Your Google Ads account reports 5x ROAS. Your Meta dashboard says 4.2x. Your TikTok dashboard says 3.8x.
ROAS ratios cannot literally be added. Yet stack those reports in one board pack and you can end up implying 13x return from spend that generated perhaps 3x in actual revenue.
This is a measurement problem disguised as a performance problem. And it is endemic in B2B Google Ads.
The ROAS number your campaigns report is not fabricated. It is simply measuring something different from what most people think it measures. Three structural problems explain most of the gap: time, credit and value.
What Google Ads ROAS Actually Tells You
Platform-reported ROAS answers a narrow question:
Reported ROAS = conversion value attributed by Google Ads ÷ Google Ads media cost
It does not tell you how much incremental, realised gross profit the business generated because of those ads. It does not automatically account for other channels, lead quality, refunds, fulfilment costs, agency fees or sales that would have happened anyway.
The number can be calculated perfectly and still be commercially wrong.
That distinction matters even more in 2026 because platform reporting can include observed and modelled conversions, while automated bidding acts on whatever conversion actions and values you have designated as important. Our deeper audit of what Google Ads ROAS hides covers each inflation source separately.
The Attribution Window Mismatch
Many website conversion actions use a 30-day click-through window, whether inherited from an account setup or deliberately selected. View-through and engaged-view windows are separate, and settings can vary by conversion action and campaign type.
Do not assume a legacy default applies to your account. Inspect every primary conversion action.
A conversion window determines whether an ad interaction remains eligible for credit. It does not, by itself, make Google count the same conversion twice.
The reporting problem is subtler.
Suppose someone clicks an ad on 29 July and converts on 20 August. Standard Google Ads conversion columns will normally assign that value back to the date of the ad interaction. Your CRM may record the deal in August. A July report exported on 1 August will also be less complete than the same report exported three weeks later.
Nothing has necessarily been duplicated. You are comparing an immature advertising cohort with revenue recorded on a different date basis.
The opposite problem appears when the real outcome happens outside the window. If your sales cycle runs 90 days from initial enquiry to signed contract, Google may capture the form submission but never connect the eventual revenue to the campaign. You then optimise the algorithm towards an action that may have little relationship with the outcome you actually care about.
The fix is a two-part audit:
- Inspect the account’s conversion-lag and days-to-conversion reporting for each primary action.
- Compare it with CRM timings from first enquiry through qualified opportunity to closed-won revenue.
Report recent cohorts as preliminary until they have matured. If stakeholders need calendar-month revenue, show that separately rather than silently mixing conversion-date and interaction-date figures.
For long sales cycles, a form submission also needs a defensible value. Use qualification rates, stage progression, close rates and realised deal economics to model lead value and pipeline ROAS without pretending forecast revenue is cash.
Cross-Platform Double-Counting
Here is where the maths breaks completely.
Google, Meta, LinkedIn and TikTok each apply their own attribution windows, identity rules and models. None of them knows that another platform has already claimed the same commercial outcome.
A B2B buyer clicks a Google ad, later engages with a LinkedIn campaign, sees a Meta retargeting ad and eventually returns through branded search. Depending on the settings, several platforms can claim some or all of the conversion. Your CRM still contains one opportunity and one deal.
You look at the combined platform dashboard and see 530 attributed conversions. Your CRM shows 300 actual opportunities. The gap is not necessarily fraud. It is self-serving scorecard accounting.
GA4 can produce another answer because it uses different attribution logic, identities, event definitions, time zones and reporting dates. The CRM may use first source, latest source or an owner-entered field. Finance records recognised or collected revenue later still.
That is why you need an explicit hierarchy to reconcile ROAS across Google Ads, GA4 and your CRM instead of averaging incompatible numbers.
Google Search often reports the highest ROAS because it sits close to the point of conversion. A meaningful portion can be demand harvesting: capturing people who were already going to find you through organic search, brand awareness, referrals or activity on another channel.
Search intercepts an already-decided buyer at the bottom of a funnel it did not necessarily build. That is not always incremental return. Sometimes it is credit-claiming.
This does not mean Google Ads does not work. It means attribution cannot prove what Google Ads added.
Incrementality testing provides the stronger answer. Depending on your volume and campaign type, that might mean an audience holdout, geo experiment or controlled brand pause. Compare the exposed group with a credible unexposed group and measure the difference.
Attribution asks who receives credit. Incrementality asks what would not have happened otherwise.
The Brand Tax
Branded keyword campaigns are the stealth ROAS inflator almost nobody talks about.
When you bid on your own name, you are often capturing traffic that would have found you through organic search anyway. The conversion rate is high because the buyer searched for you by name. The attributed ROAS looks excellent. The incremental value may be much smaller.
There are legitimate reasons to run branded campaigns:
- Defending against competitor ads
- Controlling the search results page
- Promoting specific offers or sitelinks
- Testing messaging
- Protecting high-value searches when organic coverage is weak
But branded, generic-intent and competitor campaigns perform fundamentally different jobs. Reporting one account-level ROAS across all three conceals the information needed to allocate the next pound.
An account showing 5x ROAS might be producing 2.8x from generic searches and 9x from branded searches. The first figure describes traffic the advertising had to compete for. The second largely describes demand already attached to your name.
Use separate campaigns, budgets and reporting rows. Apply brand exclusions and query controls where campaign types could absorb branded traffic. Do not allow a generic campaign to inherit the brand campaign’s economics unnoticed.
Which number should you be making decisions from?
The Costs Missing From the Dashboard
Even perfect attribution would not make platform ROAS a profit measure.
Google only knows the conversion values you send and the media cost in the account. If every form submission is assigned £1,000, Google reports that value whether the lead is qualified, duplicated, fraudulent or never contacted.
For ecommerce, the reported value may ignore returns, cancellations, discounts, payment charges, fulfilment and cost of goods. For B2B services, it may use total contract value even though revenue arrives over several months and delivery consumes a substantial share of it.
A commercially useful calculation separates the layers:
True revenue ROAS = realised net revenue ÷ fully loaded advertising cost
Contribution ROAS = contribution generated by acquired customers ÷ fully loaded advertising cost
Fully loaded advertising cost may include media, management, creative, landing-page production, tracking tools and platform fees where applicable. Net revenue should reflect refunds, returns and cancellations. Contribution should account for the direct costs and margin attached to each sale.
The exact inclusions depend on the decision, but they must be written down and applied consistently. Our worked guide explains how to calculate true ROAS with fees, returns and margin included.
A flattering 6x platform ROAS can become 3.7x after invalid leads and cancellations, then fall below the commercial target once margin and acquisition costs are applied. That is not pessimistic reporting. It is the number the business can actually spend against.
If the path from ad click to signed revenue still has gaps, Actualyse will trace each hand-off with you — book a call
What a Real B2B Attribution Audit Looks Like
Answer these six questions before trusting any ROAS figure.
1. What conversion action is receiving the value?
Is it a page view, form submission, demo request, qualified lead, proposal, signed contract or collected revenue?
Check which actions are primary, which are secondary and whether the same enquiry can trigger more than one primary conversion. If Smart Bidding sees only form submissions, it will find more people who submit forms—not necessarily more people who become clients.
2. Does the reporting window match the conversion lag?
Pull the lag data for the tracked conversion, then compare it with the CRM sales cycle.
If most form submissions occur within 14 days but deals close three months later, those are two different lags. A short advertising report may be useful for optimisation, but it cannot be presented as final revenue performance.
3. Are branded and non-branded campaigns separated?
If not, you cannot make intelligent budget-allocation decisions.
Show branded, generic and competitor performance independently. Include spend, leads, qualified opportunities, won revenue and margin—not just the platform conversion total.
4. Is offline outcome data feeding back into Google Ads?
A typical B2B journey might be:
Enquiry → qualified lead → opportunity → proposal → closed-won revenue
If only the first step is imported, the bidding system is optimising against the weakest signal.
Use unique lead identifiers and a consented CRM-to-platform process, such as enhanced conversions for leads or an equivalent offline conversion import. Feed back meaningful stages and values, not every internal status change.
The end-to-end B2B ROAS measurement framework explains how the CRM, analytics, advertising and finance layers should connect.
5. Are conversions deduplicated across systems?
Check transaction IDs, lead IDs and opportunity IDs.
Look for duplicate thank-you-page events, repeated phone-call imports, reopened opportunities, test leads and multiple platforms importing the same sale under different identifiers. Decide whether reports use interaction date, conversion date, close date or payment date.
Without these rules, reconciliation becomes spreadsheet theatre.
6. Does the calculation include the right revenue and costs?
Confirm whether the numerator is assigned lead value, pipeline value, signed contract value, recognised revenue, collected revenue or contribution.
Confirm whether the denominator contains media only or the fully loaded acquisition cost. State both definitions on the report so nobody can mistake a platform efficiency metric for profit.
The accounts that get this right—clean signals, CRM feedback, separated campaign types, matched reporting periods and consistent cost definitions—consistently outperform accounts with higher budgets but broken measurement. The algorithm is only as smart as the data you give it.
If an agency supplies the report, use this checklist to audit its claimed ROAS in one afternoon.
Which Number Should Drive the Budget?
Do not force one metric to perform every job.
- Platform ROAS is an optimisation signal. Use it to compare campaigns inside a platform once the conversion setup is clean.
- Non-brand revenue ROAS helps assess whether the channel is acquiring demand efficiently.
- Blended ROAS compares total business revenue with total paid-media spend. It is useful for strategic trends but can be affected by organic demand, seasonality and returning customers.
- MER commonly compares total revenue with total marketing expenditure, although some teams use it as another name for blended ROAS. Put the formula on the report.
- Contribution ROAS, CAC and payback help finance decide whether growth is affordable.
- Pipeline ROAS provides an early directional signal for B2B sales cycles, but it must later be reconciled with realised outcomes.
The governance rules in blended ROAS versus platform ROAS show when each metric should control a decision. It is also worth defining exactly how ROAS and MER measure different things before either appears in a board pack.
A practical decision hierarchy is:
- Finance confirms realised revenue, margin and total cost.
- The CRM confirms lead quality, opportunity progression and won customers.
- Analytics explains journeys and cross-channel behaviour.
- Advertising platforms optimise campaigns using the cleanest available signals.
Use platform data to operate the platform. Use CRM and finance data to decide whether the operation is commercially working.
The Harder Question
If you fix all of this and your ROAS comes down significantly, that is not a failure. That is clarity.
A business making decisions from accurate data is in a materially better position than one with a flattering number nobody believes. Most B2B founders already suspect their ROAS figures are inflated. They simply have not had a framework for locating the problem.
Fixing the measurement infrastructure requires CRM integration, cleaner conversion tracking, campaign restructuring and disciplined reporting. It may take implementation time, consent-aware data handling and cooperation between marketing, sales and finance. That is still cheaper than scaling an account using a false signal.
Actualyse builds high-precision B2B Google Ads frameworks that optimise for revenue, not just clicks. If you want to find the gap between claimed and commercial return, request a measurement audit.
FAQ
What is a good ROAS for B2B Google Ads?
There is no universal healthy range. The required ROAS depends on gross margin, sales costs, customer retention, cash flow and payback period.
If your gross margin is 40%, a 2.5x revenue ROAS only covers media spend before agency fees, overheads and other acquisition costs. Branded campaigns may report much higher figures without producing the same incremental value as generic campaigns.
Set the target from your unit economics, then report branded and non-branded performance separately.
Why is Google Ads ROAS higher than my revenue growth?
Common causes include cross-platform attribution, branded-search inflation, assigned lead values, immature conversion cohorts and a mismatch between interaction-date advertising reports and close-date CRM revenue.
Google also reports attributed conversion value rather than company-wide incremental revenue. Reconcile campaigns against unique CRM opportunities and finance-approved revenue before treating the number as commercial return.
Does server-side conversion tracking improve Google Ads performance?
It can improve data resilience, matching and the import of offline CRM outcomes. Its greatest B2B benefit is allowing qualified leads and won revenue to inform bidding instead of relying only on form completions.
It does not override consent requirements, recover every missing interaction or prove incrementality. Better collection cannot compensate for a weak conversion definition.
How can I tell whether Google Ads generated incremental revenue?
Run a controlled experiment where practical: an audience holdout, geo test or carefully governed brand pause. Compare conversion or revenue outcomes between exposed and credible unexposed groups.
The test needs enough volume and time to distinguish genuine lift from normal variation. Attribution reports are useful operational proxies, but they are not proof of causation.
Should platform ROAS or blended ROAS control my budget?
Use platform ROAS for campaign-level optimisation. Use blended ROAS, contribution, CAC and payback to decide the total amount the business can afford to invest.
If the two tell conflicting stories, investigate brand demand, attribution overlap, reporting dates, lead quality and missing costs before increasing spend.
Summary
- Google Ads ROAS measures attributed conversion value against media cost—not incremental profit.
- Conversion windows create eligibility and timing differences; they do not automatically duplicate the same conversion.
- Reconcile platform reports with unique CRM outcomes and finance-approved revenue.
- Separate branded, generic and competitor campaigns before making budget decisions.
- Feed qualified leads and offline revenue back into bidding with reliable deduplication.
- Include returns, fees, delivery costs and margin when calculating commercial return.
- Use platform ROAS to optimise campaigns and business-level metrics to set budgets.
Actualyse builds measurement and attribution setups that tie B2B ad spend to real revenue. Book a call to talk through where yours stands.
