SaaS PPC and Paid Ads

A $12,000 ACV product and a $120,000 ACV product keep turning up in the same media plan. Both buy category keywords on Google, both test LinkedIn job-title targeting, and both report cost per lead to the board every month. Eight months in, the cheaper product works out that its $340 cost per lead looked reasonable on a slide and was ruinous once you divide by a 4 percent lead-to-customer rate. The ad copy was never the problem.

Paid acquisition against a subscription is a different arithmetic problem from paid acquisition against a purchase. You’re buying a revenue stream with a churn rate attached, so the ceiling on any bid comes from gross-margin-adjusted lifetime value and from how long your CFO will wait to see the cash back. Every recommendation below sits on top of that single idea.

Which paid channel fits your ACV

Below roughly $10,000 ACV, high-intent search and review marketplaces beat everything else on cost per opportunity. Above $25,000, LinkedIn and account-based targeting start to justify their CPMs, because you’re paying for reach into a buying committee that a search engine cannot address directly.

Minimum viable spend in that table has almost nothing to do with platform minimums. Google will happily take $30 a day from you. You just won’t learn anything, because Target CPA bidding wants around 30 conversions in a rolling 30 days before it stops guessing, and a $400 cost per conversion means one campaign needs about $12,000 a month to feed the algorithm that signal. Spend under the learning threshold and you are renting variance.

Here’s the pattern I see most often. A Series A company with a $9,000 ACV launches on category terms like project management software at $45 a click, because that’s where the volume sits, and skips brand and competitor terms because the volume there looks trivial. Six weeks later the category campaign has spent $40,000 and produced four opportunities. The brand and competitor campaigns nobody built would have produced eleven for under $3,000.

Channel choice and budget split are different decisions, and people conflate them constantly. If you already know which channels you’re running and need the percentage split instead, SaaS PPC budget allocation works through the same problem from the money side.

What a subscription does to your maximum bid

You cannot set a bid from a cost per lead you read in a benchmark report. Set it from allowable CAC, which comes from gross-margin lifetime value divided by a payback period your finance team will sign off on, then work backwards through your own conversion rates.

Run that for three different businesses and the reason one media plan cannot serve all of them becomes obvious.

That final row settles most channel arguments before they start. The SMB column cannot buy a $45 category click at any volume and stay solvent, which is exactly why review marketplaces and long-tail problem terms matter so much at the bottom of the ACV range. The enterprise column can pay $52 for a security keyword click without flinching, and should.

The per-channel version of this calculation, including how to handle blended CAC when three channels touch the same deal, sits in the SaaS PPC bid math playbook.

Google Search is the only channel that starts with demand already formed

Search is the one paid channel where the buyer has described their own problem before you pay anything. That makes it the highest-converting and most expensive place to be, and it’s where roughly half to two thirds of a typical B2B SaaS paid budget ends up.

SaaS search demand splits into four tiers, and the economics of each are nothing alike. Brand terms convert at 20 to 40 percent and cost $2 to $8. Competitor terms convert at 5 to 12 percent and cost $12 to $60. Category terms convert at 1 to 4 percent and are where CRM, project management, HR and security keywords routinely clear $30 to $80 a click in 2026. Problem and job-to-be-done terms convert under 1 percent and belong to content, not to paid, unless you have a genuinely self-serve product.

Two positions I’ll defend. Buy your own brand even when you rank first organically, because a competitor bidding on your name costs you deals you’d otherwise close, and the defensive spend is usually under 3 percent of your budget. Buy competitor terms carefully: Google permits bidding on a rival’s trademarked name in most markets, but using that trademark in your ad text will get the ad disapproved once the owner files a complaint, so keep the competitor name in the keyword and out of the headline.

The full campaign-level treatment, including which campaign types are still worth running, is in Google Ads for SaaS. If you need headlines that survive a legal review, the SaaS ad copy templates cover the competitor and category patterns.

LinkedIn earns its CPM above about $25,000 ACV

LinkedIn’s floor is $10 a day per campaign, which tells you nothing useful about running it. Realistic CPMs sit between $30 and $90, clicks land between $8 and $16, and you need four to six weeks plus a meaningful audience before frequency and conversion data mean anything.

The targeting is the product. Nowhere else can you buy VP of Engineering at companies with 500 to 5,000 employees in financial services with that accuracy, and for a $120,000 ACV security product that precision is worth every dollar of the premium. For a $6,000 ACV scheduling tool it is not, because you’re paying enterprise CPMs to reach people who would have found you through a comparison page for a tenth of the cost.

Lead gen forms are the trap. They convert two to four times better than a landing page because the form is pre-filled, and the leads are measurably worse because pre-filled means unconsidered. Teams that switch to lead gen forms watch cost per lead halve and cost per opportunity stay flat or rise. If you run them, gate the follow-up behind a qualification step and report on opportunities, never on form completions. The channel-specific detail lives in LinkedIn Ads for SaaS.

Review marketplaces are the most underpriced line in a SaaS media plan

G2 and Capterra sell intent that is later-stage than almost anything on Google, because a person browsing a category directory has already decided to buy something in that category. Capterra starts at a $2 minimum cost per click with a $500 minimum monthly budget, which makes it the only serious paid channel a seed-stage company can test properly for the price of a decent dinner.

Costs climb quickly with category competition. Two to $15 a click is normal, above $20 is standard in crowded categories like CRM, help desk and project management, and G2’s Marketing Solutions contracts generally start around $25,000 a year for the profile features, category placement and buyer intent data bundled together. Pay-per-lead programmes on both platforms tend to run $30 to well over $100 per lead depending on category and filters.

Three honest problems with the channel. Volume is capped by the directory’s own traffic, so you cannot scale it past a point no matter how much budget you have. Your competitors appear on the same page by design, which means you’re paying for a click into a comparison you don’t control. And pay-per-lead leads are frequently sold to several vendors at once, so your sales team will hear the same objection from every one of them.

Review volume gates the whole thing. If you have 11 reviews and the category leader has 800, paid placement buys you traffic into a page where you look like the risky option. Get to 50 reviews first, then buy placement.

Meta and Reddit buy cheap attention that needs the right offer

Both are problem-aware channels, not solution-aware ones, and treating them like search is how budgets die. Nobody on Reddit was looking for your demo form.

Reddit works when you target 15 to 30 specific subreddits rather than interest categories, write the ad in the register of the forum, and send traffic to something readable rather than a trial wall. The minimum daily budget is $5, clicks often land between $1 and $4, and I’ve seen developer-tool companies run sustainable pipeline out of it at $200 to $300 per opportunity. Comments are public and moderators are unforgiving, so budget the time to reply.

Meta earns its place in two situations: retargeting site visitors and trial abandoners, and lookalike audiences built from closed-won customers rather than from all leads. For a self-serve product under $3,000 ACV it can carry real volume. For a $60,000 ACV enterprise sale it is a retargeting line item and nothing more, which is fine, because retargeting is where most of its value sat all along.

The measurement stack that keeps bids honest

Every bid recommendation above collapses if the platform is optimising to the wrong event. The fix is mechanical: capture the click ID, store it on the lead record, and send the revenue outcome back.

Self-reported attribution deserves more respect than it gets. Platform data will tell you a deal came from paid search because that was the last click. The buyer will tell you they heard about you on a podcast in March and searched your brand name in August. Both are true, and only one of them tells you where to put next quarter’s money. Run both, report the gap openly, and do not pretend the model is cleaner than it is. The mechanics of stitching this together are in paid media attribution for SaaS, and the tooling options are compared in PPC tools for SaaS teams.

What to switch on, in what order, by stage

Sequencing matters more than selection. Every stage below assumes the previous stage is saturated, meaning impression share on those terms is above 80 percent and cost per opportunity is still inside your ceiling.

Note what that table refuses to do. It does not tell a $2M ARR company to run LinkedIn, no matter how good the targeting story is, because at that spend level the budget buys more opportunities elsewhere. The stage-by-stage reasoning, including what to cut first when budget tightens, is in the SaaS advertising strategy playbook.

Where paid advertising stops working, and what it costs to find out

Paid channels saturate, and most SaaS teams notice about two quarters late. The signals are consistent: brand and competitor impression share above 80 percent with no headroom left, category CPCs rising faster than your conversion rate improves, and a cost per opportunity drifting toward allowable CAC while pipeline volume stays flat.

Costs have moved against everybody. Acquisition cost in B2B SaaS is widely reported to have risen 40 to 60 percent since 2023, driven partly by real competition and partly by attribution loss, because teams that can no longer see which clicks produced revenue bid conservatively across the board and pay more for less. That second cause is fixable with the measurement stack above. The first one is not.

The honest tradeoff nobody writes down: paid media stops the moment you stop paying. A content programme that took nine months to work keeps working for two years after you pause it. Paid gives you a readable answer in three weeks and owns nothing. Teams that build only on paid hit a ceiling where every incremental dollar buys a worse customer, and the only way out is a channel with compounding returns.

The specific ways money disappears, from broad match with no negatives to counting trial signups as conversions, are catalogued in SaaS PPC mistakes that waste budget. If you want to know whether your numbers are normal before you act on any of this, the SaaS PPC benchmarks dataset is segmented by ACV band and motion rather than averaged into a single figure.

Questions SaaS teams ask about paid ads

Does paid advertising still work for SaaS in 2026?

Yes, on high-intent surfaces, and much worse on broad ones. Search and review marketplaces still produce buyers because intent is stated rather than inferred. Broad awareness buys have degraded as attribution signal weakened and costs rose, so they now need account-level measurement to justify. If your ACV is under $5,000 and you have no self-serve motion, paid will struggle to clear payback at any spend level.

How long does SaaS PPC take to show results?

Two to three weeks for search to produce enough clicks to judge targeting, four to six weeks for LinkedIn, and a full sales cycle plus 30 days before you can honestly report cost per closed-won. A company with a 90-day sales cycle should not evaluate a channel on revenue before month five. Judge on cost per qualified opportunity in the meantime.

Should a SaaS company run ads before content and SEO?

Run brand search from day one, then use paid to test messaging before you commit a year of content budget to it. Paid answers in three weeks what SEO takes nine months to answer: which problem framing people actually click. Treating a $4,000 search test as market research is usually the highest-return use of early paid budget.

How much of a SaaS marketing budget should go to paid?

Most B2B SaaS companies between $1M and $20M ARR put 25 to 45 percent of total marketing spend into paid media, with the higher end at companies running a sales-led motion into a category with formed demand. Below 20 percent you usually cannot maintain a readable signal in any channel. Above 60 percent and you’re renting a pipeline you don’t own.

What to do this week

Pull your last 90 days of paid spend and split it by intent tier, not by campaign name. Most accounts discover that 60 percent of spend sits on category terms producing 20 percent of opportunities, and that the fix is a budget reallocation rather than a new channel.

Then do the LTV calculation with your real churn number and write the maximum CPC on a sticky note. If any campaign is bidding above it, pause that campaign today and argue about it tomorrow. Once those two things are done, connect offline conversion import before you touch a single ad, because every optimisation you make before that is being judged by the wrong scoreboard.