johnwdefeo

Speaking as an artist, many (most?) of my enduring works were the result of an accident of some kind. I call them "happy accidents" because I recognized that the mistake was better than whatever the vision was that I had at the time.

As a corollary, there are unhappy accidents, and with respect to life forms in a chaotic system, such accidents can perpetuate and endure without human recognition.

I sincerely hope not. From one of the authors:

"Scientists publish papers not because the paper is the end of science, but because it is a unit of research that is valuable to share with others so that others can use this brick of knowledge and either build with it… or find its weakness and break it down...We wrote our entire analysis in R and shared our code with the world. I tried SO hard to check every single line of code and make our pipeline clear & easy to reproduce. However, despite nearly giving myself stomach ulcers checking every line and stressing about these findings, it’s possible someone finds a mistake in our work. We don’t share this work happily - this is the saddest paper I’ve ever written. We’ve shared our code precisely for that reason: we want you to see exactly what we’ve done, and if we’ve done something wrong we are open to hearing it."

As to your original concern, it is a valid one. I wrote this is response to pre-prints popularized via the press earlier this year:

-> Make bold, unjustifiable claims in the preprint; -> Ensure widespread coverage in the science press; -> Walk back those claims during peer-review; -> Get published; and then -> Watch blue checks tout original claims as "Fact!"

Yet still relevant and yet to be answered for: https://www.bmj.com/content/375/bmj.n2635

Thanks for this. I stumbled across the 2015 study and I wasn't aware of the controversy surrounding its recent dissemination. Additionally, here is a recent article published by the author of the study: https://theconversation.com/vaccines-could-affect-how-the-co...

That said, I feel frustrated by how the author alternates between weak and strong wording that may influence a reader's interpretation. For example:

Said then: "When vaccines prevent transmission, as is the case for nearly all vaccines used in humans, this type of evolution towards increased virulence is blocked."

Said now: "...no vaccine is 100% effective...we still need more data to determine how leaky [the mRNA vaccines] are..."

Said then: "The use of leaky vaccines can facilitate the evolution of pathogen strains that put unvaccinated hosts at greater risk of severe disease. The future challenge is to identify whether there are other types of vaccines used in animals and humans that might also generate these evolutionary risks."

Said now: "Individuals and populations have always been better off when vaccinated. At every point in the 50-year history of vaccination against Marek’s disease, an individual chicken exposed to the virus was healthier if it was vaccinated. Variants may have reduced the benefit of vaccination, but they never eliminated the benefit."

If the author believed then what he says now, it would have been beneficial to include in the abstract of the 2015 study something like: In the view of the authors, this data suggests that next-generation vaccines coupled with mass vaccination programs are necessary to combat evolutionary pathogenic risks.

Thanks, likewise!

Thanks for the response. I created best bourbons page for fun after my traffic had collapsed, so that can't be the culprit.

I redirected the cloth mask page to the CDC's website. It was something that I created at the behest of friends and family members (and then forgot about). It's worth noting that I did not profit from that page when masks where in high demand. (In the early days of the pandemic, I actually spent a good chunk of time trying to wrangle available PPE and get it into the hands of first responders. I had a big moral issue with the publishers who were profiteering by selling consumers things that they didn't need at the expense of those who did.)

Didn't consider a homepage link from the footer, but it's a fair point. Added.

Sorry to hear that. I've heard a saying that "rejection is protection," and the way that you rationalized your experience lends credence to that.

Thanks for this. Good suggestions here.

Thanks, I appreciate the support.

I haven't mirrored the content, but I did republish/redirect a few pieces elsewhere and they fared much better. The issue is tied to the domain.

Even with brand recognition, it's hard to build a product review business that doesn't rely on Google. Until 2018, I was the head of insights at one of the largest online publishers in the U.S. 90% of the company's revenue was tied to organic search. Sure, branding helped to prevent huge impacts like the one that I saw, but plenty of our competitors suffered big drops that resulted in 100s or 1,000s of layoffs.

I'm fortunate enough to have a kind and supportive professional network.

Good suggestions, thank you.

I agree with this sentiment: https://www.johnwdefeo.com/articles/google-is-ruining-the-in...

That 7% conversion rate is legit (and was equally surprising to me). On top of that, my return rate is lower than typical. I know that my site isn't for everyone, but the people who do happen across it seem to get value from it.

The website is another run of the mill affiliate website built with the sole purpose of making the owner some money.

This is the exact opposite of every principle that I believe in. If that were true, at minimum I would have:

-Placed ads on the site -Opted into Skimbit's data sale program -Hyperlinked titles and images with affiliate codes -Written about huckstery things like CBD oil and salt lamps.

I feel bummed to read this take, but if you're interested in learned more about my process, here are some links:

https://www.goodcheapandfast.com/about https://www.goodcheapandfast.com/articles/value-award-winner... https://www.johnwdefeo.com/articles/amazon-review-analysis

I've spent most of my career working for "real review" sites and that was part of my motivation for creating Good, Cheap and Fast:

Good, Cheap and Fast isn't meant to be a substitute for lab-tested product reviews, but I hope it's a complement. I like Cook's Illustrated and Consumer Reports. These publications buy products, test them and cover their costs by charging a subscription. This is the best business model for protecting readers.

There are plenty of “free” websites that also test products, but visitors to these sites end up paying in less obvious ways:

Ad Bloat - The websites have great content, but it's hard to enjoy it because of excessive advertisements, autoplay videos (with sound) and third-party networks that hijack your browser and redirect you to pop-up ads claiming that you’ve won a free gift card.

Price Creep - The websites are great at reviewing products, but in order to cover costs, the tested products are more expensive than necessary. Many publishers do this (unwittingly) because the affiliate commissions that they receive are based on a percentage of a product's sale price.

These phenomena are common on legitimate websites. Nevermind sites that:

Rent studio space to stage their "lab testing;"

Don't know how to use lab equipment properly;

Fail to disclose pay-to-play relationships;

Give inflated reviews to maintain relationships with PR people; or,

Fail to investigate the issues that products have after extended use.

I’ve seen these trespasses throughout my career: Hucksters are everywhere.

You've stoked my curiosity. Which sub-$200 144Hz monitors are your favorites?

I'm surprised to hear this, not because I think my recommendations are infallible, but because the user return rate is lower than average. Are there any specific products that you think are junk?

That perception is an understandable one, yet a bitter pill all the same.

Changing algorithms also favor a small set of incumbents. The stock of my former employer is up more than 2,000% in the last 5 years. That's amazing growth for any publisher and it's largely based on SEO.

I agree with this in concept, but it has been a bitter pill to see my site overtaken by scraper/outsourced content that I consider to be pure spam. Google doesn't always get it right (regardless of my own situation).

I spent my first year scrubbing the seed corpus to minimize the GiGo effect: https://www.johnwdefeo.com/articles/amazon-review-analysis - I also review all the pricing data manually to flag and discard outliers.

Not sure about why Anandtech stopped endurance testing, but I would assume cost cutting. I left the company in Sept. 2018 which was right after it was acquired by Future PLC. They're a frugal lot.

Yes! It would be easier to write 500 word blurbs than a meaningful 50 words. I despise recipe sites for this reason. So much "content" gets in the way of the actual content. The recipes in my favorite cookbook are around 100 words on average, yet if they were published online, I doubt they would rank.

That's exactly right.

The manual vetting that I do revolves around discarding real-yet-unhelpful reviews (which aren't easy to spot algorithmically). E.g.

Off-Label Usage - Customers rate a portable jump starter 5-stars, even though they have only used the product to charge their smartphones, not to jump start a vehicle.

Self Validation - Customers rate a carbon monoxide detector 5-stars because they feel a sense of relief and validation that their purchase will protect their families.

Customer Service Uprating - A 1-star rating is later updated to 4- or 5-stars because the manufacturer offers the customer a replacement product (and suggests altering the review).

Misunderstanding - A customer leaves a negative review because he or she didn't read the product description carefully and is consequently disappointed with the product.

Ideology or Spite - A positive review is paired with a negative rating because the customer disagrees with the business practices of the manufacturer (e.g. It's a great product, but Widget Corp. is a POLLUTER!).

Wrong Model - A review for one variation of a product is lumped in with reviews of another version of the product. (Hard drive failure rates can differ by 900% depending on the size of the drive.)

Wrong Product - A product page is repurposed by a seller, thereby mixing the reviews of one product with a completely different one. E.g. A page about a protective phone case contains reviews about a wireless charger.

Shipping Issues - Customers leave negative reviews because their packages arrived late or damaged in a way that reflects negatively on the shipping carrier, not the manufacturer.

Joke Reviews - A customer uses his or her review as a platform for comedy. Sexual wellness products, or those that are gender-based, seem to be disproportionately affected.

Empathy or Pity - A customer has a bad experience with a product, yet he or she leaves a positive rating (typically, 4-stars) because "someone" might like the product.

I've never claimed to test products, but I do consider my selections to be highly-vetted: https://www.goodcheapandfast.com/about - There is around 2,000 hours of research behind them. The trick (from a UX standpoint) is deciding how much work to show when the spirit of the site is to be dead simple.

Thanks! My intention was to complement longform content rather than going head on with it.

Interesting. I've struggled with the balance of explaining my process vs. getting out of the way of the recommendations. And for that matter, how many products is the right amount to feature. The earliest iteration of the site only listed 3-5 products per category, which may have been an inadvertent sweet spot. I had hoped that my awards program would help people to learn more about my selection process: https://www.goodcheapandfast.com/articles/value-award-winner...

Fair points. I don't scrape/spin any content, and I would argue that it's easier to write 500 words than 50, but I can see how my content might run afoul of an automated system. Plenty of the comments on this thread make similar points.

Ironically, I know of a scraper site that mirrors Amazon's bestsellers list, reformatted as blog posts: https://gistgear.com - and it's growing like gangbusters.

Thank you for the thoughtful reply.

I agree that the spirit of my site is closer to "a statistics site for products" than a typical review site. Adding more content, photos, etc. may help to remedy my situation, but I would feel like I was abandoning my original purpose. I designed the site with "satisficers" in mind, i.e. allow people to make a decent decision as quickly as possible. No frills.

Until recently, I worked for product review sites like Anandtech and Tom's Hardware, and I'm not equipped to compete with them head on. My hope was to create a new niche by using data analysis, which is more of my skillset: https://www.johnwdefeo.com/articles/amazon-review-analysis

I look for fake reviews by measuring the relationship between 1, 2, 3, 4 and 5 star reviews. There is a normal distribution and when I find outliers, it usually indicates that something suspicious is happening.

If I understand what you mean by inventory co-mingling, there is actually a drill-down that's built into Amazon itself. For example, the reviews of Samsung's HW-T450 look awesome at first blush: https://smile.amazon.com/dp/B085WTFCQ7/ - but, if you filter the ratings to only display that particular model, you'll see that 22% of the reviews are 1-star: https://smile.amazon.com/product-reviews/B085WTFCQ7/ref=cm_c...

This an automated check that I perform, but I also investigate the reasons for these disparities. (In this case, it's because the soundbar lacks a digital decoder to support the surround sound speakers that are sold as an additional accessory). On rarer occasions, the negative reviews can be discounted, even if they are legit. One example of this is when lots of customers misread the dimensions of a product (e.g. I assumed it would be larger...).