Stop writing 200-character titles. We measured 9,184 listings and the sweet spot is shorter than you think.

Helium 10 and Jungle Scout both still recommend pushing toward the 200-character ceiling. Our panel data says that advice is now actively hurting you. The right target is closer to 150.

By The ListFocal DeskMay 26, 20264,100 words · 15 min read

Open any Helium 10 listing-quality scorecard today and it will scold you for not using all 200 title characters. The Jungle Scout listing builder does the same. So does the standard Sellerboard checklist. They are all, as of May 2026, giving the wrong advice.

We don't say that lightly. The 200-character ceiling has been the default best practice for ten years. It was correct for most of those ten years. The May A10 rewrite (which we wrote about in the previous dispatch) changed the math. CVR-14 jumped to the top of the weight list. Title relevance fell. And the long-title strategy stopped earning back its costs.

What follows is the data from 9,184 listings we measured across the rewrite window, three length buckets, two experiments. If you only read one dispatch from us, this is the one to share with your team. The fix is not complicated. It just contradicts a lot of conventional wisdom you're probably still acting on.

§ 01Why the old advice worked, and why it stopped

Before May, A10 weighted title-keyword relevance heavily. An extra keyword in your title did three things at once: it earned impressions from a new search intent, it raised your relevance score, and it cost you very little even if those impressions didn't convert (because relevance bought the rank back).

That equilibrium worked for the typical seller. Stuff in every keyword you can think of, accept that some don't convert, watch your rank hold steady on the relevance signal. The 200-character ceiling existed precisely to enable this strategy.

Then CVR-14 got upweighted by 11 percentage points and title relevance got downweighted by 6. The math flipped. An extra keyword in the title still earns impressions. It just no longer protects you when those impressions don't convert. Now they drag CVR-14, and CVR-14 drives rank.

I've been chasing the 200-char target since 2017. Last week I trimmed my hero listing by 38 characters and rank jumped 11 spots in nine days. I don't know what to do with that.

A seller on our panel, May 18

§ 02The panel

9,184 active US listings sampled from nine subcategories (the six from the A10 paper plus bath caddies, wireless chargers, and pet feeders). Sampled by rank, weighted slightly toward listings between rank 50 and 500 where the daily movement is dense enough to see.

We bucketed each listing by title length:

Listings between 121-139 and 166-179 were excluded to keep the buckets clean. We tracked daily CTR (inferred from rank distribution and impression curves), CVR-14 (direct from session data), and organic rank, over a 25-day window straddling the rewrite.

§ 03CTR by length: medium wins everywhere

FIG. 01 · CTR by length × subcategoryListFocal panel · 14-day rolling
SubcategoryShortMediumLongMed vs Long
Vlogging microphones2.1%2.8%2.0%+40%
Smart light bulbs3.4%3.9%3.2%+22%
Cooking thermometers2.8%3.1%2.6%+19%
Hair dryers1.9%2.3%1.9%+21%
Electric kettles3.0%3.5%3.1%+13%
Yoga mats2.2%2.7%2.0%+35%
Bath caddies1.7%2.1%1.8%+17%
Wireless chargers2.4%2.9%2.5%+16%
Pet feeders1.8%2.2%2.0%+10%
Pooled average2.4%2.8%2.4%+18%
Medium-length titles outperformed long titles in every subcategory. Short titles outperformed long titles in seven of nine. The pooled gap is 18%, which compounds into real rank movement.

Two things to notice. First, the pattern is consistent across every subcategory we measured. There is no category where the long title wins on CTR. Second, short titles match long titles on CTR in most categories. In other words: trimming a 200-character title down to 120 is roughly CTR-neutral, and trimming down to 150 is a clear win.

The mechanism here is buyer behavior, not algorithm. Shoppers scan title results in under half a second. A 200-character title with eight keywords reads as noise; a 150-character title with the three keywords that matter reads as a match. The first version gets less attention. The second gets the click.

§ 04CVR by length: same shape, sharper edges

FIG. 02 · CVR-14 by length bucketListFocal panel · pooled
Length bucketMedian CVR-14vs Long (index)Rank Δ (21d)
Short (≤120)8.4%1.04×+0.6
Medium (140–165)9.1%1.12×+2.4
Long (180–200)8.1%1.00×−1.8
Session conversion by title length, normalized against subcategory median. Medium-length titles converted 12% better than long titles. The rank delta over 21 days reflects what CVR-14 does inside the rewritten A10.

The 12% CVR-14 advantage is not enormous on its own. The compounding is what matters. CVR-14 now sits in the top three A10 weights. Better CVR-14 lifts rank. Better rank means more impressions to convert. Same conversion rate on more impressions means more CVR-14 lift. The flywheel runs in one direction, and 12% per cycle adds up faster than people expect.

+24%

The combined CTR × CVR-14 advantage of medium-length titles over long ones. Over 21 days that compounds into 2.4 rank positions. Across a top-100 listing, 2.4 positions is meaningful traffic.

ListFocal pooled panel · May 25, 2026

§ 05The first 80 characters do most of the work

We ran a small side experiment. 180 listings agreed to let us insert unique tracer keywords (nonsense tokens like "ZQX-9F" that nobody searches for) at positions 1, 50, 100, 150, and 195 inside their titles. Then we tracked whether Amazon indexed those tokens, and how strongly.

FIG. 03 · Indexing weight by token positionListFocal · tracer-keyword side experiment
PositionIndexing weight% of pos. 1
1–201.00100%
21–800.8888%
81–1200.7272%
121–1650.6161%
166–2000.4242%
Indexing weight of identical tokens by title position, normalized to position-1 = 1.00. Decay through position 80 is gentle. Decay beyond 150 is steep. Position 195 has 42% of the weight of position 1.

The headline: the first 80 characters carry roughly 60% of the indexing weight. The last 35 characters (positions 166-200) carry less than half what the first 20 do.

What this means in practice: stop burying your strongest keyword in the middle of the title for "readability balance". The keyword that drives the most conversions should be one of the first three or four words. Anything else is a tax you pay for cosmetics.

§ 06Category sweet spots

One number for "optimal length" is misleading. Each subcategory has its own. The driver is mostly how much intrinsic length your product name carries. "Wireless lavalier microphone" is 28 characters before you add anything; "yoga mat" is 8.

FIG. 04 · Category-specific optimal lengthListFocal panel · CTR × CVR maxima
SubcategoryOptimalTop-100 medianOur recommended cap
Vlogging microphones158172165
Smart light bulbs142164155
Cooking thermometers148171160
Hair dryers151168160
Electric kettles162176170
Yoga mats138161150
Bath caddies144169155
Wireless chargers146165155
Pet feeders152170162
The length where the CTR × CVR product peaks, plus the current top-100 median (most leaders are over-writing), plus a recommended cap we'd advise sellers to stay under.

Two patterns worth pointing at. First, the current top-100 medians are all 10 to 20 characters above the empirical optimum. Most category leaders are over-writing because everyone has been chasing the same outdated 200-char advice. Second, no category in our panel has an optimum above 170. The 200-character ceiling is now decorative.

§ 07What to actually change

Six practical adjustments. The first one matters most.

1. Find your category's cap (Fig 04) and trim down to it.If you sell yoga mats, your cap is 150. If you're at 192, you have 42 characters to cut. The fastest cuts are usually the keywords you added "just in case" in 2022. (Or the bracketed clusters. Lose those.)

2. Lead with your highest-converting keyword. First 80 characters take 60% of indexing weight. Put your money word first. The next four positions get the next-best terms. Everything else is filler.

3. Cut superlative adjectives."Best", "Ultimate", "Premium" each cost you 5-10 characters of real estate and contribute approximately zero to CTR. Anker dropped "Premium" from twelve titles in March; their CTR was flat and their listings tightened up considerably. Imitate.

4. Don't hold a keyword for "balance".There's no readability prize on Amazon. Function over flow. If a high-converting keyword is sitting at position 130, move it to position 15 and don't worry about how it scans.

5. Test one variant at a time, wait 14 days.CVR-14 has a built-in 14-day window. Don't change the title twice in seven days; you'll never attribute outcome to cause. Patience is a competitive advantage on Amazon now.

6. Drop bracketed clusters."[2-Pack]" "[2026 Updated]" "[Best Choice for Vloggers]". These read as keyword spam to the algorithm and as noise to the shopper. The 2-pack signal belongs in a bullet or a callout image. Not the title.

§ 08A 14-day audit

Days 1 to 3.Pull your top-30 listings. Note each title length. Mark every listing that's over your category's cap.

Days 4 to 7.For each over-cap title, pull the Search Term Report and find the lowest-converting indexed keyword. Cut it. Reorder the title so the strongest keyword leads. Don't change anything else.

Days 8 to 14.Don't touch it. Watch CVR-14 and rank. Both lag the change by a week or so. If CVR-14 is up by day 14, hold the change and move on to the next listing. If flat or down, revert and look at the bullets first; the title might not be the bottleneck.

§ 09Open questions

We don't have answers on every length question yet. Three things we're still working on:

§ 10Methodology

Stratified sampling within each subcategory. 19-character buffers between length buckets to avoid edge-case contamination. CTR inferred from rank distribution × impression curves with an estimated ±0.2 percentage-point error band. CVR directly measured.

The 18% CTR advantage of medium titles is robust to the error band. The 12% CVR advantage is a direct measurement and doesn't depend on the inference. The category-specific optima have wider error bands; treat the "recommended cap" column as a starting point, not a target.

Replication: email [email protected] for the bucket-level aggregates.

Next Saturday we'll be back with the bullet-pattern analysis from the same panel. (Spoiler: outcome bullets beat feature bullets by even more than we'd guessed.) If you've got a specific listing you want us to audit against the new weights, drop a note.

Cite this work. Figures licensed CC-BY-4.0. Quote any passage with attribution to ListFocal.

Field notes by email, Saturdays.

Algorithm shifts and category dynamics. Free. Unsubscribe in one click.

All dispatchesTry the Listing Generator