The Honesty Page
Every other dashboard prints a confident number for things it cannot actually measure. We don’t. When a source can’t give us the truth, we render an em dash — and link it to a named reason. Here is every one of them.
How to read every number we show
Nothing here is a bare figure. Each number carries one of four provenance badges, so you always know how much to trust it — and the same badges appear on every screen in the product.
Straight from a system of record we own — the double-entry ledger, entitlements, lesson progress. The closest thing to a fact.
Computed from those owned truths by a formula we publish — a link click traced through a contact to a balanced journal.
An estimate or a forecast. Useful, but a guess — so we badge it a guess and never let it decide money on its own.
The source cannot give us this, so we render an em dash and send you here. We would rather show nothing than a fabricated number.
Straight from a system of record we keep — the ledger, entitlements, lesson progress. Even here, we tell you where the platform’s own conventions bend.
Reach is a daily estimate, so we never add it up.
Meta returns reach as an approximate daily figure and says so in its own docs. Summing seven days is not a weekly reach — the same account gets counted many times over. Any “total reach” number is arithmetic on sand.
What we show instead
We show reach per day, badged estimated, and refuse to print a summed total that would only mislead you.
#ig_reach
Courses
Honest completion is low single digits. We won’t flatter it.
Measured honestly — a learner finishing every lesson — completion for self-paced online courses runs in the low single digits industry-wide. A vendor quoting 40% has quietly redefined “complete” (one lesson opened, a video 30% watched) to sell you a nicer chart.
What we show instead
We report completion from actual lesson-progress records and define the bar in the open, even when the number stings.
#course_completion
Computed from truths we own by a formula we publish. Trustworthy, with the one honest caveat noted.
Attribution
We under-credit ourselves on purpose.
We can prove the sales that ran through a link we minted — click, contact, order, balanced journal, with real ids. We cannot prove a sale from someone who saw a post and bought later without clicking; measuring that view-through needs a paid ad pixel we don’t run. So our attributed revenue is a floor, not a ceiling.
What we show instead
We credit only what traces to a journal id and say plainly: your true organic impact is at least this — and probably more.
#organic_attribution
We show these, but they’re estimates or forecasts — so we badge them a guess and never let them decide your money on their own.
An “open” is a guess, so we label it one.
Apple Mail Privacy Protection pre-fetches every image the instant a message lands — whether or not a human ever opens it — which trips the tracking pixel for a large, unknowable share of your list. The “open rate” every email tool prints is therefore a model, not a count.
What we show instead
We badge open rate modeled and lean on clicks and replies — actions a machine can’t fake — for anything that decides money.
#email_open_rate
Analytics
A prediction without a range is a fib.
Predicted LTV forecasts behavior that hasn’t happened yet. A lone dollar figure hides how wide the uncertainty is — two creators with the same point estimate can have completely different realistic outcomes.
What we show instead
We always render predicted LTV as a band with its confidence, badged modeled — never as a single confident number.
#predicted_ltv
The em dash. These sources cannot give us the truth, so we render — and link you straight here. We would rather show nothing than a fabricated number.
Before you connected, the number is simply gone.
Meta deprecated the historical impressions endpoints and does not backfill them. Impressions from before you linked the account are stored nowhere we can reach — they are unrecoverable, not merely un-fetched.
What we show instead
We render — for the pre-connection window and start counting cleanly from the day you connect, instead of back-filling a fiction.
#ig_historical_impressions
TikTok
TikTok’s API doesn’t expose it, so we don’t invent it.
The official TikTok API returns neither average watch time nor audience demographics for your content. There is no compliant way to get them — a vendor showing you these is scraping, extrapolating, or guessing.
What we show instead
We render — and name exactly which door TikTok closed, rather than filling the gap with a number we made up.
#tiktok_watch_time
Instagram never exposes this. We will never ship it.
“Who viewed your profile” has never existed in any Instagram API. Every product that claims to show it is fabricating names outright. We treat it as permanently unavailable.
What we show instead
Nothing — we won’t build a feature whose only possible implementation is a lie.
#profile_views
All platforms
A “view” means five different things — one total means nothing.
A three-second autoplay on one network, a click-to-play on another, and a bare impression on a third are all called a “view” under incompatible definitions. Adding them yields a big number that measures nothing in particular.
What we show instead
We keep each platform’s number in its own units and footnote the definition, instead of summing things that aren’t the same thing.
#cross_platform_views
Automations
No holdout, no number.
“This automation made $4,000” is only true if you know what those contacts would have done untouched. Without a randomized holdout, revenue that would have arrived anyway gets credited to the automation. The effect is unmeasurable, not merely unmeasured.
What we show instead
Every automation carries a holdout by default; its card shows incremental lift with a confidence interval — or an honest — until the data earns a number. We report lift or nothing.
#automation_lift
Why this page exists
We keep the books — a real double-entry ledger under every sale, schema-registered events behind every action. That’s what lets us prove “this post produced $47, here are the journal ids.” It’s also what forces this page: once you can prove a number, you can no longer hide the ones you can’t. Every limitation above is a place a competitor prints a confident figure and we refuse to.
Caveat registry v1.0.0 · updated 2026-07-08 · 10 named reasons.
This page is generated from the same versioned registry the product renders from, so it can’t drift from what you actually see in-app. When the registry changes, this page changes with it.
