I got tired of guessing. Every Amazon influencer has opinions about what ranks in the carousel, and almost none of them are backed by anything. So I pulled the data: over 11,000 products, more than 37,000 Amazon Influencer videos, and around 13,000 creators. Then I went looking for what actually correlates with a better carousel slot and more helpful votes.

Some of it confirmed what I suspected. Some of it surprised me. One finding flatly contradicts advice I see repeated constantly. Here’s everything I found.

What’s In This Article

Ranking and Engagement Are Two Different Games

Before any of the numbers make sense, you need this distinction, because most creators only play one of these games and wonder why the results don’t come.

Ranking is where your video sits in the product carousel. Slot one, slot two, slot five. That’s Amazon deciding how visible you are.

Engagement is what happens once people see you: clicks and helpful votes. That’s shoppers deciding whether you’re worth their time.

A great slot means nothing if nobody clicks. And clicks mean nothing if you never rank in the first place. Some of the factors below move ranking, some move engagement, and a few move both. I’ll tell you which is which, because the changes you make should depend on which problem you have.

Video Length: Longer Keeps Winning

This was the clearest signal in the entire study, and it goes against the “keep it short, attention spans are dead” advice you hear everywhere.

Videos under 30 seconds ranked the worst on average. As length goes up, average carousel position gets better, and it keeps improving the whole way. Videos at four minutes and up landed the best average position in the study at roughly 2.2 in the carousel. Anything under four minutes ranked progressively worse the shorter it got, and the curve really starts climbing once you clear the two-minute mark.

So the takeaway is simple: stop shooting 45-second clips for your storefront. Get to two minutes minimum. Three to four is the sweet spot if you want the best odds at a top slot.

How Often to Mention the Product

This one is related to length but worth separating out. Videos in the top carousel positions mentioned the product noticeably more often than videos in the bottom positions — around eight mentions per video at the top versus about 5.9 at the lowest slot.

I’m not telling you to keyword-stuff your own speech. But saying the product name naturally throughout — the full name, not just “it” and “this thing” — appears to help. And if you’re making a three or four minute video anyway, hitting eight mentions happens without forcing it.

Horizontal vs Vertical

Horizontal wins, but you have to be careful with the raw numbers here. About 78% of all videos in the sample are already horizontal, so of course horizontal shows up more in top slots — there’s just more of it.

Once I adjusted for how common horizontal already is, it still wins the top slot 57% of the time. That’s a genuine edge, not just a volume artifact. If you’re filming vertical for your storefront because that’s what you do for Reels and TikTok, you’re giving up position.

The Title Formula That Wins

Titles were the most fun part of this study because titles are 100% in your control and cost nothing to fix. Here’s what the data showed.

One word in ALL CAPS. Titles with at least one fully capitalized word averaged 12.5 helpful votes. Titles with no caps averaged 8.1. That’s a big gap for a change that takes two seconds.

Ask a question or create a curiosity hook. Give the shopper a reason to click instead of stating a flat fact. “What you need to know” beats “I love this product.” The curiosity hook showed up repeatedly as an engagement driver.

Punctuation helps — mostly. Colons, pipes, and question marks all performed well. Titles with a colon averaged 16.2 helpful votes, which was one of the stronger title signals in the whole set. Exclamation marks were the exception: they create no curiosity, they just shout, and they underperformed.

Keep it 50 to 69 characters. That range was a clear sweet spot. Go over 70 and the title gets truncated in the carousel, so the buyer literally cannot read what you wrote — which is exactly why those titles underperform. Personally I try to stay under 60.

Do Custom Thumbnails Actually Matter?

This is the question I get asked most, and I finally have a real answer instead of a guess. About 61% of videos in the sample already have a custom thumbnail, so again the raw numbers need adjusting.

Head to head inside the same carousel, custom thumbnails ranked above frame grabs 55% of the time. Adjusted for how common custom thumbnails already are, they still win position about 53% of the time. So yes, they help — but not by a landslide. A little over a coin flip.

Where thumbnails earn their keep is engagement. In slot one, videos with a custom thumbnail pulled around 16 helpful votes versus frame grabs well below that. So the honest framing is: a custom thumbnail is a modest ranking edge and a meaningful click edge. Make one. Just don’t expect it to rescue a 40-second vertical video with a bad title.

Comparison Videos: The Biggest Number in the Study

If you take one thing away from this whole article, take this one.

Comparison videos averaged 30.6 helpful votes versus about 7 for non-comparison videos in the same sample. That’s roughly 4.4 times the helpful votes. It is the single biggest helpful-vote number of any metric I pulled, and nothing else came close.

They also took a larger share of the top slot — around 13.1% of slot one versus 10% in the lower positions. So they rank better and convert dramatically better.

This makes intuitive sense. Someone looking at a product page is deciding between options. A video that puts the product next to its alternatives is answering the exact question in their head. Most creators don’t make comparison videos because they’re more work — you need multiple products in hand and a plan. That’s precisely why they win.

This is the workflow Oink’s Comparison Video Schedule was built for: figuring out which of your products can be paired up and planning the shoots so you’re not scrambling.

The “Honest Review” Mistake

Here’s the finding that surprised me most. Using the word “honest” in your title is associated with a massive decrease in helpful votes. Not a small dip — one of the clearest negatives in the dataset, in the range of a 25% drop.

My read on it: “honest review” is such a worn-out phrase that shoppers have learned to discount it. Announcing that you’re being honest makes people wonder why you had to say so. Meanwhile, actual review-style content did fine on ranking — review-style videos skewed toward the top at 39.6% of slot one versus 29.2% of slot five, and earned around 87% more helpful votes in the same sample.

So make review content. Just don’t put the word “honest” in the title.

The Playbook for Your Next Video

Everything above, condensed into what to actually do:

  1. Film at least 2 minutes, ideally 3 to 4. The clearest signal in the study. Four-minute-plus videos averaged the best slot.
  2. Shoot horizontal. Still wins the top slot 57% of the time after adjusting for how common it already is.
  3. Name the product often — aim for around eight natural mentions.
  4. Compare it to alternatives whenever you can. 4.4x the helpful votes. Nothing else in the study performs like this.
  5. Write a 50 to 69 character title with one ALL CAPS word and a question or curiosity hook. Use a colon or pipe. Skip exclamation marks. Never use “honest.”
  6. Make a bold custom thumbnail. Modest ranking edge, real click edge.

The winning formula, in one sentence: a three to four minute horizontal video that names the product often and compares it to alternatives, published with a bold custom thumbnail and a 50 to 69 character title containing an all-caps word and a curiosity hook.

Here’s the full breakdown with every chart if you want to see the numbers yourself:

Finding Products Worth Filming

Knowing the formula only helps if you’re pointing it at the right products. That’s what I built Oink for — Storefront Cross Check, Comparison Video Schedule, and Unavailable Video Matching all exist to keep you filming things worth filming and keep your storefront from rotting.

You can grab it at oinkforinfluencers.com.

A note on method: these are correlations from observational data, not a controlled experiment. Videos that are longer and include comparisons may also be made by creators who are simply better at this. Treat the findings as strong signals about what tends to accompany success in the carousel, and test them against your own results.

I'm an Amazon Influencer since 2023 and blogger since 2021. My goal is to help you succeed in the Amazon Influencer Program.

Similar Posts