The bars are the same for everyone, but they do not mean the same thing for an 8,000-follower account as for a 400,000 one. These are the three readings that matter before changing anything.
Up to 15,000 followers
Micro: the niche outweighs the network
With a small audience, switching network barely moves your rate: on every platform you sit in the low part of the table. What does move the price is the niche. On Instagram, Finance pays $23 CPM against $13 for Gaming: the difference is in the audience you have, not in the network you publish it on. Work on niche, format and engagement before considering a migration.
Before moving, compare what each network leaves per thousand impressions organically: a higher CPM with worse monetisation can cost you more.
Between 15,000 and 100,000
Mid: here the network moves money
This is the band where the network's CPM changes the year's revenue. In lifestyle, CPM runs from $8 on X to $28 on LinkedIn: 250% for the same audience and the same format. If your engagement sits above the network average, you are in a position to choose where you publish instead of accepting whatever is paid.
If the new network pays more but your audience takes a year to move, the maths only works while you publish on both.
Over 100,000
Large: the network is one line of the negotiation
With a large audience, CPM stops deciding and the package starts deciding: rights, exclusivity and formats. The network becomes one more line in the proposal, and the one with the highest CPM is not always the one that pays best overall: the same piece on two networks is two deliverables, not a bonus.
At this size you are not negotiating a migration, you are negotiating multi-platform use. They are different prices and should appear on separate lines.
Switching costs more than the table suggests: your audience does not move with you, and rebuilding the same reach takes six months to a year of steady publishing. Anyone comparing CPM and skipping that part is comparing a number with a year of work.
The switch pays when three things hold at once: the destination CPM beats your current one by at least 40%, the niche pays more there —not just the network— and you already publish on the new platform without extra effort. If one is missing, the sensible move is to add the second network and compare for three months with your own data, not the catalogue average.
What does not work is migrating for a one-off campaign. If the high CPM depends on a seasonal advertiser, next year is paid with the audience you left behind.