Reviewing the Reviews: Review Meta

There are a lot of people on the internet. A lot of voices and a lot of opinions. Often, it feels like there’s no one on your side, no one looking out for you, the little guy, the baby tater.

Well don’t worry, potat. Review Meta is here for you.

Reviewing Reviews, but Differently

While we happy-go-lucky boys of Review Party Dot Com fill our time picking apart reviews to find the funny centers, Review Meta analyzes reviews, looking for the truth. After all, a review is someone’s voice, someone’s experience, someone’s lived truth, shared with another.

Except when it isn’t.

I don’t know if you know this, but not everything you read on the internet is true, and reviews are not exempt from the untruthiness. Bias, misinformation, and malice can leak in, tainting the well. So Review Meta takes a sieve to that well and separates the good from the bad.

How’s all that work, now? Maaaagic! Or, well, computer analysis.

Review Meta is a website that scans Amazon product pages (including areas of Amazon.co.uk, Amazon.ca, Amazon.com.au, Amazon.de, Amazon.fr, Amazon.it, Amazon.co.jp, Amazon.in, Amazon.es, Amazon.cn, Amazon.com.mx, Amazon.br, and Amazon.nl), and it analyzes all the reviews on the pages, running them against a set of 12 quality tests.

Here is a look at the results of one product’s analysis:

In the Fail row, we see Deleted Reviews are flagged as a fail because a high number of reviews were deleted, meaning they were present in one scan, then missing in a later scan. This high number is problematic.

In the middle row, we see Unverified Purchases have been flagged as a warning. This means that reviews were left by people who didn’t get that “How’s your new product” email, and instead went directly to the product page to leave a review. It might not be a problem, but depending on the number of these reviews - and if they are significantly higher/lower than verified purchase - this is still worth noting.

Each test works a bit differently, and all taken together, they affect the “value” of each review, as well as the overall grade. Devalued reviews are removed from Review Meta’s analysis for a final product review:

In this case, the final review is still the same, but is now based on high-quality, non-abnormal reviews. In other words, more trustworthy seeming reviews, based on raw data.

Here’s something that’s important: you can do this with any product on Amazon, and all you need is the link to the product page.

Now, you might be wondering what product we’re looking at here. No? Well. Okay, just pretend!

We’re looking at none other than meme review classic, Tuscan Dairy Vitamin D Milk Gallon:

You’ll notice it’s no longer available. Worry not…

And since this silly product received silly attention and gained many silly reviews, it makes sense that you’d want a service like Review Meta to help you clear the clutter and get help from real reviews. Reviews that intend to aid, not entertain.

After all, consider this:

This is what large-scale data analysis can reveal. Do you really think anyone who left a review on a gallon of milk on Amazon also happened to buy and leave reviews for the Three Wolf Moon shirt and the Bic for Her pens? Those are meme review royalty.

No, this is the work of review jokesters, well-meaning in their pursuit of comedy, but nonetheless disingenuous, and therefore, not helpful to a true prospective buyer.

Review Meta has set their analyses to do hero’s work. But there’s a but.

Is there a savior out there, waiting to lend their hand? It could be you. This could be your moment to shine, to do the good you’ve always wanted, to make a difference.

That, or at least use the site once. Maybe tell a friend. We think it’s pretty cool.

Review Meta, we salute you.