September. Quarterly marketing meeting. They show you a report with 27 slides: reach, impressions, new followers, likes, comments, best posting time. Nice. Visual. Convincing.
And then you ask “OK, how much money did we generate?” and everything slows down. The classic “it depends” appears, along with “it is hard to measure” and “social media returns are long term”.
Measuring social media ROI requires consistent revenue, cost and attribution data. Where information is missing, make the limitation explicit and improve measurement. These eight KPIs help interpret activity and results; none replaces the profitability calculation on its own.
What real ROI means in social media
ROI (Return on Investment) relates attributable net profit to the costs of generating it. Include the cost of the product or service, advertising, agency fees, content and other attributable costs, without counting any cost twice:
ROI = ((Attributable revenue − Total attributable costs) / Total attributable costs) × 100
Illustrative example: if you spend €1,000 on advertising and generate €4,200 in revenue attributable to those ads, your ROAS is 4.2× (420%): €4.20 in revenue per euro of ad spend. These figures are not enough to calculate ROI: the cost of the product or service and the other attributable costs are still needed.
The next step is attributing revenue and costs to social media consistently. A customer rarely buys after a single post: they see a Reel, search your brand on Google, read a review, receive an email and call two weeks later. Who owns that customer? The honest answer is: the whole ecosystem. But there are ways to measure contributions.
ROI, ROAS, CPA: what each one measures
Before getting into specific KPIs, three concepts need to be clarified because they are constantly mixed up and they are not the same.
| Metric | What it measures | Formula | When to use it |
|---|---|---|---|
| ROI | Profitability after product or service costs and all attributable marketing costs | (Attributable revenue − Total attributable costs) / Total attributable costs × 100 | For quarterly or annual budget decisions |
| ROAS | Return only on ad spend | Revenue attributable to ads / Ad spend | To optimise individual campaigns |
| CPA / CAC | CPA: cost per defined action. CAC: cost of acquiring a new customer. | CPA = attributable cost / defined actions. CAC = acquisition costs / new customers. | To understand whether the unit economics work |
The most common mistake is reporting ROAS and calling it ROI. An agency tells you “your ROAS is 4×” and it sounds great, but if their management fee and the cost of content are not included, your real ROI may be negative. Always ask for the full formula.
The 8 KPIs that do matter in social media
At islaNet we use these indicators to analyse social media investment. Select those relevant to your objectives and interpret them alongside revenue, costs and attribution limits.
Cost per lead (CPL) is attributable acquisition cost / number of leads for the same period and scope. If measuring qualified leads, use that same qualification criterion in the denominator and comparisons. There is no universal benchmark: compare periods, channels, lead quality and conversion consistently.
What it costs to acquire a customer. Formula: Acquisition costs / New customers. Compare CAC with customer value adjusted for margin and the costs of serving that customer over the relationship. Comparing CAC with revenue alone does not show whether acquisition is profitable.
Out of every 100 people coming from Instagram, how many buy? Formula: (Conversions / Visits from channel) × 100. It tells you which network deserves more investment. A network with 0.5% conversion but high traffic may perform better than one with 4% conversion but little volume.
How much attributable revenue each euro of advertising generates. Formula: Revenue attributable to ads / Ad spend. The ROAS required for profitability depends on margins and the other business costs; there is no universal threshold. Warning: ROAS is not the same as ROI.
You can define a selected engagement rate as (comments + saves + shares) / reach × 100. Document which interactions are included and keep the denominator consistent. These actions do not prove purchase intent or predict sales on their own.
Share of Voice is brand mentions / total mentions in the analysed set × 100. Define brands, sources and period. A change may reflect conversation or measurement coverage; it does not by itself equal market share or sales.
Measure the time between the first recorded touch and conversion: total days across those journeys / number of conversions. Document the window and which touches are observable. The average helps planning but does not determine when each purchase will occur.
An estimate of customer revenue over the entire relationship. Simplified revenue formula: Average ticket × Purchases per year × Retention years. To estimate how much you can invest in acquiring a customer, adjust that revenue for margins and service costs, and consider when payments are received.
The 3 KPIs that do not matter as much as you think
These three KPIs are not useless: they are intermediate indicators. The problem appears when someone reports them as if they were business results. They are not.
Follower count indicates the size of the recorded audience, not its profitability. Review relevance and activity alongside enquiries, sales and costs; account size alone does not establish which business sells more.
Likes are an engagement signal but do not prove revenue or purchase intent. Interpret them alongside reach, content and conversions; an isolated total does not demonstrate profitability.
Distinguish reach from impressions: reach estimates people or accounts reached, while impressions count appearances and can include repeats. Check each platform’s definition and audience relevance; neither figure equals sales.
We are not saying these metrics should be ignored. They are useful as intermediate indicators of account health. But as ROI metrics or “proof of results”, they are noise. If your agency only reports these three, you have a problem.
Your minimum viable dashboard: what to measure and where
You do not need a €400/month tool. This is the minimum you need to understand what is happening with your investment. Monthly.
| Data | Where to get it | Frequency |
|---|---|---|
| Attributable product or service and marketing costs (ads + agency + production) | Invoices + bank statement | Monthly |
| Total leads generated | Meta Business + CRM or Google Sheet | Monthly |
| Closed customers attributed to social media | CRM with “source” field | Monthly |
| Conversion rate by channel | Google Analytics 4 / GA4 + UTM | Monthly |
| ROAS by active campaign | Meta Ads Manager / Google Ads | Weekly |
| Qualified engagement | Native insights from each platform | Monthly |
| Share of Voice (optional but useful) | Brand24, Mention or manual search | Quarterly |
This dashboard is a starting point. Add attributable revenue and product, service and marketing costs, and record the period, attribution window and model. Its usefulness depends on data coverage and quality; it does not represent a fixed percentage of ROI visibility.
Common mistakes when measuring social media ROI
From the audits we see every year, these are the mistakes that most compromise the reading of results:
- Not counting the cost of content. The agency fee + photo/video expenses count. If someone says “your ROI is 5×” but does not include the cost of producing the content, the calculation is lying.
- Confusing the last click with the full journey. If the customer saw your Reel, read your blog and received your email before buying, the last click shows only part of that journey. In GA4, compare data-driven attribution with last-click attribution and document the chosen approach. No model proves the causal effect of each interaction on its own.
- Ignoring average conversion time. Expecting results in 30 days when your average cycle is 60. Poor comparison leads you to stop campaigns that were actually working.
- Not segmenting by audience. An average ROAS of 4× can hide the fact that one audience performs at 8× and another at 0.5×. Without segmentation, you optimise blind.
- Only measuring what is easy. Followers and likes can be measured in two clicks. Conversion takes more work: UTM, CRM, attribution. If your agency only measures what is easy, it is not measuring what matters.
FAQs about measuring ROI
What is a good social media ROI for an SME?
There is no universally good ROI. It depends on margins, risk, the time horizon and the costs included. With the formula used in this article, a positive ROI means attributable revenue exceeds all the attributable costs considered. Set your own target and compare periods consistently; do not confuse a ROAS multiple with an ROI percentage.
How long should it take to see positive ROI from social media?
Timing depends on the starting point, budget, offer and buying cycle. Define review milestones and allow a window compatible with that cycle. There is no universal week for initial signals or guaranteed month of positive ROI.
How do I attribute a sale to social media if the customer came through several channels?
Review attribution paths and compare the models available in Google Analytics 4, such as data-driven and last-click attribution. Cross-check the data against your CRM and its recorded sources. Define the model, time window and included channels, then apply a consistent approach when comparing periods. Do not automatically split credit equally: attribution is an estimate, not proof of causality.
Why is my ROAS high but my ROI low?
ROAS compares revenue attributable to ads with advertising spend. ROI must also consider the cost of the product or service, agency fees, creatives, production and other attributable costs. A high ROAS can therefore coexist with a low or negative ROI. Those costs are needed to calculate it; ROAS alone is not enough.
What basic tool do I need to measure ROI?
You can combine correctly configured web analytics, advertising reports, a sheet of attributable revenue and costs, and a CRM with each lead’s source. Document the period, attribution model and window and review the available data; no tool combination guarantees a fixed coverage percentage.
Is it normal that my agency does not give me clear ROI?
It is not normal, but it is common. Many agencies prefer to report activity metrics — posts, followers, reach — because they are easy to show. A professional agency gives you, at least quarterly, a dashboard with total investment, leads, closed customers, CAC and ROI. If yours only gives you likes and reach, ask for more. If they cannot or will not, you know.
How do I measure ROI if I sell in a physical shop with no e-commerce?
Combine campaign codes, the customer’s stated source and campaign-linked bookings where appropriate. Cross-check against sales and costs and document duplication, bias and unobserved touches. The coverage needed depends on the decision; there is no universally sufficient percentage for allocating budget.
DO YOU WANT A REAL DASHBOARD FOR YOUR SOCIAL MEDIA?
At islaNet, we implement custom ROI dashboards: we integrate GA4, Meta Ads, your CRM and your investment sheet into one dashboard that updates automatically. No more endless Excel files, no more meetings with decorative metrics. Actionable data, faster decisions.
We work with local and enterprise brands from Palma de Mallorca. We are an accredited Kit Digital digitalisation agent and Kit Consulting digital adviser. As of 1 October 2026, the calls for new applications to both programmes are closed. An already-awarded voucher remains subject to its service category, agreements and deadlines; Kit Consulting covers advice, not ongoing social media execution.
