Amazon's 75-character title limit: what we see in the data

Amazon's 75-character title rule took effect July 27, 2026. Most of what has been written about it explains the policy. This page reports what is actually happening to real listings, because we snapshot per-ASIN product titles daily and can therefore date each rewrite instead of guessing.

Last updated: August 26, 2026

Updated August 26, four weeks after the deadline. Two things changed. The measurement below now covers roughly 22,000 top-selling US listings. First, Amazon's rollout has effectively stopped with more than half of tracked listings still over the limit — compliance moved about a tenth of a point per day over the last twelve days, against half a point a day before that. Second, the click-through null hides a large amount of movement: the median listing barely moved, but 63% of rewritten listings moved more than 20%, and what separates the winners from the losers is how much the rewrite removed. See "What the performance data says".

We also said in early August that a conversion and ACOS read was coming mid-August. It is not, and the reason is now clear rather than a matter of waiting. That is explained below rather than quietly dropped.

If you checked your account and saw nothing, that is the finding

Most sellers looked at their numbers after July 27, saw nothing obviously broken, and moved on. So did we. Across roughly 22,000 listings the average effect on click-through came out at about zero.

Then we split the same listings by how much each rewrite had actually removed. A third of them had lost more than a fifth of their clicks. Another quarter had gained about as much. The average is flat because those two groups cancel out, and the average is what your account-level reporting shows you.

The ones that got hurt are identifiable, though, and they are not the ones you would expect. Three things worth doing this week:

1. Work out which of your titles Amazon rewrote. Seven in ten seller feeds still hold the old long title, so your own listing data won't tell you — you have to compare it against what Amazon displays.
2. Start with the titles that were 90–130 characters. Not your longest ones; those came through nearly unharmed. It is the middling ones that lost clicks.
3. Move what got cut into Item Highlights. 125 searchable characters, free, and the median brand is using them on 19% of listings.

The rest of this page is the evidence, including the parts that didn't work out.

Summary of what we found

Amazon's 75-character title rule took effect on 27 July 2026. Four weeks later, 53.7% of the top-selling US listings we track still displayed a title over the limit, and enforcement had slowed roughly six-fold. Amazon made about 60% of the shortenings and sellers made 40%. Shortening a title had no average effect on click-through, but 63% of rewritten listings moved more than 20% in one direction or the other, and how much text the rewrite removed predicted which way.

The numbers in one place

Measured 24 August 2026 across roughly 22,000 top-selling US listings, snapshotted daily since before the deadline. Figures are shares, medians and distributions; segment figures are medians across brands.

Measure	Figure
Listings still displaying a title over 75 characters	53.7%
Median displayed title length	90 characters (129 at the June baseline)
Of listings we can compare: cut to compliant	38.7%
Of listings we can compare: never changed, still over	50.8%
Enforcement pace, 12–24 August	~0.09 points/day (~0.56 earlier in August)
Title changes that landed before the 27 July deadline	65%
Median rewritten title length	71 characters
Median cut	76 characters and 9 words
Shortenings made by Amazon vs by the seller	60% / 40%
Seller feeds still holding an over-limit title	70%
Listings with an Item Highlight set	41% (median brand: 19%)
Rewritten listings whose click-through moved more than 20%	63%
Click-through change where more than 100 characters were removed	−1.6%
Click-through change where 26–50 characters were removed	−20%
Click-through change, difference vs never-changed listings	−0.3 points (a null)

The rule, briefly

• Titles must be ≤ 75 characters, including spaces and the brand name.
• Every category except media (books, music, video).
• Over-limit titles get rewritten by Amazon, with its own model, gradually and on Amazon's schedule — not all at once on the deadline.
• Brand-registered sellers get 14 days in Review Listings Changes to review, edit or approve a rewrite before it goes live.
• Item Highlights (title_differentiation) is a 125-character searchable field that renders below the title — but only when the title is under 75 characters.

That last point is the one most easily missed: the limit and the new field are a package. Shortening a title without filling in the Item Highlight means the keywords you dropped are simply gone.

How many Amazon titles are still over 75 characters?

Four weeks after the deadline, 53.7% of the listings we track still display a title over 75 characters. Among listings we can compare against a pre-wave baseline:

What happened to the listing	Share
Shortened to compliant (≤75)	38.7%
Never changed, still over the limit	50.8%
Never changed, already compliant	7.9%
Got longer	1.2%
Shortened but still over the limit	1.0%
Reworded at the same length	0.4%

Median title length across the cohort has moved from 129 characters at the June baseline to 90 today — real movement, but nowhere near a finished rollout.

Has Amazon stopped enforcing the 75-character limit?

This is the most significant change since our last update. Enforcement is not grinding slowly toward completion; it has largely stopped:

Period	Over-limit share fell by
August 4 → August 12	about 0.56 points per day
August 12 → August 24	about 0.09 points per day

That is a six-fold deceleration. The daily change counts say the same thing from another direction: we recorded 200–400 listings changing per day through late July, and 6–40 per day across the second half of August. The cohort grew over that period, so the later days are measured on a larger panel and still show an order of magnitude fewer changes — this is a real slowdown, not a shrinking sample.

What this means: if your titles are still over the limit, the evidence no longer says "your turn is coming shortly." It says the wave that was going to reach you may already have passed. That is not permission to ignore the rule — Amazon can resume at any time, and a rewrite still arrives without warning — but it does mean you have more time to write your own version than the July messaging implied.

The category spread is the striking part. These are categories where at least five brands in our cohort contribute, so they are directional rather than precise:

Category	Median title before → now	Over 75 before → now	Listings changed
Hair Shampoo	164 → 62	94% → 12%	82%
Boys' Costume Accessories	178 → 73	100% → 17%	75%
Hair Conditioner	171 → 66	93% → 24%	65%
Body Lotions	177 → 60	100% → 12%	62%
Facial Serums	179 → 68	100% → 22%	59%
Detox & Cleanse Weight Loss	180 → 83	100% → 53%	47%
Collagen Supplements	156 → 98	90% → 64%	36%
Dog Treats	106 → 99	90% → 53%	35%
Facial Cleansing Washes	166 → 74	88% → 47%	33%
Kids' Costumes	176 → 72	83% → 33%	33%
Face Moisturizers	189 → 91	100% → 50%	23%

A caveat, and it is a real one: read that table as directional only. Categories in our cohort are thin, and a single large seller can carry a whole row — when we checked in August, one brand supplied 91% of a category's listings in the worst case, and that row was still being published. So the low-movement rows in particular may describe one seller who hasn't been touched rather than a category Amazon is leaving alone. We stopped quoting category-vs-category contrasts because of it. The rows above now have to clear two tests: at least five distinct brands, and no single brand supplying more than half the listings. Most of Amazon's browse categories are too granular to clear both, which is why so few appear here.

What to do with this: if your titles look untouched today, do not read that as an exemption. It means you still have the chance to write the 75-character version yourself, and a queue position you cannot see.

How long are the titles Amazon rewrites?

Where titles were cut, they cluster tightly just under the cap. The median rewritten title is 71 characters, with the middle half of them between 65 and 74. The typical cut removes 76 characters and 9 words. Roughly a quarter dropped a number (a size, count or pack quantity) and one in eight dropped a bracketed or parenthetical segment. These figures barely moved as the cohort grew, which is what you would expect from a mechanical rule.

A person rewriting a title for clarity does not reliably stop nought to five characters short of the limit. That distribution is a machine truncating to fit.

What to do with this: the rewrite optimises for length, not for the keywords that earn your clicks. A 75-character title you wrote will keep different words than a 75-character title Amazon cut, and only one of those was chosen with your search-term report in hand.

When did Amazon actually start shortening titles?

65% of all the title changes we recorded landed before July 27, not after it. The wave began weeks earlier and peaked in the final week of the run-up. That share was 81% on July 30 and has drifted down as post-deadline changes accumulated — but the bulk of it still predates the date itself.

One thing we deliberately do not claim: who made those changes. A daily snapshot tells us a title changed, not who changed it. A pre-deadline change is at least as likely to be a seller getting ahead of the rule as Amazon acting early, and we have no way to separate the two per change. The measurement point below holds either way.

What to do with this: if you are measuring the impact of this rule with a clean before-July-27 / after-July-27 split, you are putting about two-thirds of the changes in the wrong bucket and will read the result as "no effect." Impact has to be measured against each listing's own change date — which is exactly how the measurement on this page works.

Which words does Amazon delete when it shortens a title?

Across the cohort, these are the words most often present in a title before the rewrite and gone after it. Only words dropped by at least five different brands are listed, and brand names are excluded:

Words dropped	Brands affected
black	22
premium	20
men	19
free, women	18
white	17
all, blue, gift, kids	16
green, made	15
box, home, natural	14
adjustable, gifts, medium, size	13

They fall into three clusters:

1. Variation attributes — black, white, blue, large, medium. These are the words shoppers use to pick between your own variants. Losing them from the title is survivable only if your variation data and Item Highlights carry them.
2. Audience terms — men, women, kids. Often the single most important qualifier a shopper searches with, and these are being deleted.
3. Puffery — premium, easy, extra, free. Losing these costs you very little, and it is worth noticing that the machine is not distinguishing between them and the two clusters above.

What to do with this: before Amazon rewrites a title, check whether the colour, size and audience words in it are searchable anywhere else on your listing. If they only exist in the title, a rewrite takes them out of your keyword coverage entirely.

Why does my Seller Central title differ from the one Amazon shows?

Compare the title you submitted with the title Amazon displays and they frequently disagree. 70% of seller feeds still hold a title over 75 characters. For 20% of listings, a short displayed title is sitting on top of a long submitted one — Amazon rewrote it and the seller's own listing data was never updated to match. Your feed and your storefront are telling different stories.

Who actually shortened the title: Amazon 60%, sellers 40%

In our last update we said this comparison could not attribute a shortening to either party, because "compliant on both sides" mixes the seller fixed it with it was always short. It can now, by restricting the comparison to listings we watched cross the limit — listings that were over 75 characters at the pre-wave baseline and are compliant today. For those, a feed still holding the long title means the edit was not the seller's.

Across 4,466 listings observed making that crossing:

Who made the edit	Share
Amazon — feed still over the limit	59.9%
The seller — feed compliant too	40.2%

So Amazon did the majority of this work, but two in five of these listings were rewritten by the seller themselves. "Amazon rewrote everyone's titles" is wrong as a blanket statement, and so is the opposite.

Honest limit: displayed-vs-submitted is a strong signal, not an audit trail. Amazon normalises titles for other reasons too, and on a shared listing another contributor's copy can win.

How many sellers are using Amazon's Item Highlights field?

This is still the biggest open goal in the whole change, though less lopsided than it was. Across the listings we sampled, 41% had an Item Highlight set, and the median brand has now set them on 19% of its listings. That median was zero in early August, so this is the one number in the study still moving in a straight line — adoption is spreading across sellers rather than sitting with a handful of them.

The sellers who do use it use nearly all of it: a median of 113 of the 125 characters available.

What to do with this: if a title was cut from 190 characters to 71, roughly 78 characters of keywords left your listing. Item Highlights is a searchable field designed to hold exactly that, and it costs nothing but the time to write it. Right now it is the cheapest keyword real estate on Amazon precisely because so few people have claimed it.

Did shortening titles hurt click-through rate?

This is the question everyone is asking, and as of this update we can answer half of it.

Shortening a title had no measurable effect on click-through rate. Comparing each listing against a seven-day window on either side of its own change date:

Group	Listings	Median CTR change	Share that improved
Shortened to compliant	1,705	−6.6%	44%
Never touched, still over the limit	1,898	−6.3%	43%
Never touched, already compliant	341	−7.3%	46%
Got longer	64	−3.2%	48%

Every group drifted down together, including the ones nothing happened to. The gap between listings Amazon rewrote and listings it left alone is 0.3 of a percentage point. Across three separate runs that gap has been −0.6, +0.7 and −0.3 points — it changes sign and never clears a point, which is what a real null looks like. Whatever pushed click-through down was market-wide, not the title rule.

We had a different number a few days ago, and it did not survive. Our July 30 read showed a 2.5-point gap. That was an artifact: the rewritten group had five days of post-change data while the control group had only two. With both at a full seven days and three times the listings, the gap collapsed to noise.

A pattern we reported on July 30 that turned out to be wrong

That first read also showed shorter titles having higher click-through — a clean gradient from 0.89% at 61–75 characters down to 0.53% at 151–200. With the full cohort, that gradient disappears:

Title length	Listings (Jul 30 → Aug 4)	Median CTR (Jul 30 → Aug 4)
≤ 60 chars	89 → 499	0.75% → 0.54%
61–75	30 → 453	0.89% → 0.61%
76–100	87 → 449	0.66% → 0.65%
101–150	495 → 2,181	0.55% → 0.55%
151–200	1,290 → 2,797	0.53% → 0.60%

Click-through is essentially flat across every title length. The earlier gradient came from the short buckets being desperately thin — 30 and 89 listings — and a bug on our side that dropped large-catalog sellers from the sample. Both are fixed, the short buckets grew five- to fifteen-fold, and the pattern evaporated. We are calling that out rather than quietly editing it, because a number that moves this much on sample size was never a finding.

The null is real, and it is hiding a great deal of movement

A median near zero does not mean nothing happened to anyone. Among the listings Amazon shortened:

Movement in click-through	Share of rewritten listings
Gained more than 20%	27.6%
Lost more than 20%	35.4%
Moved more than 20% either way	63%
Moved more than 50% either way	29%

The 10th percentile is −58% and the 90th is +78%. The cohort median of −6.6% describes almost none of these listings; it is two large groups moving in opposite directions and cancelling out. If you checked your account after the rule and concluded it did not affect you, an account-level average is capable of telling you that while a third of your rewritten listings lost a fifth of their clicks.

Which listings lost the most clicks?

The obvious hypothesis is that it matters who wrote the replacement title. It does not. Comparing Amazon-written against seller-written short titles inside the same brand — which holds catalogue quality, bidding and imagery fixed — the difference is 0.1 of a point, and exactly half the brands favour each side. A coin flip.

What does predict the outcome is how much the rewrite removed:

Characters removed	Median brand's CTR change
More than 100	−1.6%
51–100	−4%
26–50	−20%

And the same pattern appears from the other direction, by how long the title was to begin with:

Starting title length	Median brand's CTR change
151–200	−3.8%
101–150	−5.2%
76–100	−14.6%

Two independent ways of slicing the same listings, and they agree. The deepest cuts were the safest ones. Impressions held roughly flat across every group, so these are genuine click-through moves rather than listings being shown less.

The explanation is straightforward once you see it. Cutting a 200-character title to 70 removes keyword stuffing that shoppers were not reading. Cutting a 110-character title to 70 removes the specifics that made the product distinguishable — and those are the listings that lost clicks.

What to do with this: the listings to review first are not your longest ones. They are the ones that were already moderately short, between roughly 90 and 130 characters, before Amazon cut them. Those titles had no padding left to lose.

These figures are medians across brands rather than across listings, so that a single large catalogue cannot drive them. Segments backed by fewer than five brands, fewer than thirty listings, or dominated by one seller are not reported.

---

Which of your listings are in the group that lost a fifth of its clicks? The 75-Character Title Audit dashboard lines your submitted titles up against what Amazon actually displays, then sorts the mismatches by what they are costing you. Connecting Seller Central and Amazon Ads is free. Connect your accounts.

---

What is still open — and why it is not simply a matter of waiting. In early August we said a conversion and ACOS read would arrive mid-August, once Amazon's 14-day attribution window closed on the July rewrites. That window has now closed: 97% of the listings in this measurement have fully settled sales data. We are still not publishing a conversion figure, and the reason is worth stating plainly rather than deferring again.

The difficulty is structural. Rewritten listings are measured around their own change date, while the control group is measured around the deadline — so the two groups' post-change windows sit on different calendar days and mature at different times. When we first looked, that made shortened listings appear to convert 6.5 points better than the control. Twelve days later the same comparison had decayed to 2.6 points, which is the behaviour of a measurement artifact rather than an effect. We built a filter that admits a listing only once its own window has fully settled, expecting it to resolve the question — it did not, because by late August nearly every listing already qualified.

What remains is a small apparent advantage for shortened listings that we cannot distinguish from noise: in the same data, a control group that nothing happened to posted a 52% swing in ad conversion rate. When a group with no treatment moves that far, a 2.6-point difference is not a finding. The honest summary is that shortening a title does not appear to have hurt conversion, and may have helped slightly, but the effect is not separable from the noise around it. We would rather say that than publish a number that reads as authoritative and is not.

What this means for you

Do not panic about the average click-through consequence of a shortened title — measured across thousands of listings, there isn't one. But do not stop at the average either: nearly two-thirds of rewritten listings moved by more than a fifth, and the ones that got hurt are identifiable in advance. The reason to write your own 75-character title is keyword control, not click-through defence: an automated cut keeps the words Amazon judges important, and those are not necessarily the ones your search-term report says earn your sales.

A checklist that follows from the data

1. Start with the 90–130 character titles, not the longest ones. This is the counterintuitive result above: listings cut from a moderate length lost roughly four times the click-through of listings cut from 150+. A very long title mostly loses padding. A moderately long one loses its specifics.
2. Find your over-limit titles and rewrite the important ones yourself. A rewrite may still come; the only question is whether it keeps your converting keywords or Amazon's guess at them. Prioritise by revenue, not by how far over the limit a title is.
3. Do not read the stall as an exemption. Enforcement has slowed to a near-stop, which buys you time rather than a pass. Amazon can resume without notice, and it does not announce your turn.
4. Fill in Item Highlights. 125 searchable characters, the median brand has them on only 19% of its listings, and it is the designated home for the keywords a cut removed. Still the highest-value, lowest-effort item here.
5. Reconcile your feed against what Amazon displays. Seven in ten feeds still hold the old long title. If Amazon has already cut one, your submitted copy is stale — update it so you control the version of record.
6. Watch Review Listings Changes if you are brand-registered. Fourteen days is a real window, and it closes quietly.
7. Do not judge the change on a July-27 split. Date each listing's change and measure from there.
8. Segment before you conclude nothing happened. An account-level average is the single easiest way to miss this entirely.

Method

• Source: daily per-ASIN displayed-title snapshots (Amazon's Catalog Items data) for each account's top-selling listings, plus sellers' own submitted titles, ad performance by advertised ASIN, and Seller Central sessions and units.
• Change detection: per listing, comparing the displayed title across consecutive daily snapshots, which yields a change date per listing rather than an aggregate before/after.
• Baseline: a pre-wave reference point, because most changes predate the deadline. Comparing to the day before July 27 would miss four-fifths of them.
• Alignment: pre/post windows anchored on each listing's own change date; never-changed listings anchored on the cutover as a control group.
• Scope: US, and each account's top-selling listings rather than its full catalog. The long tail, where over-limit titles almost certainly concentrate, is under-represented — so the true share of listings still over the limit is likely higher than what we report.
• Disclosure: aggregate and scale-free only — shares, medians and distributions across listings and brands. No individual seller's titles or figures appear here, and categories are reported only where several brands contribute.

How this page gets updated

• August 4 (done): click-through settled — a null result — and a larger cohort that overturned an earlier length-vs-clicks pattern. Both are written up above rather than quietly edited.
• August 26 (done): the rollout stalled; the click-through null turned out to hide a large amount of movement, and how much a rewrite removed turned out to predict it; who made the edit turned out not to.
• The conversion read we promised for mid-August: not delivered, and explained above rather than postponed again. The obstacle is structural, not a matter of waiting longer.
• September: a third compliance sweep, to establish whether the stall is permanent or a pause.
• Ongoing: Item Highlight adoption, the one number still climbing. Once enough listings have gained a highlight we can measure what the field is worth directly, by watching the same listing before and after it appears — which is the only clean way to answer that question.

Frequently asked questions

When did Amazon's 75-character title limit take effect?

July 27, 2026. From that date Amazon requires product titles in every category except media (books, music, video) to be 75 characters or fewer, including spaces and the brand name. Enforcement is not instant: Amazon rewrites over-limit titles gradually, on its own schedule, using its own model. In our daily catalog snapshots the changes arrive listing by listing across weeks rather than all at once, and 65% of the title changes we recorded landed before the deadline rather than after it. Four weeks on, enforcement has slowed to roughly a tenth of a percentage point per day, from more than half a point a day earlier in August.

What happens if my Amazon title is longer than 75 characters?

Amazon rewrites it for you with its own model, which keeps the keywords Amazon judges important rather than the ones you know convert. Brand-registered sellers get a 14-day window in the Review Listings Changes tool to review, edit or approve the rewrite before it goes live. The rewrites we measured land at a median of 71 characters, with the middle half falling between 65 and 74 — a machine cutting to fit, not an editorial rewrite. The typical cut removes 76 characters and 9 words.

What is Amazon's Item Highlights field?

Item Highlights (the title_differentiation attribute) is a 125-character searchable field Amazon introduced alongside the 75-character title limit. It renders below the product title in search results and on the detail page, but only when the item name is under 75 characters. It is where the secondary keywords, materials, sizes and use cases that no longer fit in the title are meant to live. Adoption is still low but rising: across the listings we sampled, 41% had one set, and the median brand has them on 19% of its listings, up from none at all in early August. Sellers who do use it use nearly all of it, at a median length of 113 of the 125 characters available.

Has Amazon already shortened everyone's titles?

No, and the rollout has largely stopped. Four weeks after the deadline, 53.7% of the roughly 22,000 top-selling listings we track still displayed a title over 75 characters. Among listings we can compare against a pre-wave baseline, 39% had been cut to compliance while 51% were still over the limit, most of them never touched at all. The pace has fallen roughly six-fold, from about half a percentage point per day in early August to about a tenth of a point per day since. If your titles are untouched today, that is not evidence you are exempt — Amazon can resume at any time — but the wave that was going to reach you may already have passed.

Did Amazon shorten my title, or did my listing change?

You can tell them apart by comparing what you submitted against what Amazon displays. The Listings report holds your submitted title; the catalog holds the title shoppers see. We found 70% of seller feeds still hold a title over 75 characters, and 20% of listings show a short displayed title sitting on top of a long submitted one — those are cases where Amazon rewrote it and the seller's own listing data was never updated to match. Restricting the comparison to 4,466 listings we watched cross from over-limit to compliant, Amazon made 60% of those edits and the sellers themselves made 40%.

Is Amazon still shortening titles?

No, and that is the most important change since the rule launched. Enforcement has slowed roughly six-fold: the share of listings still over the limit fell about half a percentage point per day in early August, and about a tenth of a point per day from mid-August onward. Daily title changes across our panel dropped from 200-400 a day in late July to between 6 and 40 a day through the back half of August, measured on a panel that grew over the same period. Four weeks after the deadline, 53.7% of the top-selling US listings we track still display a title over 75 characters. Amazon can resume at any time and gives no warning before a rewrite, so this is more time rather than an exemption.

Which listings lost the most clicks from Amazon's title change?

The ones cut from a moderate starting length, not the longest ones. Grouping rewritten listings by how much text was removed, the median brand saw a click-through change of -1.6% where more than 100 characters came out, -4% where 51-100 came out, and -20% where only 26-50 came out. Sorting the same listings by their starting length shows the same pattern: titles that began at 151-200 characters lost 3.8%, while titles that began at 76-100 characters lost 14.6%. Cutting a 200-character title mostly removes keyword stuffing shoppers were not reading, while cutting a 110-character title removes the specifics that made the product distinguishable. Impressions stayed roughly flat across every group, so these are click-through changes rather than listings being shown less often.

Do shorter Amazon titles convert better?

On average, no. Listings Amazon shortened saw a median click-through change of -6.6%, against -6.3% for listings it left alone — every group drifted down together, and across three separate measurements that gap has been -0.6, +0.7 and -0.3 points, changing sign each time. But the average hides a great deal: 63% of rewritten listings moved more than 20%, and what predicts the direction is how much the rewrite removed. Listings that lost more than 100 characters saw a median brand-level change of -1.6%, while listings that lost only 26-50 characters saw -20%. Cutting a very long title removes padding; cutting a moderate one removes the specifics that made the product distinguishable. On conversion rate and ACOS we still decline to publish a figure: the treated and control groups are measured around different calendar dates by construction, and the remaining difference is not separable from noise.

Related guides

• Amazon Prime Day 2026 ad benchmarks
• Best Amazon ads management tools in 2026
• Best Amazon ads reporting & analytics tools in 2026

---

You started this brand to build a product, not to audit character counts. But a machine has been editing your storefront copy, and your own feed still shows the old version.

Find out which of your titles Amazon rewrote, and which are still yours to fix. The 75-Character Title Audit dashboard shows your own before and after, and ranks what to fix by what it is costing you. Connect your accounts — free, and it runs on the data you already have.