What RPS is and why it matters

The formula is total revenue divided by total sessions over the same period. A store generating $50,000 from 25,000 sessions last month has an RPS of $2.00, and what makes the number useful is that it captures two things at once: how often people buy, which is conversion rate, and how much they spend when they do, which is average order value. Conversion rate reflects only the first half, AOV only the second, and RPS combines them into a single figure representing the actual revenue productivity of your traffic.

Here’s the distinction working in practice. Say you run a 25 percent sitewide promotion and conversion rate jumps from 2.5 to 3.8 percent, a clear win by that dashboard. But average order value drops from $85 to $58 as people buy smaller orders with a discount cutting into every transaction, so RPS moves from $2.13 to $2.20, a 3 percent improvement, not the 52 percent lift the conversion dashboard celebrated, and after the margin impact of the discount, the promotion may have lost money outright. Without RPS, you’d run that promotion again. With it, you see the whole picture and decide better.

Where RPS gets really useful: segmentation

The aggregate number is a fine topline health check, but the value comes from breaking it down, because segmented RPS finds problems and opportunities other metrics hide.

By traffic source. We see this constantly in client accounts: a paid social campaign driving heavy sessions at a low cost per click, the team excited about volume, and an RPS for that source of $0.40 against $3.50 for branded search and $2.10 for email. The campaign is filling the funnel with low-intent visitors who browse and leave, and once you calculate the actual cost of generating a dollar of revenue from the channel, the cheap traffic isn’t cheap at all.

By device. Mobile RPS trailing desktop is a quantifiable UX problem, not the vague “mobile converts worse” everyone already knows, but a dollar amount attached to every underperforming session. At 60 percent mobile traffic with mobile RPS of $1.20 against $3.80 on desktop, you can compute exactly what the gap costs per month, and that number gets attention in rooms where “we should improve the mobile experience” never has.

By landing page. RPS by entry point shows which pages attract visitors who spend money rather than visitors who click. A blog post might drive real traffic at an RPS near zero because its readers are researching, while a well-built collection page drives fewer sessions at $4.50. That doesn’t make the blog post worthless, it makes the case for thinking differently about which entry points deserve paid promotion budget.

By customer type. New-versus-returning RPS is one of the most telling comparisons a store can run. Returning visitors typically produce two to five times the RPS of new ones, which is expected, but the ratio carries information: an extreme gap says acquisition is bringing in poor-fit visitors or the site isn’t converting first-timers well, and a narrowing gap over time says top-of-funnel targeting is improving.

What RPS doesn’t tell you

Positioning RPS as the only metric that matters would be a disservice, so here are its limits. It doesn’t account for margin: a $5 session on an 80 percent margin product beats an $8 session at 20 percent, and gross margin per session is the more powerful version if your data can support it, which most stores’ can’t without work, the gap we mapped in connecting marketing metrics to margins. It doesn’t capture lifetime value: a $1.50-RPS first-timer who makes six repeat purchases beats a $4.00 one-time buyer, so subscription businesses and strong-repeat brands need LTV-informed metrics alongside it. And it misleads during promotions, launches, and seasonal peaks, a Black Friday RPS spike says nothing about baseline performance, so compare like-for-like periods and smooth with seven- or twenty-eight-day rolling windows. None of these limits make RPS less useful. They mean it works best as a primary lens combined with others, not a replacement for all of them.

How we use RPS in client work

When we evaluate a store’s performance, RPS across segments is usually the first thing we calculate, not because it answers every question but because it surfaces the right questions faster than anything else. Flat overall RPS while traffic grows means the new traffic is worth less than the existing traffic, pointing at acquisition quality. Declining RPS with stable conversion rate means order values are dropping, pointing at merchandising, pricing, or product mix. Wild variation between traffic sources means the media mix needs rebalancing, since high-volume, low-RPS channels are usually over-budgeted relative to quieter channels earning multiples more per session. And mobile lagging desktop means a UX problem with a price tag attached, ready to be prioritized. Each is a starting point for deeper investigation rather than a conclusion, but RPS reaches the right starting point with more confidence than conversion rate, AOV, or traffic volume ever manage alone.

The practical takeaway

If you’re not tracking RPS by segment, start. GA4 supports it as a custom metric or a calculation from standard reports, and in Shopify you can pull revenue and session data into a spreadsheet and do the division. The segments that matter for most stores are traffic source, device, landing page, and new versus returning, and a month or two of baseline data across those four will surface patterns that tell you where optimization effort belongs. RPS won’t tell you everything. But it tells you things conversion rate and AOV separately can’t, and those tend to be the things that actually move revenue.