How it works
Every number on the screen can be traced back to where it came from.
Most social analytics tools are confident. Very few will tell you what their confidence is built on. This page is the method.
Connecting an account
You sign in to LinkedIn, Instagram or TikTok through their own login screen. Hype Machine never sees your password, and the access it receives is read-only: it can read your posts and their engagement, and it cannot publish, delete, message, or follow.
You can disconnect an account at any time, from inside Hype Machine or from the provider’s own settings. Disconnecting stops all future reads immediately.
The cohort
A post is only ever compared to posts it could reasonably be compared to. That set — we call it the cohort — is one project, one platform, one side of the owned/competitor line.
This sounds obvious and almost nothing does it. A tool that ranks your “top posts” across LinkedIn and TikTok together is ranking two completely different engagement economies against each other, and the platform with the looser definition of a “like” wins every time.
How a post is ranked
Each post gets an engagement score, and each score is placed against the distribution of every other score in its cohort. The tiers are quartiles, not fixed thresholds: your best posts are measured against your own channel, not against a number someone picked for a blog post about industry benchmarks.
Below a minimum number of posts, the cohort is unranked. A handful of posts is not a distribution, and a badge drawn from one is a guess wearing a uniform.
Where the numbers come from
Different ways of reading a platform return different metrics. An account you have connected exposes more than a public profile page does — and what is available differs per platform on top of that.
So every stored metric records its source, and every comparison is computed over the metrics that all participating sources can actually report. If one side cannot report shares, shares are dropped from both sides rather than counted as zero for one of them.
Why this matters more than it sounds. Treating an unreadable metric as zero does not produce a slightly pessimistic number. It produces a number that is wrong in a direction that depends on which tool read which account — and it is invisible, because zero looks like data.
Competitor accounts
You can add an account you don’t own by pasting its public profile URL. Hype Machine reads what is publicly visible on that profile, on the same platform as the channel you’re comparing it to, and applies the same rules: same platform only, shared metrics only.
This is the part that costs money, because reading a public profile costs us money per request.
Analysis and ideas
Once there is enough history, Hype Machine can tell you what is working on a channel and why, and suggest what to post next based on what has actually performed. That analysis is generated, it costs money to run, and it is always optional.