Why Your ROI Numbers Feel Wrong (And Often Are)

Your dashboard says leads are up, spend is down, and cost per click looks great — so why can't anyone tell you if marketing actually made money? If that question stings, you're in good company. Most B2B teams are drowning in metrics while starving for one honest answer.

The uncomfortable truth is that many ROI numbers aren't just unclear — they're structurally broken. When Metadata analyzed $57.6M in B2B ad spend across 153 advertisers, 42 of them — 31.8% — had zero CRM signal. Nothing their ads produced could be traced past the lead stage. For those teams, every ROI number on every dashboard is a guess dressed up as a metric.

The measurement gaps don't stop there. Benchmarkit's B2B benchmarks found that only 52% of marketing teams measure cost per dollar of pipeline, and only about 10% of participants could report marketing expense per dollar of new logo ARR. You cannot benchmark yourself against an industry if you can't benchmark yourself at all.

Even teams with clean data face a quieter distortion: attribution. Research on growth channel ROI shows last-click attribution underestimates content marketing value by roughly 40% on average, and systematically undervalues top-funnel channels like SEO and LinkedIn. As Metadata puts it bluntly: "Last-click attribution is useless in B2B. The last thing someone did before they bought is usually 'talk to sales' or 'sign the contract.'"

Three measurement failures account for most of the confusion:

• Broken feedback loops — spend never connects to CRM revenue, so closed-won deals are invisible to marketing reporting.
• Wrong attribution models — last-click credits the bottom of the funnel and starves the channels building future pipeline.
• Too-short windows — a 90-day ROI view over-credits demand capture and under-credits SEO, content, and nurture, which influence deals closing 6–18 months later.

This is why Directive Consulting's benchmark analysis insists you treat ROI benchmarks as "guardrails, not goals." They exist to help you spot when something is deeply off — not to hand you a target your unique sales cycle, deal size, and margins can't support.

That's the philosophy we take at Worqd: one integrated report tied to pipeline and booked calls, no vanity metrics. Before comparing yourself to any benchmark, fix the plumbing — connect spend to revenue, use multi-touch attribution, and measure each channel against its own time horizon.

Once your numbers are honest, the benchmarks in the next section become genuinely useful.

B2B ROI Benchmarks by Channel: The Real Ranges

Here's what the numbers actually say — and they say very different things depending on which channel you're looking at. B2B marketing ROI isn't one number; it's a set of ranges, each with its own clock.

The most widely cited channel-level benchmarks come from First Page Sage data compiled by Directive Consulting. Paid search sits at roughly 36% ROI, with a working range of 0–100%. Paid social on LinkedIn averages 229%, ranging from 50–250%. Email lands at 261%.

Then the long-cycle channels pull far ahead. SEO averages 748% ROI, with a realistic range of 200–800% or more, and content marketing hits 844% — but measured over three years, not one quarter. The channels with the highest ROI are also the slowest to pay off. SEO typically needs 12–36 months before results show, while paid search and paid social return returns in 1–3 months.

Real-world ad data backs this up with a dose of honesty. Metadata analyzed 153 B2B advertisers spending $57.6M and found 7.97x in sourced pipeline per $1 — but closed-won ROAS often starts below 1x, and only one segment (SMB on LinkedIn) beat 1x on closed-won revenue. Pipeline multiples look impressive; actual revenue takes longer to show up.

The email numbers illustrate how methodology bends everything. The DMA benchmark claims $36–42 back per $1 spent — a 3,500%+ return — but only with a qualified, engaged opt-in list. First Page Sage's 261% reflects a different measurement approach entirely. Same channel, wildly different "benchmark."

A few practical takeaways when you compare your own numbers:

• Judge each channel against its own range and time horizon, not one company-wide ROI target.
• Give SEO and content 12–36 months before calling them failures — short windows systematically defund your future pipeline.
• Expect paid channels to show pipeline fast but closed-won revenue slowly; a clean 4x in month one usually means the measurement is looser than it looks.
• Pair ROI with unit economics — a 300% ROI means little if your CAC is $15,000 and your LTV is $18,000.

One caveat worth repeating: these figures are guardrails, not goals. Last-click attribution underestimates content marketing value by about 40%, and a 90-day measurement window over-credits demand capture while under-crediting SEO, content, and nurture. That's why at Worqd we report against the whole path — first click through booked call — rather than a single vanity multiple. When your benchmarks match your time horizons, the numbers finally start meaning something.

The Measurement Traps That Distort Every Benchmark

Two companies can run the same LinkedIn campaign, spend the same budget, and report ROI numbers that look like they came from different planets. The difference usually isn't performance — it's measurement.

The first trap is attribution model choice. Last-click attribution gives all the credit to whatever happened right before the deal closed, which in B2B is usually "talk to sales." According to growth marketing research, last-click underestimates content marketing value by roughly 40% and systematically undervalues top-funnel channels like SEO and LinkedIn. Switch to a multi-touch model, and suddenly your "underperforming" blog looks like your best investment.

The second trap is the measurement window. A 90-day ROI view over-credits demand capture — the channels that harvest buyers already in market — and under-credits the channels that create demand. As Directive Consulting's benchmark analysis notes, SEO, content, email nurture, and ABM influence deals that close 6 to 18 months later. Measure only what closes this quarter, and you will defund the channels building next year's pipeline.

The third trap is the gap between pipeline and closed-won revenue. Metadata's analysis of 153 B2B advertisers and $57.6M in spend found 7.97x sourced pipeline per dollar — but honest closed-won ROAS started below 1x, with only one segment (SMB on LinkedIn, at 1.37x) clearly above it. A pipeline multiple is not revenue. It's a forecast wearing a revenue costume.

Then there's the unit economics warning that makes even "good" ROI meaningless. Directive puts it bluntly: a 300% ROI sounds great until you realize your CAC is $15,000 and your LTV is $18,000. High ROI with thin margins is not a growth engine — it's a treadmill. Always pair ROI with CAC-to-LTV ratio and payback period before celebrating any number.

This is why CFO-level ROI means revenue, not clicks. Metadata's framing is the one to remember: your CFO doesn't care about your click-through rate — they care whether the money came back with friends. Yet most teams can't even run that math today:

• Only 52% of marketing teams measure cost per dollar of pipeline
• Only 46% measure cost per dollar of new logo ARR
• 31.8% of B2B advertisers have zero CRM signal — nothing traceable past the lead stage

That last number, from Metadata's first-party data, is the real scandal. Nearly a third of advertisers literally cannot measure ROI because their ad spend never connects to revenue in the CRM.

The fix starts with plumbing, not dashboards: connect spend to CRM outcomes, compare attribution models side by side, and give long-cycle channels a fair time horizon. It's also why one integrated report beats fragmented vendor reporting — the approach Worqd builds around, tracking the whole path from first click to booked call so no channel gets credit it didn't earn and none loses credit it did.

Before you compare your numbers to any benchmark, ask which traps are baked into both. Otherwise you're not benchmarking performance — you're benchmarking accounting choices.

How to Benchmark Your Own Marketing the Right Way

Benchmarks only matter if your numbers are honest in the first place. A Metadata study of 153 B2B advertisers found that 31.8% had zero CRM signal — nothing their ads produced could be traced past the lead stage. Before you compare yourself to anyone, connect ad spend to CRM revenue.

Once your data is trustworthy, judge each channel against its own range and time horizon. A 90-day window systematically over-credits demand capture and under-credits SEO, content, and email nurture, which influence deals closing 6–18 months later — so allow SEO and content 12–36 months before rendering a verdict.

Attribution model matters just as much. Last-click underestimates content marketing value by roughly 40% and undervalues top-funnel channels like SEO and LinkedIn. As Metadata bluntly puts it, the last thing someone did before buying is usually "talk to sales" — not an ad. Use a multi-touch model (U-shaped or W-shaped) and compare models before cutting anything.

ROI alone can also mislead. A 300% ROI sounds great until your CAC is $15,000 and your LTV is $18,000 — high ROI with thin margins is not a growth engine. Pair ROI with CAC-to-LTV and payback period, and remember that real-world cost per customer ranges from about $31,939 on average to $103,331 for enterprise deals.

Finally, run a balanced portfolio rather than betting everything on one channel:

• Keep 3–5 channels active so a single algorithm change can't sink your pipeline.
• Put roughly 80% of budget into proven channels and 20% into tests.
• Reallocate only after 90 days of proven performance — not after a slow week.
• Weight toward the strongest-evidence channels: SEO, LinkedIn, and short-form video.

This is why Worqd runs every engagement as one plan and one report — ads, creative, and follow-up measured against the same revenue line, with no vanity metrics in between. Benchmarks are guardrails, not goals; the point is spotting when something is deeply off, then fixing the bottleneck instead of chasing someone else's number.