I’m going on a tangent to the topic of the article: MLM companies are getting wise about this though.

If you search YouTube for some MLM companies you’ll find many of them presented as an independent opinion/study on whether a specific company is a scam or not. They’ll point out surface-level flaws in a company’s business in the beginning that would suggest a pyramid scheme. They know they can’t hide certain dirty laundry so they put it out there. But then they start talking about other studies, without citing references of course, showing these companies are not a scam. It goes on from there and sadly a lot of people still fall for these when they’re trying research these companies.

Yeah, that's another bad smell, when you google for "<name of pyramid> scam" and you get zillions of links asking "Is X a scam?" and they all try to convince you -- again, using the same or similar language -- that no, seriously, it's totally legit. This is how Scientology tried to google-bomb itself to counter the effects of Operation Clambake, et al.

The error there is that people assume that online reviews are by real, independent reviewers that have bought the product with their own money.

Sampling online reviews does not at all have the same kind of confidence levels as independently interviewing actual users of a product.

Here's the thing: In a multilevel marketing scheme, even the word of the product users is unreliable. This is because MLMs recruit their end users as a sales and marketing force that works, essentially, for free -- with the promise if big riches if you sell enough in the program.

You know "The Secret" and all that positive thinking Aladdin's genie hogwash? That's actually the public facing side of MLM propaganda. Internally, MLMs pump their customer base up with big promises about how if you think happy thoughts you are guaranteed to fly with the hidden (or even explicit!) threat that even the tiniest sad thought will send you crashing to the ground.

And this extends to what people are allowed to say. The key word is "edify". You're supposed to "edify the product" (talk about how great it is) and "edify your upline" (talk about what a great guy/gal the person who introduced you into the program is). This is often backed up with threats and verbal intimidation and abuse. (e.g., "Only losers say things like that. Losers and quitters. You're not a loser, are you?" and it gets worse from there.)

So for some of the most prevalent scams, objectivity can't even be expected from the end purchasers.

I think his simple test somewhat accounts for this. Botted reviews usually don't bother with the sophistication of generating noise in their ratings, banking instead on the possible sale from a spendy shopper instead of trying to convince the already skeptical.

With the assumption that most botted reviews aren't going to be any lower than the two highest options, the same suspicions the article raises should come into play here as well, botted responses or shill responses.

Though, reviews in general are not really all that good. Southwark's Yelp episode was a long but cute take on it, but keep in mind that almost everyone sees reviews differently. On Newegg, it's incredibly common for a part/device to get 1 egg because it didn't work for whatever reason, even when it's the buyer's own fault (e.g., buying an incompatible CPU/mobo, which happens). DoA items, misconfigured computers (e.g., forgetting to plug in a power cable), and so on all warrant 1 star, even though they have nothing to do with the performance of the item in question. (DoA being an exception, though I would argue a device being DoA by chance is not the same as widespread DoA, and that the rating system should have a special category for DoA which gets its own tally and is independent of the ratings)