The dashboard can look healthy while the business is not

You spend on Meta to create demand. A customer sees the ad, later searches your business name on Google, clicks a search ad and buys. Meta may report the purchase because of the ad impression or click. Google may report the same purchase because of the search click. Add email, direct traffic or an organic result and several systems can describe one order as their success.

That does not necessarily mean either platform is lying. Google Ads uses attribution models to distribute credit across interactions, and its data-driven model estimates the contribution of interactions along the conversion path. Google also includes modeled conversions where it believes there is enough evidence to estimate conversions it cannot directly observe. (support.google.com)

Meta operates with its own attribution settings and reporting logic. In 2026, practitioner Jon Loomer reports that the standard website conversion setup uses a seven-day click and one-day engaged-view or view-through framework, depending on the campaign configuration. That window affects both reporting and what the system learns to pursue. (jonloomer.com)

So the answer is yes: when platforms claim overlapping conversions, a small business should judge advertising by the money left after the sale, across the whole business. But “blended profit” is not a magic replacement for attribution. It is the business-level scorecard. Platform ROAS remains a diagnostic. Incrementality is the test of cause.

Blended profit beats ROAS for the decision that matters

Use contribution profit, not revenue, as the bridge to reality

Blended ROAS, also called MER, is usually calculated as total revenue divided by total marketing spend. Shopify describes it as a high-level measure of overall marketing efficiency, useful for comparing the business over time but unable to identify which campaign drove an individual sale. (shopify.com)

That is useful, but revenue is still one step removed from profit. A sale can carry a strong ROAS and leave little behind after product cost, shipping, payment processing, discounts, refunds and staff time. Klaviyo makes the same distinction between ROAS and profit on ad spend: ROAS measures gross revenue, while a profit-based measure accounts for the costs attached to fulfilling the order. (klaviyo.com)

For a small business, the practical measure is contribution profit after variable costs and advertising. You do not need a perfect finance department to begin. Pick a consistent definition. For ecommerce, that may be net sales minus product cost, fulfilment, payment fees, discounts, refunds and paid media. For a local service business, it may be collected revenue minus contractor labour, materials and advertising.

Then compare that number with the same period’s total marketing cost. Include the channels you actually pay for, and decide whether creative, software, agency or freelance costs belong in the number. Shopify notes that MER changes depending on what a business includes in spend, so consistency matters more than pretending there is one universal formula. (shopify.com)

This is also where a fast, clear website usually converts better than an impressive one becomes a measurement issue, not just a design preference. If the landing page leaks demand, platform reports may still show clicks and attributed conversions, while the business absorbs the cost of the friction.

The numbers show why small businesses need a simpler scorecard

The measurement problem is not limited to large ecommerce brands. Shopify’s survey of store owners in the fourth quarter of 2025 found that seventy-seven percent tracked sales or total revenue, while fewer than half tracked profit margin, traffic, average order value or conversion rate. Among businesses with at least one million dollars in revenue, Shopify reported that revenue tracking was at eighty-five percent, compared with fifty-seven percent for profit margin and forty-four percent for ROAS. (shopify.com)

Intuit’s March 2025 survey of US small and medium-sized businesses found that twenty-five percent said they could always measure advertising ROI, forty-seven percent usually could, and twenty-three percent sometimes could. The figures suggest a practical truth: many owners are being asked to make budget decisions with partial evidence, not a clean attribution system. (digitalasset.intuit.com)

That is why I would not make a small business owner reconcile every disagreement between Google Ads, Meta Ads, analytics software and the CRM before making a decision. First establish a trusted cash and profit view. Is total collected revenue rising? Is contribution profit rising? Is customer acquisition costing more? Are refunds, discounts or fulfilment costs changing? If those answers worsen while both platforms report excellent ROAS, the business has its answer.

Blended profit also protects you from a common mistake: scaling the platform with the best reported ROAS when that platform is harvesting demand created elsewhere. Branded search and retargeting can appear unusually valuable because customers who were already close to buying are more likely to encounter them. Klaviyo specifically flags branded search and retargeting as channels that can look more valuable than they are when the customer would have converted through a free or owned channel anyway. (klaviyo.com)

A practical measurement order for the next reporting cycle

1. Start with collected sales

   Use the payment system, ecommerce platform or CRM as the primary revenue source, rather than adding each platform’s attributed revenue together.

2. Subtract the costs that move with each sale

   Use a consistent contribution-profit definition that reflects product, labour, fulfilment, fees, discounts, refunds and other variable costs.

3. Add all paid marketing cost

   Include Google, Meta and other paid channels, then decide whether creative, software and outside help belong in the same management view.

4. Track platform ROAS separately

   Keep the figures for campaign optimisation, but label them as platform-reported attribution rather than total business return.

5. Compare changes over time

   Look for whether contribution profit, total sales and acquisition cost improve together after a budget or creative change.

6. Run a lift test when the decision is expensive

   Use a geographic, audience or platform experiment when you need to know whether a channel caused additional sales rather than merely received credit.

Blended profit still cannot tell you what the ads caused

Here is the important limitation. If you turn off Meta and total sales fall, that does not prove every sale previously attributed to Meta was incremental. Seasonality, competitor activity, email timing, promotions and Google demand may have changed at the same time. If total sales stay flat, that does not prove Meta had no value either. A prospect may have seen an ad, remembered the business and converted through another route.

Blended profit is therefore a strong operating measure, not a causal measurement method. It tells you whether the combined marketing system is producing enough contribution to justify its cost. It does not allocate credit cleanly between channels.

For larger decisions, controlled testing is stronger. Google describes incrementality experiments as a way to measure what happened because of advertising and what would not have happened otherwise. Its current measurement guidance puts attribution, marketing mix modelling and incrementality together because each answers a different question. Google also says its updated experiments can be run at lower spend thresholds than in the past, although the practical feasibility still depends on conversion volume, geography and the value of the decision. (support.google.com)

Most small businesses will not run a formal test every week. They can still create useful evidence. Pause branded search for a carefully chosen period if the risk is acceptable. Split regions when the service area allows it. Hold back a small audience. Compare a new prospecting campaign with a similar control. Treat the result as evidence with uncertainty, not a perfect answer.

Use ROAS to steer campaigns, and profit to decide whether to keep driving

The right answer is not to delete ROAS from the reporting sheet. It is to put it in its proper place.

Use platform ROAS for local decisions inside a platform. Which creative earns cheaper qualified traffic? Which audience produces better leads? Which search themes deserve more budget? Those questions need granular data, and platform reporting is often the fastest available signal.

Use blended contribution profit for the business decision. Are all paid channels together producing enough margin after variable costs? Is advertising improving cash generation, or only making the dashboards look busy? If the blended number deteriorates while individual ROAS stays strong, investigate overlap, branded demand, discounting, tracking changes and website friction before increasing spend.

The best small-business reporting page is usually modest: collected revenue, contribution profit, total marketing cost, new-customer cost, refunds or cancellations, and platform-reported results shown underneath. That is enough to stop a duplicated conversion from becoming a duplicated victory.

A final caution: do not compare one month’s blended result with another month’s platform ROAS and call the difference a finding. Keep the revenue definition, cost definition, attribution windows and time period consistent. Then use experiments when a budget change is large enough that being wrong would be expensive.