Measuring Social Media ROI: The Five Metrics That Actually Predict Revenue

Published 22 May 2026 · Last updated 22 May 2026 · 9 min read

The problem with most social reports

Most social media reports are theatre. Reach, impressions, follower growth, engagement rate. Numbers that go up, presented in coloured charts, with no connection to whether anyone bought anything. I've sat through more of these than I'd like to admit. The business owner nods, the agency nods, the report goes into a folder, and the next month's invoice gets paid.

Vanity metrics are not useless. They're just not predictive of revenue. If the question you want answered is "is our social media producing money", you need a different set of numbers. Here's the set we use with our clients.

The five metrics that actually predict revenue

1. Branded search volume lift

If your social media is working, more people are searching for your brand. They saw something on Instagram, they're now Googling your business name. This shows up in Search Console as growth in branded queries.

To measure: pull branded impressions and clicks from Search Console month over month. Cross-reference with your social posting cadence. If you posted more in March and branded searches went up in April, social is doing its job. If they didn't, it isn't. For the broader measurement framework, see SEO KPIs.

2. Assisted conversions from social

Most conversions from social don't happen on the click. Someone sees a post on Instagram on Monday, thinks "I should check that out", Googles you on Thursday, fills the form on Friday. In Google Analytics 4 that shows as an organic conversion. The social touch was real but invisible in last-click attribution.

Use GA4's data-driven attribution model (not last-click) to see assisted conversion paths. Add a "how did you hear about us?" question on the contact form. The honest answer for most service businesses is that social's contribution is roughly 2 to 4 times what last-click reporting shows.

3. Dark traffic correlated with posting cadence

Dark traffic is the chunk of your direct or referral traffic that has no UTM tag and no clear source. It's people pasting your URL from a DM, clicking from inside an app, or arriving via an AI search referrer that doesn't pass referrer headers cleanly.

The way to attribute this to social: track direct traffic and look for correlation with posting cadence. If your direct traffic spikes the day after a strong post, you have a working signal even if the analytics can't tell you so. ChatGPT and Perplexity referrers often arrive as direct or with a partial referrer string. Treat the trend, not the individual number.

4. Cost per qualified lead, by channel

This is the metric that tells you whether to keep going. Track:

• Hours spent on social per month (or agency fee, or paid spend)
• Qualified leads attributed to social (using #1, #2 and #3 above as your attribution stack)
• Divide to get cost per qualified lead

Compare to other channels. A Perth professional services firm might find SEO leads cost $80 per qualified lead, Google Ads leads cost $140, and social leads cost $230. That tells you something useful about where to invest the next dollar. For the structural side of CRO and measurement, see CRO for SEO.

5. Lifetime value of social-sourced customers

The cost-per-lead number alone is misleading. If social leads cost $230 but spend $4,000 over their lifetime and SEO leads cost $80 but spend $900, social wins. The full picture needs LTV.

This requires you to tag the source of every new customer somewhere persistent (CRM, accounting software, anywhere durable) and look at it twelve months later. Most Australian SMEs never close the loop. The ones that do make smarter marketing decisions for years afterward.

The vanity metrics worth tracking anyway

Not because they predict revenue, but because they're useful leading indicators or sanity checks:

• Engagement rate: useful as a check on whether content is reaching the right audience, not a revenue predictor
• Saves and shares: stronger signals of usefulness than likes; saves in particular correlate with future action
• Profile visits: a leading indicator of interest that often precedes a website click
• DM and reply volume: for service businesses that get inquiries via DM, this can be a direct revenue signal

Per-channel framing

Instagram

For Instagram, profile visits + saves + DM volume are the strongest leading indicators. Reach and follower growth are vanity for most SMEs. We've written more on the platform's evolution at Instagram for small business in 2026.

Facebook

For Facebook, group activity, event responses and Marketplace inquiries (for relevant businesses) outweigh feed metrics. Branded search lift is the most useful long-term measure.

LinkedIn

For LinkedIn (B2B), profile visits + connection requests + inbound DMs from decision-makers are the predictive metrics. Likes on posts mean almost nothing. Saves on posts mean something.

TikTok

For TikTok, view-through rates above 50 per cent and a steady climb in profile visits are the predictive signals. Vanity metrics on TikTok (views, follower count) often look great and produce nothing. The watchful metric is brand-search lift in the week following a viral post.

What to report to non-marketing stakeholders

If you're reporting to a board, a partner, or a non-marketing owner who just wants to know whether the spend is working, present this:

1. Qualified leads from social this month (with comparison to last month)
2. Cost per qualified lead vs other channels
3. Branded search volume trend (12-month chart)
4. One sentence on what worked and one sentence on what didn't

That's it. That's the report. No screenshots of Instagram insights, no slides about "reach". The four numbers above answer the actual question: is this making us money or not?

Setting up honest social attribution

Practical setup, in order:

1. Add UTM tags to every link you put in social bio sections and posts where supported
2. Set GA4 to use data-driven attribution as the default report
3. Add "how did you hear about us?" to every form and inquiry
4. Track CRM-tagged source of every new customer for at least 12 months
5. Build a monthly correlation chart: posting cadence vs branded search vs direct traffic

If you want the broader measurement architecture, see our SEO measurement pillar.

How long until the data is meaningful

Three to six months of consistent activity before the numbers tell you anything reliable. Most Australian SMEs measure social ROI after six weeks, conclude it doesn't work, and walk away. The data wasn't there yet. The work was just starting to compound.

If you're three months in and the five metrics above are flat, that's a real signal. If you're six weeks in and the metrics are flat, it's noise.

A simple reporting template

If you want a structure to copy, here's the monthly social ROI report we use with our clients. Five sections, half a page each:

1. Headline numbers: qualified leads this month, cost per qualified lead, total spend (cash + time)
2. Trend chart: branded search volume over the last 12 months with posting cadence overlaid
3. Best performing content: the three posts with the highest profile-visit-to-follower ratio, and what they had in common
4. What didn't work: two or three posts that produced nothing, and a hypothesis why
5. Next month's focus: the one or two experiments planned, with what you're testing

This format takes 30 minutes to compile if you've set the metrics up well. It produces a report that a non-marketing owner can actually read and act on, which is the point.

Want us to set up proper social attribution and reporting for your business? Get a free SEO audit, or look at our social media marketing and CRO services. We work with Perth SMEs from the metro through to Bunbury and the regions.

Related reading

• Digital marketing for Australian businesses: a 2026 field guide
• Instagram for small business in 2026
• Core Web Vitals and conversion

FAQ

What is the best metric for measuring social media ROI?

There isn't a single metric. Five together give a useful picture: branded search volume lift, assisted conversions from social, direct/dark traffic correlated with posting cadence, cost per qualified lead by channel, and customer lifetime value of social-sourced customers. Vanity metrics like followers and reach do not predict revenue.

How do you attribute revenue to social when most traffic shows as direct or dark?

Use a combination of UTM-tagged links where possible, post-purchase surveys ("how did you hear about us?"), and correlation analysis between posting cadence and branded search lift. Perfect attribution is impossible. Directional attribution is enough for decision-making.

Is engagement rate a useful social media metric?

Engagement rate is a check on whether content is reaching the right audience, not a predictor of revenue. A post with 10 per cent engagement and zero conversions isn't doing your business any good. Use engagement as a leading indicator alongside the revenue metrics.

How long until you can measure social media ROI honestly?

Three to six months of consistent activity before the data is meaningful. Anything shorter is noise. Most Australian SMEs measure too early, conclude social doesn't work, and walk away just as the compounding starts.

Which social platform delivers the highest ROI for Australian SMEs?

It depends on the business. Trades and lifestyle brands tend to see best returns on Instagram and TikTok. B2B and professional services on LinkedIn. Local service businesses often find Facebook still works for the demographic. The right answer is the platform where your buyers actually scroll.