Social media ROI calculator

ROAS answers one narrow question: for every dollar of media, how many dollars came back. It is the right metric for a media buyer optimising a campaign, and the wrong one for deciding whether a social programme is worth running.

Consider a programme doing $18,000 a month in attributed revenue on $2,500 of ad spend. That is a 7.2× ROAS — excellent by any standard. Now add a 60% gross margin, $300 of tools, and one person spending fifteen hours a week at a $45 loaded cost. Gross profit is $10,800; total cost is about $5,725. The ROI is roughly 89% — still good, but a completely different conversation than 7.2×.

Drop the margin to 25%, as it would be for many physical products, and the same campaign is losing money every month while its ROAS dashboard stays green.

The fastest way to make this calculation useless is to be generous with what counts as social revenue. If your attribution model credits social for every purchase that ever touched a social post, the ROI number is a story rather than a measurement.

Consistency matters more than picking the theoretically correct model. Whatever window and model you use, use the same one every month so the trend means something. Consistent UTM tagging is the unglamorous foundation of all of it — inconsistent casing alone will split one channel into three rows in GA4 and quietly understate your results.

Social also carries value that direct attribution structurally cannot see: branded search lift, sales that close over the phone, the customer who found you eight months ago. That is an argument for treating a modest positive ROI as a genuine win, not an argument for inflating the input.

When ROI comes out negative, the instinct is to cut ad budget. But look at the cost split first — on most in-house programmes labour is 60–80% of the total. Halving a $2,500 ad budget saves less than removing four hours a week of production time.

That does not mean working less. It means the same output taking fewer hours: batching a month of content in one sitting instead of daily, adapting one idea across networks rather than starting fresh each time, and letting scheduling remove the manual publish step entirely.

Run the same numbers with your hours cut by a third and watch what happens to the ROI figure. It moves far more than any realistic change to ad spend or conversion rate.