquisition comfortably below what a customer is worth to you. That ceiling is your maximum allowable CPA, set by your gross margin and customer lifetime value, not by an industry average. If a customer delivers $400 in gross profit over their life, a $120 CPA is excellent and a $380 CPA is a slow bleed.

"Average CPA by industry" is the most searched and least useful number in paid media. Two companies selling the same product can have maximum allowable CPAs that differ by 5x on repeat purchase behavior alone. Here is the math that actually tells you whether your CPA is good, plus the traps that make your dashboard number wrong in both directions.

How to calculate CPA

CPA is total ad spend divided by conversions produced. Spend $6,000 on Google Ads, get 150 conversions, and your CPA is $40. Its simplicity is the problem: everything interesting hides in the word "conversion."

CPA = total ad spend ÷ conversions

If your conversion is a purchase, your CPA is a customer acquisition cost and compares directly to customer value. If it is a form fill, a demo request, or a phone call, your CPA is a cost per lead, and comparing that to customer value makes your ads look roughly ten times better than they are. Most disappointed advertisers I meet are quietly making this mistake.

Spend should also include everything you pay to run the ads, not just media. An agency taking 15 percent of a $10,000 budget makes your real CPA 15 percent higher than the platform reports.

CPA vs CAC: the gap most accounts ignore

CPA usually measures the cost of a conversion event, while CAC measures the cost of a paying customer, and your lead-to-customer close rate is the bridge between them:

CAC = CPA ÷ lead-to-customer close rate

Generate leads at $65, close 12 percent of them, and your real CAC is 65 ÷ 0.12 = $542. That is the number to compare against customer value. If leadership celebrates $65 while the business actually pays $542 per customer, every budget, pricing and headcount decision rests on a number that was never true.

The math cuts the other way too. Improving close rate from 12 to 18 percent drops CAC from $542 to $361 without touching the ad account. If your CPA looks stuck, the cheapest lever is often lead quality or speed to first contact, not bidding.

How to calculate your maximum allowable CPA

Your maximum allowable CPA is the gross profit a customer generates over their lifetime, and your target CPA sits well below it so there is room for overhead and profit. You can plug your numbers into our max allowable CPA calculator or work it by hand in three steps.

Gross profit per sale: average order value minus cost of goods, shipping, payment processing and fulfillment. A $150 order at 55 percent margin yields $82.50. Lifetime value: multiply by purchases per customer before they lapse. At 3.2 orders, LTV in gross profit terms is about $264. Use gross profit, not revenue; revenue LTV is a vanity number that has convinced plenty of founders to overspend. The target: most businesses aim for roughly a third of LTV gross profit, leaving two thirds for overhead and profit. On $264 of LTV that is a target near $88, with $264 as the ceiling you never cross.

Business	Gross profit per order	Orders per customer	LTV (gross profit)	Max allowable CPA	Healthy target CPA
DTC supplement brand	$28	4.5	$126	$126	$40 to $45
Furniture ecommerce (one-off)	$310	1.2	$372	$372	$120 to $125
B2B SaaS, $99/mo, 70% margin	$69/mo	22 months	$1,520	$1,520	$500
Local HVAC service	$540	2.1	$1,134	$1,134	$350 to $380
Law firm, personal injury	$6,000+	1.0	$6,000	$6,000	$1,800 to $2,000

The hardest input here is "orders per customer," and almost everyone guesses. Do not. Pull the real repeat rate from your order data, segmented by acquisition channel, because customers acquired on a discount code behave nothing like customers acquired on branded search. If SQL is not your thing, you can ask your customer database in plain English what a cohort actually repurchased over 18 months and get a real number instead of a hopeful one. That figure moves your allowable CPA more than any bidding change you will make this year.

Why CPA varies so wildly by industry and channel

CPA differences across industries come down to two things: how much a customer is worth, and how many advertisers are bidding on the same intent. Legal, insurance and B2B software tolerate four-figure acquisition costs because a customer is worth thousands. Apparel ecommerce cannot, because the first order clears $20 of profit. Channel matters as much as vertical. Typical US ranges in 2026, useful for sanity checking and not for target setting:

Channel	Typical cost per lead	Typical ecommerce CPA (purchase)	What it is good at
Google Search (non-brand)	$40 to $150	$30 to $90	Existing demand, high intent, fast payback
Google Search (brand)	$5 to $25	$8 to $25	Cheap, but often buying customers you already had
Performance Max / Shopping	n/a	$20 to $70	Volume on product catalogs
Meta (Facebook and Instagram)	$15 to $80	$25 to $75	Creating demand, cheap reach, creative-driven
TikTok	$10 to $60	$20 to $65	Low CPMs, younger buyers, fast creative burnout

Branded search always looks like the best line in the account and almost never is, incrementally speaking. Pause it, and if total orders barely move, you were paying to intercept people who typed your name. Defensible cost, not a growth one.

LTV:CAC ratio and payback period

The common healthy benchmark is an LTV:CAC ratio of 3:1, meaning a customer returns three times what you paid to get them. Below 2:1 the model is fragile: one bad retention quarter wipes out the margin. The more interesting failure is at the other end. A 5:1 or 6:1 usually means you are under-investing, having found something that works and not bought enough of it. The right response is to raise budgets until the ratio drifts toward 3:1, accepting a worse average for far more absolute profit.

Payback period decides whether you can act on that. It is CAC divided by monthly gross profit, so a $500 CAC against $69 per month waits about 7.2 months for its money back. Survivable if you are funded, fatal if you are not, since every new customer is a cash outflow first. Most self-funded businesses should target payback inside 3 months, which often means accepting a lower LTV:CAC ratio for faster cash recovery. If you think in revenue multiples instead, the same logic runs through return on ad spend, and you can pressure test the floor with a break-even ROAS calculator.

When a falling CPA is bad news

A CPA that drops while conversion volume collapses is not an improvement, it is a retreat. CPA is a ratio, and the easiest way to improve any ratio is to shrink the denominator: cut everything except your five best keywords and your warmest retargeting audience and your CPA looks fantastic on a business that is quietly starving.

Judge CPA and volume together. Going from $70 CPA on 400 conversions to $55 on 180 traded 220 customers for a prettier dashboard. A rise from $70 to $85 while volume doubles is usually the right trade, as long as $85 sits under your ceiling. Marginal CPA on the next dollar is always worse than average CPA on the last one, and scaling means accepting that on purpose.

Target CPA bidding, and how to not break it

Target CPA bidding tells Google or Meta to buy as many conversions as it can at roughly the average cost you specify, adjusting bids per auction based on predicted conversion probability. It works well under two conditions: enough conversion volume to learn from (a rough floor of 30 conversions in 30 days per campaign) and clean tracking.

The classic mistake is setting the target too aggressively. Drop it from $90 to $45 overnight and the algorithm does the only rational thing: it stops bidding on nearly every auction it cannot win cheaply, impressions fall off a cliff, and the campaign re-enters learning with too little data to recover. Move targets in steps of 10 to 15 percent and let each settle for a conversion cycle. Watch for the opposite failure too, where a generous target quietly pays $90 for conversions available at $60.

This re-targeting across campaigns and channels is unglamorous daily work, and it is the part AI PPC software does better than a human checking in weekly. AdBot adjusts target CPA across Google, Meta and TikTok daily against your real allowable number, which is the difference between automated Google Ads management and switching on smart bidding and hoping.

Why your reported CPA is probably wrong

Every CPA you see is an estimate produced by an attribution system with blind spots. iOS privacy changes cut the signal Meta gets from web conversions, so platforms now model conversions statistically as much as they observe them. Windows change the story completely: 7-day-click and 30-day report different CPAs for identical performance. View-through conversions credit impressions nobody clicked, which flatters display and video.

The fixes are boring and they work. Send conversions server side so platforms see events the browser drops. Compare channels only on matched windows. Reconcile reported conversions against real orders monthly and carry the gap as a correction factor. Run occasional holdout tests, since the only unarguable measure of incrementality is turning something off and watching total revenue. If Meta is your problem area, the tactical levers are in our guide on how to lower CPA on Facebook.

Frequently asked questions

What is a good CPA for Google Ads?

A good Google Ads CPA is one below your maximum allowable CPA, which is your customer lifetime gross profit. As a sanity check, US non-brand search leads commonly cost $40 to $150 and ecommerce purchase CPAs land between $30 and $90. Legal and B2B software profitably pay several hundred dollars per conversion.

What is a good CPA for Facebook ads?

Meta CPAs run lower than search because you pay for attention rather than intent: $15 to $80 per lead and $25 to $75 per ecommerce purchase are common US ranges in 2026. Lead quality is usually lower too, so compare Meta on cost per closed customer, not cost per lead.

What is the difference between CPA and CAC?

CPA is the cost of a tracked conversion event, which is often a lead rather than a sale. CAC is the cost of an actual paying customer. Divide your CPA by your lead-to-customer close rate to convert one into the other: $65 leads at a 12 percent close rate means a real CAC of $542.

Is a 3:1 LTV to CAC ratio good?

Yes, 3:1 is the widely used benchmark for a healthy business, leaving enough gross profit after acquisition to cover overhead and still make money. Below 2:1 you are fragile. Above 5:1 you are usually under-spending on a channel that works, and you should raise budgets until the ratio settles closer to 3:1.

How do I lower my CPA without losing volume?

Improve conversion rate before touching bids. A landing page converting at 4 percent instead of 2.5 percent cuts CPA by 37 percent at identical traffic cost. Then cut spend on segments that never clear your allowable CPA, tighten targets in 10 to 15 percent steps, and fix tracking so bidding optimizes toward real conversions instead of noise.

If you would rather not run this arithmetic every week, that is AdBot's job: give it your URL and budget, and it builds, launches and optimizes your Google, Meta and TikTok campaigns toward your target CPA daily, for a flat monthly fee from $297 and no percentage of ad spend.