You posted the same thing to TikTok and LinkedIn. TikTok gave you 50,000 views. LinkedIn gave you 3,000. Obvious which one won, and obvious where you should spend next week.
Except you have not looked at what either one earned. When you do, the picture often inverts completely. The 3,000-impression post drove $60. The 50,000-impression post drove $50. Per thousand people reached, LinkedIn was worth twenty times more. Every hour you shift toward the bigger number is an hour spent on the worse channel.
RPM, revenue per 1,000 impressions, is the number that catches this. It is borrowed from advertising, where it has been the standard unit for decades precisely because it makes channels of wildly different sizes comparable.
The formula
RPM = (revenue ÷ impressions) × 1,000.
Run the example above. LinkedIn: 60 divided by 3,000 is 0.02, times 1,000 gives an RPM of $20. TikTok: 50 divided by 50,000 is 0.001, times 1,000 gives an RPM of $1. LinkedIn is twenty times more valuable per unit of reach on this comparison.
The arithmetic is trivial. What makes RPM hard is the numerator. Almost nobody knows revenue by platform, because platform analytics report impressions and engagement while payment processors report money, and nothing joins them. Getting the numerator is the actual work; the division takes ten seconds.
Why RPM changes decisions when engagement rate does not
Engagement rate answers a question about attention. RPM answers a question about money. Those two answers diverge more often than most people expect, and when they diverge, RPM is the one that should drive your calendar.
Short-form video platforms are the clearest case. They are exceptionally good at distributing content to people who are entertained by it and exceptionally bad at distributing it to people who will buy something. A high-engagement, low-RPM platform is not a failure. It is a brand channel, and it should be judged and resourced as one.
The reverse case is just as common. Niche communities, developer platforms, and professional networks routinely deliver small numbers that convert at rates that would look like errors on a consumer platform. If you judge them by reach they look like a waste of time. If you judge them by RPM they are often your best channel by a wide margin.
How to actually compute it
You need three inputs per platform, over the same time window.
- Impressions. Every major platform reports this natively, though the definitions differ, which matters and is covered below.
- Attributed revenue. This is the hard part. You need tracked links per post and a payment processor connection that carries the tag through to the payment.
- A consistent window. Thirty days is a reasonable default. Use the same window on both sides or the ratio is meaningless.
If your attribution is set up so revenue lands against individual posts, platform-level RPM is a rollup you get for free. This is the specific reason seenpaid connects to Stripe read-only rather than stopping at scheduling: without the revenue side, RPM is uncomputable, and impressions on their own cannot rank anything.
The impression-definition problem
Be honest about this one, because it is the biggest weakness of cross-platform RPM. An impression does not mean the same thing on every network.
A TikTok view can be counted almost immediately after the video starts. An X impression counts when the post enters the viewport. LinkedIn counts impressions on scroll past. Some platforms count repeat views from the same person, some deduplicate. These are genuinely different units wearing the same label.
This means cross-platform RPM comparisons are directional, not exact. A platform with 20x the RPM of another is almost certainly better for you. A platform with 1.2x is within the noise created by definitional differences. Do not make decisions on small gaps.
Where RPM is much more reliable is within a single platform over time. Comparing your own LinkedIn RPM in March against your LinkedIn RPM in June uses one consistent definition on both sides. That comparison is clean, and it is the one that should drive your iteration.
Other denominators worth trying
RPM is one member of a family. Swapping the denominator answers a different question, and the other members are often more actionable.
- Revenue per click. Removes the impression-definition problem entirely, since a click is a click everywhere. It measures how well your landing page and offer convert, rather than how well the post travels.
- Revenue per follower. Useful for judging whether audience growth on a platform is worth pursuing at all.
- Revenue per hour of production. Ranks formats rather than platforms. A carousel that takes ninety minutes and a text post that takes six can have identical revenue and radically different returns on effort.
Revenue per hour is the one most people should compute and almost nobody does. Your time is the genuinely scarce input, and it is the only denominator that reflects it.
What to do with a low-RPM platform
The instinct is to cut it. Sometimes that is right, but check three things first, because low RPM has more than one cause.
Check whether you are giving the platform a way to convert. Instagram suppresses outbound links in captions. If your only path is a bio link, low RPM may reflect a missing route rather than a bad audience. Check whether your offer matches the audience. A $2,000 consulting package will show a terrible RPM on a platform full of students, and that is a targeting problem, not a platform problem. Check whether the platform is doing top-of-funnel work that last-click attribution cannot see, which is the most common reason a genuinely valuable channel scores badly.
If none of those explain it, and the platform has had a fair sample of posts over a couple of months, then cutting it is a reasonable call. Reallocating those hours to your highest-RPM channel is usually the single highest-leverage change a small team can make.
Start here
Pick your last thirty days. For each platform you post on, write down total impressions and total attributed revenue. Do the division. Rank the list.
If you cannot fill in the revenue column, that is the problem to solve first, and it is a tracking problem rather than a content problem. Everything else in this article depends on having a numerator. Once you have one, RPM takes ten minutes a month and will quietly reshape where you spend your week.