B2B SaaS cost-per-lead is climbing, and blaming your bids is the wrong diagnosis. The real story in 2026 is a squeeze from two directions at once: average SaaS CPCs are up 15-29% year-over-year, and AI Overviews now sit at the top of the results page, pushing paid ads down and cutting the click-through rate on the same ad. When you pay more per click and earn fewer clicks per impression, the cost to produce each lead rises faster than the CPC number alone implies. This post lays out the mechanism, the data behind it, and the response that actually protects your CAC.
The reason this is worth a dedicated diagnosis rather than another “CPCs are up” post is that the two forces interact. As eMarketer reports, Google Ads cost per lead is going up while conversion rates fall — so both halves of the CPL equation are moving against you. Understanding why lets you cut the spend that AI Overviews have quietly devalued and protect the spend that still converts, instead of raising targets across the board and hoping.
The mechanism: two forces compounding
Cost per lead is a product of two things — what you pay per click and how many clicks it takes to get a lead. In a normal inflation year, only the first rises: more advertisers bid on the same keywords, CPC creeps up, CPL follows proportionally. 2026 is different because the second factor is deteriorating at the same time. AI Overviews answer the query inline, so fewer users scroll to the ads, which lowers click-through rate and, downstream, the volume and quality of leads per impression. The two effects multiply rather than add.
Concretely: if CPC rises 20% and the effective conversion path lengthens so it takes more clicks per lead, a 20% CPC increase can show up as a 30-40% CPL increase. That gap between the CPC line and the CPL line is the signature of the squeeze, and it is why teams watching only CPC are surprised when their lead costs jump more than their click costs. This compounding is the same dynamic driving the broader CAC pressure we cover in why B2B SaaS CAC keeps rising — the AI Overviews squeeze is one of the newest and least-understood inputs to it.
What AI Overviews do to paid click-through rate
AI Overviews reduce paid click-through rate by answering the query before the user reaches the ads. When Google generates an inline answer at the top of the page, a meaningful share of searchers get what they came for and never scroll — so the same ad, in the same position, earns fewer clicks than it did a year ago. Industry CTR studies through 2026 show the effect is real and concentrated on informational queries, where the AI can fully satisfy the intent, and much milder on commercial queries, where the searcher still needs to visit a site to compare, price, or buy.
For B2B SaaS this asymmetry is the whole game. Top-of-funnel research terms — “what is X,” “how does Y work” — are exactly the queries AI Overviews answer best, so the cheap informational traffic that used to seed the funnel is drying up or getting more expensive per surviving click. Bottom-of-funnel terms — “X pricing,” “X vs Y,” “X free trial” — still push buyers past the AI answer to click through. We go deeper on the CTR data in how AI Overviews are reshaping SaaS paid CTR, and on the stranger cases where the AI answer and your ad disagree in when AI Overviews contradict paid ads.
Why the surviving clicks convert worse
The clicks that make it through the AI Overview are not just fewer — on average they convert at a lower rate, which is the second and more painful half of the squeeze. Part of the reason is selection: AI Overviews skim off the highest-intent informational searchers who found their answer, and the remaining clicks skew toward users who were not fully served, a slightly noisier pool. Part is structural to B2B — longer, more considered buying journeys mean more clicks and touches per eventual conversion, so any given click is less likely to be the one that converts.
The practical consequence is that you cannot fix rising CPL by conversion-rate optimisation on the landing page alone, because a chunk of the decline is coming from the composition of the traffic, not the page. This is where measurement discipline matters more than ever: if you are counting raw form fills as leads, a falling conversion rate looks like a landing-page problem when it is really a traffic-quality problem. Optimising to genuine pipeline instead of raw leads — the argument in cost per lead vs cost per SQL — keeps you from over-investing in fixing the wrong thing.
The response: move budget down the funnel
The single highest-leverage response is to shift budget toward the bottom of the funnel, where AI Overviews have the least effect. High-intent commercial queries still send buyers past the AI answer to click and convert, so their clicks retain their value even as informational CTR erodes. Concentrating spend there means you are buying the clicks that still perform and stepping back from the ones AI has devalued — a reallocation, not a cut. For most SaaS accounts this looks like protecting pricing, comparison, alternative, and trial-intent keywords while trimming aggressive bids on broad research terms.
The corollary is to stop treating paid search as your top-of-funnel demand engine. The cheap informational clicks that used to seed the funnel are the exact traffic AI Overviews absorb, so that demand-generation work has to move to channels AI does not intercept — organic content, community, and paid social. We lay out a fuller version of this reallocation in our response playbook for rising CPLs. The principle is simple: let paid search do what it is now best at, capturing bottom-funnel intent, and stop paying premium CPCs to compete with a free AI answer at the top of the funnel.
The response: tighten measurement so bids follow revenue
A squeeze on click economics makes measurement the difference between defensible and reckless spending. If Smart Bidding is optimising to raw lead volume, it will keep chasing the cheap top-of-funnel clicks that AI Overviews have quietly made less valuable, because it cannot see that those leads convert worse downstream. Feeding offline conversion data — MQL, SQL, closed-won — back into Google Ads lets the algorithm bid to real pipeline, so it naturally reallocates toward the bottom-funnel clicks that survive the AI answer and convert.
This is not a new-tooling problem; it is a discipline problem. The offline-import pipeline that makes revenue-based bidding possible is the same one every serious SaaS account should already run, and it is more valuable now than it was a year ago precisely because the traffic mix has shifted. If you have not built it, our offline conversion stack guide is the place to start, and bidding strategies for B2B SaaS covers how to point Smart Bidding at pipeline once the data is flowing. Buy the clicks that convert; stop subsidising the ones AI Overviews devalued.
The bottom line for 2026 planning
Rising CPLs in 2026 are not a bidding failure and not a reason to abandon Google Ads — they are the visible symptom of a channel narrowing to what it does best. AI Overviews are compressing the top of the funnel out of paid search and concentrating value in bottom-funnel, purchase-ready queries. The accounts that struggle are the ones still paying premium prices to compete with a free AI answer on informational terms; the accounts that thrive are the ones that moved budget down the funnel and pointed Smart Bidding at pipeline.
Plan next year’s paid search budget as a bottom-funnel intent-capture channel, not a demand-generation one, and move the demand-generation dollars to channels AI does not intercept. Judge the channel on cost per SQL and pipeline, not cost per lead, so the falling conversion rate on devalued traffic does not distort your read. Done right, the squeeze is an argument for sharper focus rather than retreat: fewer, better clicks, measured against revenue, on the queries where a human still needs to click to buy.