LinkToLooks · Guides

Which creator analytics actually predict revenue?

Part of: For creators

The chain, and where it actually narrows

Between somebody seeing your post and somebody buying, there are five drops:

The drop between any two of those steps is usually far larger than the variance in the step above it. Doubling your reach and keeping every rate the same doubles revenue; doubling your click-to-order rate does the same thing for free. Since reach is largely rented from a platform and rates are yours to change, the rates are the ones to watch.

The six metrics that move with revenue

1. Link clicks per 1,000 reached

• The first metric downstream of intent, and the one that most cleanly separates content that sells from content that performs.
• Normalising by reach rather than by followers is what makes it comparable between a post that travelled and one that did not.
• Track it per format. Static posts, short video and stories usually behave nothing alike.

2. Click-to-order rate

• Orders divided by clicks, from the affiliate dashboard. This is where product choice shows up: the same audience converts very differently on a $20 accessory and a $300 appliance.
• It is also where broken links and out-of-stock listings appear first, as a rate that drops without any change in your content.

3. Earnings per click

• Commission divided by clicks. It folds order value and commission rate into one number, which is what makes it the fairest way to compare two categories.
• Because it is a per-click figure, it stays meaningful in a bad reach week, when totals do not.

4. Saves and shares

• A save is the closest free proxy for purchase intent: people bookmark things they are not ready to buy yet.
• Shares extend reach to people outside your follower graph, which is where new buyers come from rather than repeat ones.
• Both are frequently unavailable on some formats. Record that as unavailable rather than entering zero, which quietly understates you.

5. Returning-audience share

• The proportion of engagement coming from people who have engaged before. It predicts whether recommendation carries weight, which is the thing brands are really buying.
• Reported differently on every platform and sometimes not at all, so treat it as a trend, not a level.

6. Reversal rate

• The share of commission clawed back after returns and cancellations. It is reported late and it is the number most often left out of a creator’s own accounting.
• It varies enormously by category. Apparel and footwear return heavily; media and hardware barely at all.

The four that mostly do not

• Follower count. It sets a ceiling on reach and nothing else. It is also the denominator that makes engagement rate fall as an account grows, which is arithmetic rather than decline.
• Impressions. Counts showings, not people, and inflates with format changes you did not make.
• Likes. The cheapest possible action, and the one most decoupled from spending money.
• Engagement rate quoted with no formula. The same post can be honestly reported at several different rates depending on the denominator chosen, so a bare percentage is a selection rather than a measurement.

None of these are useless — they are how brands shortlist, so they still get you paid. They just do not tell you which of your own posts to make more of.

What quietly breaks the count

• Attribution windows. A purchase outside the window is a purchase you made and did not get paid for. It does not appear anywhere as a loss, which is what makes it easy to miss.
• Dead links. A marketplace listing is one-of-one and stops earning the moment it sells. Clicks keep counting, orders stop, and the click-to-order rate falls for reasons that have nothing to do with the content.
• Out-of-stock listings. A live link to a sold-out product converts at zero while still absorbing clicks. We have measured how common that is across feeds, and it is not rare.
• In-app browsers. Clicks that open inside a social app behave differently from ones that open in the system browser, and neither dashboard tells you which you got.
• Multiple links in one post. Without separate tracking per link you can measure that a post worked, but not which product did the work.

A weekly dashboard that fits on one line

Six numbers, once a week, in a spreadsheet you own rather than in a platform you do not:

1. Reach, all formats.
2. Tracked link clicks.
3. Clicks per 1,000 reached.
4. Orders.
5. Click-to-order rate.
6. Earnings per click, with reversals subtracted once they land.

Keep it for twelve weeks before drawing any conclusion. Weekly creator numbers are noisy enough that a single good post can look like a trend, and the most common analytics mistake we see is changing strategy on one week of data.

Last verified 1 September 2026 against no new primary source. This page describes measurement practice and cites no platform benchmark, conversion rate or industry average

Some links here are affiliate links. LinkToLooks has earned $0 from them to date — no network has approved us yet — so nothing on this page is picked to hit a payout.

Frequently asked

Which creator metric predicts revenue best?

Earnings per click, because it folds order value and commission rate into one figure and stays comparable across categories and across weeks with very different reach.

Is follower count a vanity metric?

Partly. It sets a ceiling on reach and it is how brands shortlist, so it affects what you get paid. It does not tell you which of your posts to make more of, which is what analytics are for.

Why did my conversion rate drop without my content changing?

The usual causes are outside your content: a link that now points at a sold-out or delisted product, a marketplace listing that has been bought, or an attribution window that closed before the purchase.

How long should I collect data before changing strategy?

Twelve weeks. Weekly creator numbers are noisy enough that one strong post reads as a trend, and acting on that is the most common analytics mistake.

What is a good engagement rate?

We decline to publish one, because the same post can be honestly reported at several different rates depending on the denominator. Compute all four formulas on your own posts and compare yourself to yourself.

Sources and scope

• The four competing engagement formulas, worked longhand on one set of inputs, are on our engagement rate calculator.
• Out-of-stock rates in live affiliate feeds are measured in how many affiliate listings are actually in stock.
• Link decay and how to check for it: why affiliate links break.