Cost per lead fell for the first time in five years
WordStream by LocaliQ’s 2026 search advertising benchmarks put the average cost per lead at $66.69, decreasing overall for the first time in five years 1. The dataset behind the number is the 10th annual edition, built on 13,474 US search advertising campaigns running from April 1, 2025 through March 31, 2026, with all-industries averages of $5.42 per click, a 6.64% click-through rate, and an 8.18% conversion rate 1. A companion cut of the same data runs at WordStream 2.
Here is the position before the analysis: benchmarks are diagnostic ranges, not targets. The only CPL target that matters comes from your unit economics. Use the industry tables to find order-of-magnitude anomalies, then stop looking at them.
The spread across industries is wider than the averages suggest
The industry range runs roughly six to one. Attorneys & Legal pays the highest average CPC at $9.87, followed by Home Improvement at $8.33 and Dentists & Dental Services at $8.00, while the cheapest clicks sit in Arts & Entertainment at $1.63, Restaurants & Food at $2.05, and Travel at $2.14 1.
That spread is the first reason the all-industries average makes a poor target. $5.42 describes no one. A personal injury firm reading it would panic; a restaurant group reading it would relax. Both would be reacting to a number that carries no information about their market. The only row in the table that means anything to you is your own, and even that row blends business models, geographies, and intent mixes that may not resemble yours.
Why CPL fell, as far as the data supports
Two things in the data line up with a falling CPL, and the discipline here is to go no further than they support. Tinuiti’s Q1 2026 benchmark report found Google search ad spend up 14% year over year, the highest growth in nearly two years, with flat average CPC and brand keyword CPCs down 9% year over year 3. Set that flat-CPC environment next to the 8.18% average conversion rate in the LocaliQ dataset 1, and the arithmetic explains itself: when clicks cost about the same and a higher share of them convert, cost per lead falls.
Anything past that arithmetic would be speculation. The temptation is to write a story about smarter bidding or better creative across the market, and no figure in either dataset supports one. Speculation dressed up as benchmark analysis is exactly how bad CPL targets get set.
What a benchmark is for
Used correctly, Google Ads benchmarks 2026 by industry answer three questions and no more.
First, sanity-checking a new account or channel. If you launch search for a dental group and the first month prints $23 clicks against an industry average of $8.00, something structural is wrong: match types, geography, or landing-page relevance, in roughly that order. The benchmark cannot diagnose the problem, but it can tell you a problem exists before three more months of budget confirms it.
Second, spotting order-of-magnitude anomalies in a running account. A benchmark cannot tell you whether a $95 CPL is good. It can tell you that a $660 CPL in a $66.69 world deserves a teardown before another dollar goes in.
Third, setting expectations with stakeholders. A CFO who expects $30 leads in a legal category paying $9.87 per click will not be persuaded by campaign screenshots. The industry table is the fastest way to move that conversation from opinion to arithmetic.
What a benchmark is not for
Three misuses show up in almost every account the firm audits.
The first is setting CPL targets from the industry average. Your target is a function of your close rate and your gross margin, not of what other advertisers in your category pay. More on the math below.
The second is judging a campaign as fine because it matches the average. The average includes every neglected account and every mis-set bid strategy that ran during the measurement window. Matching it proves nothing beyond membership in the dataset.
The third is comparing across mismatched intent mixes. An account weighted toward brand terms will beat the benchmark on every metric while doing less work; Tinuiti’s Q1 data shows brand keyword CPCs down 9% while overall CPC held flat 3, so a brand-heavy mix drifts cheaper for reasons that have nothing to do with management quality. Before comparing any account to a table, ask what share of its volume is brand.
Your CPL target is a math problem you already have the inputs for
The formula fits in one line: a lead is worth your lead-to-close rate times your gross profit per closed customer, and your CPL target sits below that with room to breathe. A firm that closes 20% of leads at $1,200 of gross profit can pay up to $240 per lead, so a $150 CPL is healthy there even though it runs at more than double the dataset average. A business closing 3% of leads at $500 of margin loses money at $15 per lead, less than a quarter of that same average. Same benchmark, opposite conclusions, and the benchmark contributed nothing to either one.
That is why the industry table cannot set the target. It contains no information about your close rate or your margin, and those two numbers do all the work. Whether the economics support paid search at all is a fit question, and the firm has already answered it in the FAQ.
Once the target is set from economics, the lever that moves CPL is not in the bidding menu. The firm has made this argument at length: the landing page matters more than the bid strategy, because the platforms have automated the bidding layer and the page is where conversion rate, and therefore cost per lead, gets decided. That is where the hours go in a managed account, and it is why two advertisers in the same benchmark row can sit 2 to 3x apart on CPL with identical bids.
So give the 2026 tables their twenty minutes. Check your industry row, flag anything an order of magnitude off, note that the CPL trend broke downward. Then close the tab. The benchmark tells you what the market pays. It cannot tell you what you can afford, and no industry table will ever contain the two numbers that decide it.