In WordStream and LocaliQ’s 2026 search ad benchmarks, the average cost per lead across Google and Microsoft Ads was $66.69. Cost per lead by industry ranged from $26.84 for arts and entertainment to $131.63 for attorneys and legal services.
Facebook lead campaigns ran lower, and Google Local Services Ads for home services contractors averaged $53 per lead in February 2026.
What cost per lead measures and how it differs from cost per customer
Cost per lead (CPL) is your ad spend divided by the number of leads that spend produced. If a campaign costs $1,000 and brings in 20 form fills and calls, your cost per lead is $50.
What counts as a lead changes the number a great deal. One account counts a form submission. Another counts any phone call longer than a few seconds. A third counts a message sent through a platform. WordStream makes the same point in its 2025 Google Ads benchmarks: your cost per lead depends on how you track and assign value to conversion actions, and lead quality belongs in the evaluation too.
A lead is also not a customer. 99 Calls, writing about home-service leads in 2026, separates the price you pay per lead from the cost of a closed job. The closed-job cost depends on lead quality, your sales process and whether someone answers the phone. A pricier lead that books often can cost less per job than a cheap lead that rarely books.
Search ad cost per lead by industry
The WordStream/LocaliQ 2026 Google Ads benchmarks, last updated September 16, 2026, cover Google and Microsoft search ads. The overall average cost per lead was $66.69, down from $70.11 in the 2025 edition of the report. Clique Studios, summarizing the same LocaliQ and WordStream data, notes that the search average fell for the first time in five years.

| Industry or segment | Average search ad cost per lead | Source and period |
|---|---|---|
| All industries | $66.69 | WordStream/LocaliQ, 2026 report |
| All industries (prior year) | $70.11 | WordStream/LocaliQ, 2025 report |
| Arts and entertainment | $26.84 | WordStream/LocaliQ, 2026 report |
| Automotive repair, service and parts | $29.96 | WordStream/LocaliQ, 2026 report |
| Restaurants and food | $30.57 | WordStream/LocaliQ, 2026 report |
| Real estate | $102.51 | WordStream/LocaliQ, 2026 report |
| Furniture | $106.70 | WordStream/LocaliQ, 2026 report |
| Attorneys and legal services | $131.63 | WordStream/LocaliQ, 2026 report |
| Home services (category average) | $90.92 | LocaliQ, 3,211 home services search campaigns, April 2024 to March 2025 |
The home services row comes from a separate LocaliQ dataset with its own sample and measurement window, as reported by Now Booked Marketing. Read it as its own benchmark, not as a line from the 2026 table. If you run a trade business, our guide to advertising for home service businesses covers channel choices by trade. Law firms, insurance agents and other high-ticket offices sit at the top of this table, and our professional services advertising pages explain how those buyers plan around expensive leads.
High lead costs are not automatically bad. A signed legal case can be worth far more than a restaurant visit, so a $131.63 lead can pay for itself in law while a restaurant needs leads closer to $30.
Facebook and Local Services Ads cost per lead
Facebook lead campaigns cost a median $27.39 per lead in WordStream’s 2026 Facebook ads benchmarks, as reported by Clique Studios. WordStream publishes that data by industry on its conversion rate and cost per lead benchmarks page. This figure measures the leads objective: people who tap through a lead form, often with their contact details filled in automatically. They were not searching for your service when they saw the ad, which is part of why the leads are cheaper and why many need more follow-up before they book. If you are weighing Meta placements, see our guide to Instagram ads for local businesses.

Google Local Services Ads charge per lead rather than per click, and SearchLight Digital’s benchmark, summarized by Ryn Digital, puts the blended cost at $53 per lead in February 2026, based on $6.72 million in spend across 888 home services contractors and 126,650 leads, measured before lead credits. Electrical was the lowest trade at $39 per lead, the average book rate was 43.9%, and the average cost per paying customer was $233. Plumbers can compare these figures with our page on lead generation for plumbers.
Local Services Ads are changing as well. According to Nimbata’s 2026 Local Services Ads guide, Google began moving them into Google Ads as a specialized Performance Max campaign type with pay-per-lead goals in August 2026. Starting October 1, 2026, a call that stays connected for more than 20 seconds during your stated business hours can become a billable lead even if nobody answers. February benchmarks were measured before both changes.
We’d avoid ranking these channels against search CPL as if they measured one outcome. A lead-form submission, a search click that turns into a call, and a pay-per-lead phone call answer different questions and use different counting rules.
How to turn a benchmark into a lead budget
- Set your maximum cost per lead. Multiply what a new customer is worth to you (profit, not revenue) by your close rate. The result is the most you can pay per lead before you lose money.
- Use the benchmark as a starting range. Compare your maximum to the industry figures above. If your ceiling sits well below your industry’s average, fix the offer, the landing page or the close rate before you raise spend.
- Estimate lead volume from your daily budget. Divide daily spend by your target CPL to get a rough number of leads per day once the campaign settles, and plan for the first weeks to run differently.
- Track cost per booked job, not only cost per lead. Count how many leads become appointments and paying customers.
For example, take the roofing figure Now Booked Marketing cites: $228.15 per Google Search lead at a 15% booking rate. Divide $228.15 by 0.15 and each booked appointment costs about $1,521. A roofer who knows the average job value can decide whether that works. Our page on advertising for roofing companies covers how roofers plan around that math, and the local advertising cost guide compares what different local channels cost to run.

If you test this in Adwave, daily budgets start at $30 and cover media, ads and management, with no subscription. TV, Google (Search included), Facebook and Instagram can run from that minimum. Reddit runs from about $35 a day, and TikTok is opt-in from $91 a day. You install a tracking snippet and choose which actions count as conversions, so the cost per lead you see reflects your own definition. Your actual cost per lead will depend on your market, offer and follow-up, and no benchmark predicts it.
What these benchmarks leave out
Each figure above averages many advertisers, periods and lead definitions, and your account will count leads its own way, so treat the numbers as a range to check against, not a price you’ll be charged.
Response time changes the real cost more than any benchmark captures. Now Booked Marketing cites Blue Grid Media, which reports from the Local Services Ads accounts it manages: contractors who answer within 1 hour book about 31% of answered calls, while those taking 24 hours or more book about 15%. Slow follow-up roughly doubles the cost of each booked job regardless of which channel the lead came from.
Shared leads and exclusive leads aren’t the same product. 99 Calls points out that exclusive leads cost more upfront but close at a higher rate, since you aren’t competing with every other contractor who bought the same lead.
A city with many contractors bidding on the same searches will run above a national average, and none of the benchmarks here break results out by individual market.
The benchmarks on this page don’t include a cost-per-lead figure for TV or CTV. For costs on that channel, see what TV advertising costs.
These figures also describe local service businesses, not B2B lead generation, which runs on different sales cycles and lead values; see HubSpot’s 2025 CPL and CAC benchmarks if that’s your category.




