What 229 affiliate programs actually pay
Commission rate is the number every directory sorts on and it explains almost nothing. Here is the arithmetic across every program we publish, with the correlation calculated.
LinkApprove Partnerships Team · Program research and advertiser relations
· 3 min read
Every affiliate directory on the internet, including the parts of ours that let you sort, puts commission rate in the biggest type. It is the number brands lead with, the number creators compare, and the number that decides which programs get promoted.
It is also, on our own published data, almost unrelated to what a program pays you.
This is not an opinion. Every figure below is computed from the 229 program pages on this site, each of which publishes its commission rate, cookie window and average order value. You can check any of them by opening the page.
The shape of the directory
Before the argument, the ground it stands on.
- Programs published
- 229Programs published
- Pay a percentage
- 200Pay a percentage
- Pay a flat bounty
- 29Pay a flat bounty
- Publish an order value
- 228Publish an order value
The median commission rate is 6%. The median average order value is $145. The median cookie window is 30 days.
Those three medians describe a program that does not exist. Half the directory sits well away from them in at least one dimension, and the distance is where all the money is.
Rate and payout are different questions
Here are the eight highest-rate programs in the directory, with what a single average-sized sale on each actually pays.
| Program | Rate | Avg order | Per sale |
|---|---|---|---|
| Tripadvisor | 50% | $240 | $120.00 |
| Il Makiage | 12% | $68 | $8.16 |
| Gymshark | 10% | $85 | $8.50 |
| Brooklinen | 10% | $195 | $19.50 |
| Glossier | 10% | $65 | $6.50 |
| Drunk Elephant | 10% | $90 | $9.00 |
| SkinCeuticals | 10% | $170 | $17.00 |
| Sunday Riley | 10% | $85 | $8.50 |
Now the same directory sorted by what a sale is worth.
| Program | Rate | Avg order | Per sale |
|---|---|---|---|
| Tripadvisor | 50% | $240 | $120.00 |
| Blue Nile | 5% | $2,400 | $120.00 |
| James Allen | 5% | $2,400 | $120.00 |
| Tempur-Pedic | 5% | $2,100 | $105.00 |
| Hydrow | 5% | $1,900 | $95.00 |
| Helix Sleep | 7% | $1,250 | $87.50 |
| Burrow | 6% | $1,450 | $87.00 |
| Brilliant Earth | 4% | $2,100 | $84.00 |
Ignore Tripadvisor, which appears in both lists for the reason given above, and the two tables share nothing. Not one of the ten-percent beauty programs appears in the payout list. Every entry in the payout list except one advertises a rate at or below the median.
Blue Nile pays 5% and returns $120 a sale. Glossier pays twice the rate and returns $6.50. That is an eighteen-fold difference in the opposite direction to the ranking.
By category, the same story
median 6% on a $330 order
median 5.5% on a $345 order
median 5% on a $408 order
median 4% on a $350 order
median 8% on a $165 order
median 6% on a $165 order
median 7% on a $135 order
median 7% on a $130 order
median 6% on a $133 order
median 8% on an $85 order
median 6% on a $100 order
median 4% on an $85 order
Median values across the 200 percentage-rate programs in the directory. Bounty programs are excluded. Pets and food are small samples and should be read as indicative rather than settled.
Beauty and skincare pays the joint-highest median rate in the directory at 8%, and the second-lowest amount per sale at $6.65. Travel pays the lowest median rate at 4% and returns more than twice as much per sale.
If you selected programs on rate alone, you would systematically choose the categories that pay least.
The number nobody publishes
There is a third term, and it is the one that decides whether either of the first two matters.
The 30-day window is the industry default and three quarters of the directory sits in that band. The tail is where the risk is.
Forty-five programs, close to one in five, run a window of fourteen days or fewer. On those, a reader who clicks during research and buys a fortnight later pays you nothing. The two programs at 24 hours or less are effectively same-session only.
A short window is not automatically bad. It is bad when it is paired with a purchase people deliberate over, which is precisely where the high-value programs live. Best Buy pays 1% on a one-day window against a $210 average order, and electronics buyers routinely research for weeks.
What to do with this
Three practical consequences, in the order they change your earnings.
Sort by rate multiplied by order value, not by rate. Both numbers are on every program page here. The arithmetic is one multiplication and it reorders the directory completely.
Read the cookie window before you write anything long-form. Evergreen content on a seven-day window is a gift to whoever publishes the review that ranks above yours.
Stop treating a low rate as an insult. A 1% rate in consumer electronics is not a bad deal, it is what retailer margins in that category support. The relevant question is what 1% of the basket is, and in that category the basket is large.
None of this requires data we hold privately. It is all on the program pages, published before you sign up rather than after, which is the only reason this analysis was possible to write at all.
Programs mentioned
LinkApprove Partnerships Team · Program research and advertiser relations
The team that reconciles commission terms against live advertiser feeds, reviews partner applications and writes the program notes in the directory. Collective byline for reference material rather than opinion.
Keep reading
Related posts
What changed: follower minimum and $100 threshold
Two changes to how LinkApprove works, why each was made, and who is worse off. Written the way we would want to read it if it were someone else's platform.
Moosa · · 3 min
Spotlight: Herman Miller and the $1,180 basket
A 5% rate that pays more per sale than almost anything in beauty, on a purchase that takes weeks to decide. What converts on it, what does not, and who should not bother.
LinkApprove Partnerships Team · · 3 min
Sub-IDs, or how to know which post actually earned
Untagged affiliate traffic cannot be broken down after the fact. A sub-ID naming scheme that still reads clearly in six months, what to keep out of one, and the things sub-IDs do not fix.
Moosa · · 6 min
