ROAS Formula and Benchmarks for B2B LinkedIn Campaigns

The standard ROAS formula divides revenue attributed to ads by ad spend, and it works fine when the gap between a click and a purchase is short. That formula assumes the revenue from a campaign is captured inside the window someone is measuring. In B2B, that assumption falls apart, because the click and the closed deal can sit nearly a year apart.

ROI tries to fix part of this by factoring in the cost of closing a sale and the cost of running the campaign itself, expressed as net profit minus total marketing spend, divided by total marketing spend, which produces a cleaner number than ROAS. But it still needs a long enough time window to produce anything other than zero in a B2B sales cycle.

The data backs up just how long that window has to be. Dreamdata's 2026 benchmarks put the median time from first LinkedIn ad impression to closed revenue at 281 days. Oddly, the time from a first ad conversion (a form fill) to revenue and the time from a first ad engagement to revenue land almost on top of each other. That's a real problem for anyone who assumes a form fill signals stronger buying intent than a simple engagement with an ad. The data says otherwise.

The lengthening timeline is concentrated entirely before the deal reaches sales. Once a lead becomes a sales-qualified lead (SQL), the time to closed-won has actually gotten shorter, and time from SQL to closed-won and MQL-to-SQL have both compressed. Buyers are showing up to sales conversations later in their own research process, then moving through the pipeline faster once they arrive.

Marketing teams facing a quarterly reporting deadline will push back on all of this, and fairly so. Waiting 281 days for a number isn't practical when a budget review happens every three months. The fix is a different metric, not a shorter timeline. Switching to pipeline ROAS, measured at a meaningful point mid-cycle rather than waiting for the deal to close, gives a number that's non-zero and useful for decisions well before revenue actually lands.

The correct B2B ROAS formula

Pipeline ROAS replaces "revenue closed" with "revenue on track to close," and that single swap is what makes the number usable inside a normal reporting cycle. The formula: multiply target opportunities by average contract value (ACV), multiply that by close rate, then divide by monthly budget.

Each piece of that formula forces a real conversation between marketing and sales. Target opportunities" needs CRM-connected attribution. ACV and close rate come straight from sales data, not marketing's guess at either. Building the formula, in other words, requires marketing and sales to agree on what an opportunity actually is before anyone can calculate anything.

Say a company spends $20,000 on LinkedIn in a month and expects target opportunities to result from that spend, at a $30,000 ACV and a 25% close rate. That's a number a budget committee can act on in the same quarter the spend happened, instead of waiting to see whether any of those opportunities actually close.

Budget-setting has its own companion formula. Target cost per SQL equals ACV multiplied by a small percentage, lower for SMB deals and higher for enterprise ones. Target cost per opportunity equals cost per SQL divided by the SQL-to-opportunity rate, and that rate typically falls somewhere between a quarter and two-fifths across the industry. Set those two numbers correctly, and a monthly LinkedIn budget stops being a guess and starts being a target with math behind it.

Most companies aren't doing any of this. Only a minority of B2B SaaS companies have pipeline attribution that actually connects LinkedIn spend to CRM revenue. The rest fall back on cost per lead (CPL), and that default choice trains the ad platform to optimize toward the wrong people entirely.

The trap works like this. When CPL is the metric an account is optimized against, the platform learns to find people most likely to fill out forms: students, job seekers, competitors, and contacts below ICP seniority. Reported CPL drops, which looks like a win. SQL volume stays flat and meeting acceptance rates fall, because the leads coming in were never going to buy anything. A higher CPL paired with a high MQL-to-SQL conversion rate beats a lower CPL paired with a low one every time, because pipeline generated is the only number in this whole exercise that connects ad spend to what the business actually needs.

What LinkedIn's ROAS numbers look like in 2026

LinkedIn is the only major B2B advertising platform posting a positive aggregate ROAS in 2026. Dreamdata's 2026 benchmarks report shows LinkedIn ROAS climbing from 2024 into 2025, with companies averaging more than two dollars in attributed revenue for every dollar spent. Put that next to the same report's findings on Google Search and Meta, both of which came in below breakeven, using a data-driven attribution model applied to closed-won deals over a 12-month window.

The gap between LinkedIn and its nearest competitors widens further among well-structured programs. LinkedIn's share of influenced new business deals grew from 2024 into 2025, and its influence gets stronger, not weaker, as a deal moves closer to close, showing up in a growing share of sessions all the way from MQL through SQL, and staying elevated at the new-business stage. That pattern cuts against a common assumption about paid social, that it's a top-of-funnel tool with fading relevance the closer a deal gets to signing. LinkedIn's data says the opposite.

Budget allocation is following the same signal. LinkedIn now takes the largest share of total B2B paid media budgets among Dreamdata's customer base, up from the year before, while non-branded Google Search has lost share as AI-generated answers inside search results cut click-through rates and push cost per click higher. Companies that use LinkedIn's Revenue Attribution Report, which connects ad engagement directly to CRM pipeline and revenue data, tend to put more budget behind LinkedIn as a result. That's a telling detail: the ability to measure a channel accurately is driving budget toward it, independent of whatever the raw performance number would say on its own.

The upper end of what's possible sits well above the platform average. Top-quartile conquesting programs, the ones targeting competitors' audiences directly, post 180-day ROAS figures that reach double-digit multiples in the best cases. The platform-wide average of "more than two dollars back for every dollar spent" is a floor for a well-run program, not a ceiling.

Why LinkedIn delivers that premium

LinkedIn's ROAS advantage doesn't come free. Programs earn it by matching targeting precision, funnel architecture, and format selection to the platform's mechanics, and programs that collapse these into a single step give back the premium.

Start with targeting. Narrow, specific job-title targeting pushes CPMs and CPCs higher, but it often improves ROAS for B2B SaaS anyway, because the audience reached is closer to who actually buys. That tradeoff plays out differently by industry: Sponsored Content CPMs vary widely across the platform in 2026, with enterprise conquesting audiences (director-level and above, at companies with 500 or more employees) at the top of that range. Paying more to reach fewer, better-fit people is a deliberate tradeoff.

Cost per company influenced, the cost of reaching a company where at least one stakeholder had contact with an ad, is a better unit of measurement than cost per click. That figure dropped materially on LinkedIn from 2024 into 2025. It's the right way to think about cost, because B2B deals close at the company level. A single contact clicking an ad was never the unit that mattered.

Funnel structure matters just as much as targeting. The most successful LinkedIn advertisers build distinct stages, each with its own campaign objective, its own creative, and its own conversion goal, rather than running one campaign meant to do everything. Most programs skip this. They collapse a three-stage sequence, awareness, consideration, conversion, into a single step: put a cold ICP audience in front of a demo-request ad and hope for the best. Gartner's 2026 research shows most B2B buyers do their own research before ever contacting a vendor. Asking a cold audience to request a demo means asking people who haven't even settled on what problem they're solving to commit to a sales conversation, a mismatch between where the buyer is and what the ad is asking for that shows up directly in wasted spend.

Format choice is the third lever, and it works best as a mix rather than a single choice. LinkedIn's Lead Gen Forms, which let someone submit their info without leaving LinkedIn, produce a competitive median CPL across B2B advertisers, well below the CPL for driving traffic to a landing page instead. Lead Gen Forms win on conversion rate. Landing pages often win on lead qualification for complex sales cycles, because the person had to actually leave LinkedIn and engage with a website to convert. Running both, rather than picking one, captures volume through forms and captures higher-intent signal, plus the retargeting and analytics that come with an actual site visit, through landing pages.

Creative has to match the environment it runs in. LinkedIn's platform-wide median CTR for Sponsored Content is 0.44% in 2026, well below Facebook's, because people move through a professional feed faster and with narrower attention than they do on a personal one. Creative built to stop someone scrolling through vacation photos won't necessarily stop someone scrolling through their work feed. Whatever earns attention on LinkedIn has to be built for LinkedIn specifically, not repurposed from a Meta campaign.

None of this happens by accident. When an account is optimized to CPL, the algorithm does what it's told and finds more people who'll fill out a form, whether or not they'll ever become a customer. The mechanism runs the same way every time; the outcome depends entirely on what target gets fed into it.

The benchmark table B2B programs should be measured against

Healthy LinkedIn programs get judged on cost per SQL, pipeline ROAS, and CAC payback, not CTR, CPC, or CPL. The numbers that matter shift depending on ACV tier and how far out the measurement window runs.

Program-level health starts with three targets: a 3:1 ratio of lifetime value to customer acquisition cost, a CAC payback period under 12 months, and a 180-day pipeline ROAS between 4x and 8x. Those three numbers form the floor any program should clear before anyone spends time optimizing individual campaigns.

Below that sit the top-quartile thresholds, drawn from SaaS Hero's benchmarks, which also pull from GrowthSpree and Dreamdata data.

• Click-through rate on Sponsored Content: top-quartile programs run above the normal range. CTR above the normal range lowers effective CPM cost but does not predict SQL quality.
• Cost per click on Sponsored Content: top-quartile programs run below the normal range. CPC alone doesn't say anything about pipeline quality either, and it should always get checked against cost per SQL before anyone calls it a win.
• Cost per SQL: top-quartile programs run materially below the normal range, and this is the number to actually optimize toward, always checked against ACV tier.
• 180-day pipeline ROAS: top-quartile programs run materially above the normal range.

Cost per SQL also shifts by ACV tier, and comparing across tiers without adjusting for this is where a lot of benchmarking goes wrong. At the SMB tier, median cost per SQL runs higher than most channels can justify at that deal size. Mid-market cost per SQL climbs along with deal complexity. Enterprise cost per SQL climbs further still, but larger deal sizes justify the higher spend. Strategic accounts, the highest ACV tier, carry the highest median cost per SQL of all, offset by deal values large enough to make the math work anyway.

Rising costs don't automatically mean a channel is getting worse. HockeyStack data tracking more than 70 B2B SaaS companies with substantial LinkedIn spend found CPC climbed materially within a single year, while pipeline ROI held steady. Treat rising LinkedIn costs as a baseline planning assumption each year, the same way a marketer would plan for rent going up. The channel stays viable as long as ACV justifies whatever the cost per SQL turns out to be.

Conquesting campaigns, the ones aimed at competitors' audiences, cost more in CPM than branded campaigns but deliver roughly double the conversion rate at every stage of the funnel. Median CPL-to-SQL conversion for conquesting in B2B SaaS runs in a band of roughly a quarter to a third. Vertical economics matter too: cybersecurity campaigns support higher CPLs because high ACVs still deliver strong ROAS. "Good" performance is always relative to the ACV tier and industry a program sits in. There's no single number that applies across the board.

The attribution architecture that makes these benchmarks computable

None of the benchmarks above mean anything without attribution built to actually track them. CAC payback needs revenue data tied back to the specific campaign that touched the account, tracked over the months it takes that account to move through the pipeline.

This is the same gap the earlier CPL trap exposed: only a minority of B2B SaaS companies have that kind of pipeline attribution in place. Without it, a program can only report CPL and CTR. Building attribution that connects ad engagement to CRM data is the infrastructure the strategy depends on. It's the infrastructure the whole pipeline ROAS formula depends on, and without it, every benchmark in the table above stays out of reach.