Unit Session %
Whether the people landing on your Amazon listings actually buy.
What it means
Unit Session % is Units OrderedUnits OrderedThe volume side of demand, before value. divided by SessionsSessionsTraffic to your listings; only lifts sales if conversion holds. on your Amazon listings, shown as a percentage. Sessions are product-page visits; units are the individual items ordered, so one order of 3 units counts as 3. That's why it reads higher than any order-based conversion rate, and why a catalogue sold in multipacks reads higher than one sold singly. Refunds aren't taken out.
Show the math
Formula and a worked example
Units Ordered is every individual item ordered in the period, before refunds.
Sessions is the product-page visits Amazon reports for your listings — traffic that reached the page, from ads and from search alike.
Worked example. A month brings 10,000 sessions and 900 units ordered. Unit Session % = 900 ÷ 10,000 = 9%. Past the 6% line that needs work, short of the 12% that reads healthy.
Now split those units: 600 came one at a time and 300 came from 100 three-packs. That's 700 orders, so an order-based rate on the same month reads 7%. Same shoppers, two points apart. Never compare this figure to a conversion rate from another tool without checking which one it counts.
It answers the question
Of everyone who reached the listing, how many left with something? Traffic is the half you pay for and conversion is the half you don't, so this is where the cheapest growth in an Amazon account usually hides.
Why it matters
It separates a listing problem from a demand problem. Sessions falling is a traffic story you fix with ads and ranking; sessions holding while this falls is a page story you fix with images, price and reviews. They cost completely different amounts to solve.
It also multiplies. Every point here applies to traffic you've already bought, so lifting a listing from 8% to 10% is a quarter more units on exactly the same spend.
What good looks like
These bands read across a whole account. A catalogue sold in multipacks sits higher by construction, and a considered, high-price product sits lower without anything being wrong — so compare a listing to its own history before you compare it to the band. One thing the number can't tell you: on days you don't hold the Buy Box, the traffic still arrives and the units go elsewhere.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Stop buying the wrong clicks | Pull the search-term report and add negatives for terms that don't match what the listing sells | Sessions fall, Unit Session % rises | 1–2 weeks | The rate improved because the denominator shrank. Units can stay flat or fall with it — check Units Ordered, not only the percentage. |
| Fast Close the price gap | Compare your landed price to the competing offers and add a coupon on the listings with the lowest rate | More units on the same sessions | 1 week | Every point of price comes out of margin. ACOS can improve while what each unit leaves you falls, and the second number is the one that pays the bills. |
| Slow Rebuild the top of the listing | New main image, three bullets that answer the objection, A+ content underneath | Same traffic converts better | 3–6 weeks | It's tempting to change everything at once, which makes the result impossible to attribute. Price and stock move underneath you meanwhile, so give it weeks before you judge it. |
| Slow Earn more reviews | Work the review request flow on every delivered order, and fix the product complaints behind the 1-star ones | A durable lift across the whole listing | 1 quarter | Nothing moves for weeks, and review programmes cost either product or a fee on every unit enrolled. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
A conversion rate and the traffic under it move in opposite directions all the time. Reading one without the other is how a shrinking account looks like an improving one.
Working at both ends
More people arrived and more of them bought. Usually a listing change landing at the same time as a ranking gain, and the rarest of the four.
Traffic you didn't want
Visits grew and the share buying fell. Broad keywords and a widened audience do this: you're paying for people the listing was never going to convert.
Smaller, better traffic
You cut the clicks that never converted and the rate rose. A real improvement — as long as the units came with it.
The real alarm
Fewer people and fewer of them buying. This is rarely a bidding problem — look for a stockout, a lost Buy Box, a price change, or a review that landed badly.
A listing you don't own the Add to Cart button on can't convert for you, however good the page is. Unit Session % falling while Buy Box % slips is not a listing problem at all — the visits arrived and the order went to another seller. Buy Box % steady with conversion falling puts it back on the page. Neither number separates those two alone.
Ads buy clicks at roughly the same price whatever the listing does, so conversion is what turns those clicks into sales. Unit Session % down with no campaign change is the usual reason ACOS climbs out of nowhere: same spend, same traffic, fewer orders at the end of it. Fix the listing and ACOS recovers without touching a bid.
Common misreads
It counts units, not buyers. Multipacks and multi-item orders push it above the share of people who actually converted, and the gap is widest on exactly the listings you'd want to read most carefully.
The denominator moves too. A burst of broad ad traffic, a deal badge, or an off-Amazon promotion adds sessions that were never going to buy, and the page they landed on is unchanged.
It says nothing about volume or margin. A 20% rate on 200 sessions is 40 units — an excellent percentage on a listing nobody can find. Read it beside Sessions every time.
Also called
Unit session percentage · Amazon conversion rate · session conversion rate · unit session rate
See yoursYour Unit Session % for the period, and the same figure per SKU on SKU Performance.
Open Amazon Summary →