Hacker News new | more | comments | ask | show | jobs | submit login
What If Google Doesn't Reward White Hat SEO? (johnwdefeo.com)
99 points by midef 24 days ago | hide | past | web | favorite | 86 comments



You're a professional "SEO guy" and still surprised that your four-month-old, 50-page, one-level-deep, affiliate-linking site is not outranking the hundreds of established review sites? Come on, man. If it were that easy...


Right? I feel like this article was just another way to get publicity and social traffic to his domain. Anyone worth their salt would know he's probably stuck in the "Google Sandbox" https://ahrefs.com/blog/google-sandbox/


It seems to be working. We are all here commenting on a page that's linked to a page, that's linked to his "slightly better than thin" collection of affiliate link loaded articles.

And, despite all the various measurable quality signals, content quality, etc...Google organic rankings still mostly care about links.


It's a good article. There are some quotes in it that galvanize my point. E.g.

"I know for a fact that it’s possible to circumvent these sandbox effects. If a site goes even somewhat viral and is getting a few hundred to a few thousand daily unique visits, some social media mentions each day, and real backlinks, that site will immediately be taken seriously by Google’s algorithms. I’ve seen this happen twice so far with brand new websites, eventually ranking #1 for their top keywords."

My experience has been the same. Except in this case...and it just struck me as strange.


You mean... it was, itself, an attempt at SEO? :P


we see this dark pattern deployedall the time on hn


I'm genuinely surprised that the site hasn't found more traction than it has. Don't get me wrong: I wasn't expecting to set the world on fire or outrank the big players, but I have seen young sites do a lot more with a lot less (including some of my own).


This is site of which the author is speaking

https://www.goodcheapandfast.com/

I know little about SEO ... but I'm sorry man you are CRAZY if you think your are going to rank for the words

good, fast, OR cheap

for something that came out in the past year.

And outside of that im sure there is strong competition to rank for all of the categories of devices that appear on the site...

This is just not a think the author can make the conclusion in their title based on...


He's not trying to rank for those - he's going after product search terms such as "best robot vacuum", which is also quite hard to rank for.


Over the last 2 years, I've watched my business go from the 4th page of results for important searches, to always ranking as #2-4 on page 1. More than anything else, what made the difference was time. In my experience, Google heavily rewards old backlinks and old websites -- much more than it used to around 2005. Maybe it's an anti-spam technique, since age is a hard thing for black hat SEOs to fake.


> More than anything else, what made the difference was time.

Because it's impossible to fake the age of a domain. If you have a long-running site with lots of old links from other long-running sites, it has phenomenal pagerank power.

I am aware of this because I have been a volunteer admin and the DNS contact for blogs dating back to circa 2004. My inbox is an absolute river of people sending me offers to write articles, infographics, helpful "corrections" for "broken" links etc etc et bloody c.


What about buying an expired domain for a relative site and redirecting it to your existing one?


Google have said they know about change of ownership of domains and don't count links from them.

Having said that brands with long running sites still have links to pages from 95 - 97 that are worth redirecting I found one last week :-)


I know google has said this. But from my testing and experiments it is not true. Can't say anything else.


There's a reason so many people buy up expired domains, it's common to either start a new site with an aged domain or 301 redirect them to your site for the extra boost


Which is not detectable at all...


> since age is a hard thing for black hat SEOs to fake.

You would be surprised.


Loooong-time digital & affiliate marketer here. Interestingly, I'd say the #1 culprit here is time...and the second is probably content thinness.

Each page is certainly useful, but doesn't really show much information that google _believes_ is useful to a user.

Example: https://www.goodcheapandfast.com/articles/best-cordless-vacu...

The page itself is quite useful to someone searching for the right cordless vacuum, but google's algorithm is looking for signs of quality that would assist it in better understanding the potential value.

Examples here include:

1. Structured data in the page such as a standard price / discount price / features / avg rating table

2. A longer description of the product with more information, such as features

3. Shorter backlink anchor text to Amazon that have less data crammed in them (try and separate out the features from the actual product name here)

4. Crosslinking: you have reams of content that is related, so cross link it! You're making it hard for google to discover relevant content paths. Cordless Vacuums -> Robot Vacuums -> Dyson Vacuums -> Dyson Vacuum Accessories, is one example. The content journey one can take is pretty much Home -> Search -> Article -> Amazon, much too short.

5. Everything is stacked under tld/articles/article-name and you have very little variation, making it harder for users to easily see this sort of indexing. Interestingly, category pages, landers with aggregation of multiple areas of content, etc. can help because you assist in discoverability for the user. Keep your canonical tags sorted, no-index variants of category pages so you don't end up with reams of pages indexed, and you should see better potential user journeys.

This is just off the cuff and YMMV, but is my first take. Hit me on LinkedIn if you want to chat.


Don't spammy affiliate sites do all these relatively easy to automate things already? Presumably they don't have decent backlinks though.


Sort of - spammy sites can create a lot of formulaic structured content. It is fairly easy to see that though as they tend to have lots of auto generated content, duplicate content, thin pages with lots of links, etc.


I'm not seeing how Google is going to tell the difference based on that. It's not hard to auto generate content that has fewer links, doesn't have duplicate content etc. The website the OP is about looks formulaic as well. Not trying to knock the effort put in to it but my immediate reaction to the site was it looks like a spam affiliate website.


Thanks for the encouragement and double thanks for emphasizing the word "believes." I work hard to make the content as short as possible while still being useful, but it's true that Google may not be able to see the value in that (at least, without a lot of user feedback signals).


1. The site isn't very user friendly.

2. What's your average dwell time? If people are visiting then bouncing after 1 minute that sends a bad signal to Google. Average dwell time of a first page result is 3 minutes 10 seconds, according to Brian Dean.

3. You feature 0 images. Where's the alt text? Looking at these walls of text makes me want to bounce ASAP, which isn't obviously good for your dwell time.

4. I think you need to step up your content game A LOT. Look at what Wirecutter does: https://thewirecutter.com/reviews/best-high-end-ranges/

5. You say the site features no ads-- but that's kind-of a lie. You are an affiliate site, which I personally never trust anyway, when they're created in this sort of way. It's like look how bad the mattress industry is with these sites. You can't trust any of them.

6. Good content that ranks in Google is comprehensive. You should think about making your post format meatier. Help people choose which one is right FOR THEM. Explain why you chose what you chose. Use descriptive headlines so people can skim. People want to see pics before they click through.

7. I highly recommend learning more about what constitutes good content vs bad content.

Resources:

- https://backlinko.com/seo-this-year#rankbrain

- https://medium.com/@laurenholliday_/here-is-the-real-reason-...


Thanks for detailed comments. Many of the design choices that I made for the site are polarizing (like the decision to include no images). Same goes for the content: My goal is to help people get in and out of the site in five minutes or less. It's a bummer when that seems like a cop out. I've found it much harder to write short (and meaningful) content than it is to write longform.


The authors site (https://www.goodcheapandfast.com/) is really annoying BECAUSE of the SEO content, there’s like a page of pointless text about each category and then each product has a bunch of text about it which is totally useless.

Is this really the best way that google can distinguish good from bad content? IMO google results are starting to become stale and it makes the internet feel a lot smaller, this is probably why. I know there are tonnes of far more interesting things out there but they seem to come up less and less now.


I didn't write any of that content for SEO purposes. In many cases, it's an opportunity to clear the air about misconceptions that people have (sometimes, dangerous misconceptions). Is there a specific page that you found useless?


Maybe the ranking algorithm doesn't punish sites for having contextual information about links.


My results are stale too! It sucks. =( They're never good anymore.


Seems like the amount of affiliate links is the issue. I guess it's hard for search engines to tell the difference between spammy review sites made for affiliate links and the author's site.


At first glance it's hard for me to tell that this isn't regular run of the mill affiliate search result spam.


I can't tell the difference between this and a spammy review site just for affiliate links. He's paraphrasing the main points in amazon reviews in a short paragraph and putting a link at the bottom. I don't really see what I gain from this vs just looking at popular products on amazon directly.


I weed out the products that have excessive fake reviews, above-average prices and some other interesting "tells" that I've discovered (e.g. the 4-star reviews are amended 1-star reviews that have been changed because the customer received a replacement product).


Thanks for the reply. I do see how you could have honed a knack for picking out particular tells like those, that knowledge isn't going to be immediately obvious to most consumers.


I considered this and decided to withhold affiliate links for the first couple months of the site's existence. It didn't seem to make much of a difference one way or another.


This article is pretty hard to see in its default rendition. Light text on a light background? Not sure that's the best design choice...


It makes me wonder if another part of this story is that Google doesn't reward low-contrast text, either.


The default font is hard to read regardless of contrast.


It's true, Google and other sites will deindex a site like this into oblivion - as at first, it has many of the characteristics of a spammy affiliate site.

I think there is a real need for this kind of information, and I would even suggest Google or (any search company really) go as far as to build a feature/small-platform specifically for content like this. _Quality_ affiliates linking.


Intrigued, I fed your site into a couple of CI tools. Some initial thoughts....

- Keyword targeting: seems like you're targeting a lot of product based keywords; these tend to be hotly contested, generally favoring brands and top retailers for top slots - Not quite enough content on the pages, in terms of word count... - As you mentioned, Google is probably seeing the Amazon links on your landing pages and penalizing your rankings

So - potential thoughts on how to change it up...

- Switch from talking about products to problems... probably can get rankings a lot easier for those keywords, especially if you're willing to target long tail traffic. - Move amazon links off the landing pages; at a minimum, cloak the link but better yet try having a page or two that discussed the problem then make the user click to get to another page which recommends specific products - Also getting the vibe you might need more content density for each topic; more pages addressing a single area, with a higher word count.


I wonder if replacing amazon links with a google-shortened link to the same would improve SEO.


I believe this is against Amazon's terms (though large sites do some variation of this all the time).


I have enormously mixed feelings about anyone who says they are a professional in SEO. I don't doubt there are people who show results to their clients, and they will range from amateurs to people who charge to do it for a living and they will be black, grey and white hatted variants.

My own experience is that carefully reading the google documents which return when you do the obvious searches on how to improve SEO are very clear: do some structural work, be very clear about your sites underlying page relations, proffer keywords which contextually make sense, but do not try and game the system.

If you pay search engines money they increase visibility of your site. If you host adverts since they place adverts it increases visibility of your site in their systems because its a "virtuous circle"

If you attempt to use other mechanisms to direct traffic to your site which the search engine can (a) detect and (b) proscribe, you will be penalised.

What else is there?


The biggest problem nowadays, though, is that we're building sites more and more that aren't crawlable by the search engines. Basic site structure and usability issues plague many sites. A good SEO will fix those issues.

It's not the link building or content that needs to be fixed, what is needed is the skills of a competent technical SEO who understands web dev. In many cases a site just needs to fix it's site structure and the crawling issues and it will rank well.

For example, I've seen plenty of sites that have over 1 million pages of content but the only way to get to those pages is through an XML sitemap--that's not SEO.


Well said. The biggest problems these days stem from things like client-side rendering and making crawler-accessible versions of things that are using new technologies (like simple pagination for infinite scroll, etc.).


>> carefully reading the google documents which return when you do the obvious searches on how to improve SEO are very clear: do some structural work, be very clear about your sites underlying page relations, proffer keywords which contextually make sense, but do not try and game the system.

I honestly wish it was that easy. Trust me, it's not.


It is that easy if the site is worth the front page, which is the point of the doc.

I think everyone just assumes the front page is fair game for them. If there are ~10 better sites then yours (not out of the question unless your in a tiny niche or started the site long ago) then doing the work still won't reward you with the front page, and there isn't anything necessarily wrong with that.


Google in general doesn't like anything that intersperses an additional page between the searcher and their destination.

Part of that is user experience - their goal is to get the searcher to the information they need as fast as possible, and it's hard to slow them down more than by putting an additional page between them - and part of it is corporate strategy. Anything that helps organize the world's information - particularly information connected to a purchase, which is Google's revenue stream - is potentially competition.


> Google in general doesn't like anything that intersperses an additional page between the searcher and their destination.

Unless it's an AMP page of course!


IIUC the intent behind AMP is that the AMP page is the destination.


Except when it doesn't work.


I mean, I dislike broken AMP pages as much as any other user. But from the perspective of Google management, a broken AMP page is a bug report that you file with the relevant engineering team. An interstitial or aggregation page that some third-party puts up is outside of your control, and so you disincentivize it with product incentives.


Your website does not contain a single image. Try to use optimized images and use proper markup, use alt tags etc.


It's a (polarizing) design choice that I'm sticking with (for now). Some folks tell me it's a dealbreaker, but others say they love it. My inspiration was: https://text.npr.org


If you're an "SEO guy" you should know that any SEO techniques that clearly violate Google's Webmaster Guidelines are not "white hat SEO".

For example, specifically take a look at the line that says, "Links with optimized anchor text in articles or press releases distributed on other sites" in the Google Webmaster Guidelines.

How do you explain the hundreds of press releases, with optimized anchor text, linking to good cheap and fast dot com? Such as this one: http://www.kitv.com/story/39477963/presenting-good-cheap-and...


The only link in the press release is the name of the website. The goal of the press release was to get press, not links. It was a dud (save your money if you're considering announcing a new website this way).


The anchor text on that PR is the name of the website. Maybe i'm missing something?


Mass distribution of press releases became an outdated, spammy technique about 5 years ago. Not considered white hat by any stretch of anyone's imagination.


SEO is my main issue as well. I write good content, share with the right folks, write guest posts (that got featured of 4 different big websites), yet my traffic is tanking!

It was growing steadily, but after I switched to https, it started consistently falling, every day.

Google, I try to keep my website healthy and have good content, why are you testing my patience?

Is link building still the #1 strategy to boost your page rank? OR has the game changed?

Can anyone point to a good SEO guide? (Would be willing to pay $$$)


In general, I have found this blog to be very useful https://medium.com/@KeywordsHeaven and their main site https://www.keywordsheaven.com/.

They seem to publish new articles/guides every week or so and have addressed link building in the past which, according to their case study, appears to not be the most important thing, although it is important.


> Can anyone point to a good SEO guide? (Would be willing to pay $$$)

I wrote a concise and actionable guide for on-page SEO here:

https://www.checkbot.io/guide/seo/

I also wrote a Chrome extension that will check the suggested SEO guidelines for you automatically, checking multiple pages at a time:

https://www.checkbot.io


When you moved to HTTPs, and the site tanked, then most likely it's due to a bad HTTP/HTTPs migration. I see that all the time, and there's a lot more involved to migrating to HTTPs then you'd think.

The biggest problem I see is that sites are migrated to HTTPs but they fail to set up the redirects properly or they forget to updated all the internal links on the site so they point to HTTPs.


I did set up the 301 redirect for main page. Every existing link was also updated to HTTPS.

If you navigate directly to the http url for an article, it does NOT redirect, but if you goto http://www.mysite.com, it redirects to https://www.mysite.com

I also see that when I Google my site's content, the links in google are https://, which means Google now knows that my website is running https.

What may I be missing here?


> If you navigate directly to the http url for an article, it does NOT redirect

That would be your problem. As far as Google is concerned you are operating two different sites. Redirect all your old http URLs to the HTTPS versions of the same content. But it might be too late if that old http content is considered “gone” from Google’s index.




Offtopic -

I dont understand HN ranking at all.

At the time of writing Apple's earnings sits at #10 with 45 points (36 comments) and this one sits at #6 with 20 points (3 comments).


https://news.ycombinator.com/newsfaq.html

> "How are stories ranked?"

> "The basic algorithm divides points by a power of the time since a story was submitted. Comments in threads are ranked the same way."

> "Other factors affecting rank include user flags, anti-abuse software, software which demotes overheated discussions, and moderator intervention."


"software which demotes overheated discussions" is pretty noticeable. It seems as soon as a story gets more comments than it has points, it drops like a rock from the front page.


IMO this is by far one of the best features of HN - discussion quality tends to nosedive on such threads, or more accurately, be buried under unrelated (frequently political, emotionally-charged, and/or low-effort) comments.


I disagree. Usually they have a lot of comments because they are worth discussing.


meh, Have patience at the end of the day its an Affiliate website. Google has to sort through thousands of websites that appear similar and rank them.


Not an expert, but if you add "The Best Cheap" [1] at the beginning of every title tag of your product pages I guess Google bot will think is something shady.

[1] https://www.google.com/search?q=site%3Agoodcheapandfast.com


> No ads, trackers, slideshows or data-hogging images*

No images? Why not hide the description too?

How a product looks is a significant factor when deciding whether to buy it.

Bring back photos if you want real humans to buy from your website.


Google doesn't reward white hat SEO but it does punish black hat SEO. I wish they would do it more often of course but they do totally wreck black hat SEO techniques a couple of times per year.


I have four 10-12 year old domains I was letting expire. This article inspired me to renew them just in case. They used to get good rankings, I assume they’re worth holding on to?


If those domains actually had content on them and are still live sites, then yes, it could be worth keeping. As long as you keep the site(s) live.

But just because a domain name is 10-12 years old, if it never had content and links, doesn't mean it's valuable. A pieceofjunkkeywordkeywordkeywordnonsense.com domain name, even if it's 10-12 years old, is still worthless and only valuable to you.


They had content for ten years up to around a year ago. Then they went down.

So maybe that soured them too much?


It had a one year lapse this past year. So just let it expire?


Sry, but you don't know seo at all. pity for your years lost.


Google ranking is actually real estate. It takes time and works to reach premium position.


so, a google users googles

"cheap smartwatches"

sees a SERP with

"The Best Cheap Smartwatches For Sale on Amazon in 2019 - Good, Cheap and Fast"

https://www.google.com/search?q=site%3Ahttps%3A%2F%2Fwww.goo...

comes to a page like https://www.goodcheapandfast.com/articles/best-smartwatches

which is really slow for the first time mobile user according to PSI https://developers.google.com/speed/pagespeed/insights/?hl=e...

where the headline is "The Best Cheap Smartwatches For Sale on Amazon in 2019"

and then a text blob that nobody reads (when was the last time you googled something and really really read the paragraphs of text? i mean stuff you googled and not hacker news comments. and don't tell me your read the paragraphs of text on stack overflow answers)

knows nothing else what this site is (search users dont know your story, and dont reach your startpage (i.e.v what does the startpage of stack overflow lookn like?)).

then two list items with again textblobs, where the headlines link to amazon. giving the users no incentive to use your site more or google your brand in the future again.

if it looks like spam, uses words like spam and offers content nobody reads on a first user slow page, it probably is not a good search user experience.

please stop doing SEO, start thinking about the search user.

note: i wrote a book about it

https://www.fullstackoptimization.com/b/understanding-seo

reviews on amazon: https://www.amazon.de/Understanding-SEO-Systematic-Approach-...

and .com https://www.amazon.com/Understanding-SEO-Systematic-Approach...

disclaimer: i don't know how googles algorithms work, and i so do not care


It's true that the JS bundle is a bottleneck on mobile page speed.

I didn't set out to do SEO for this site. Rather, I'm building something that I would use. (I like to read the full text of a page, so I suppose I'm an outlier.)


The space you're competing in:

- Consumer product and tech reviews, vs heavyweights like "Wirecutter, Consumer Reports, Cook's Illustrated and Crutchfield" as you mention, and also TechRadar, TomsGuide, Lifewire, etc

- That's not as difficult as credit card or insurance industries, but there are sophisticated players. Do you truly believe your page on "best video doorbells" is a better result than ConsumerReports.org's?

Your site:

- Has 121 backlinks from different domains (great). But Wirecutter has 20,000+

- Has links from big press, which is cool, but topically relevant linking sites is important as well

- Is thin on review pages. See your wireless earbuds page: https://www.goodcheapandfast.com/articles/best-wireless-earb... vs the #1 organic ranking site for "best cheap wireless earbuds" (4,000 searches/mo): https://www.techradar.com/deals/the-best-cheap-wireless-head... (very important)

- Doesn't have many links to individual pages, so you're not seeing page level link signals, mainly just to the home page (very important)

- Is young, and while that's only one small part of it, is still part of it

Google broadly ranks on:

- Matching intent of query to the best page that serves the goal of the user, and is constantly testing - this does not mean they always reward long blog posts, as it depends on the intent of the query

- Topical authority and relevance - combination of keyword usage, topic coverage, etc

- Seems to reward topically relevant sites more than generalist sites, see the About.com split into different entities case study (though not always)

- Quality of backlinks at the page level, and also the domain level

- Relevancy of backlinks at page and domain level

- Quantity of backlinks at page and domain level

- Other link factors like referring link anchor text, placement on referring page, etc

- On-page optimizations (title tag, keyword usage in content, internal links, headers)

- Tons more, but the above cover ~40-90% of the factors, depending on query, industry, competition, intent, etc

Congrats on the press and growth!

If you truly want to optimize for organic, look at what competitors are doing, page by page, and model off them, with your own unique angle. Content & links, content & links, and repeat.


All true (and thanks for the detailed response!). I wasn't expecting the world, but, I did think the site would see some traction by this point based on other experiences that I've had.

Funny aside: My pre-sabbatical day job was working for the publisher of Tom's Guide and TechRadar.

Tom's Guide is a point of pride for me. It was a fledgling site when our company acquired it in 2013 and I had the pleasure of watching its organic traffic grow 9,900% and hit 40MM visitors in a month. Man, did that feel good!


it's really really hard to rank post 2014 era just because you want it to work. The internet is a big-boys-only club now. Things were easy 10 years ago and those who benefited from the free ride by Google SEO, Facebook and Twitter around a decade ago, are now multi millionaires and billionaires and lecturing everywhere how to build a great business while it wasn't actually their unique genius that made it happen in the first place.


> It's really really hard to rank post 2014 era just because you want it to work.

You can blame grey-hat and black-hat SEO for that. It's also why the quality of Google Search more generally has tanked, especially for non-trivial searches: it seems that Google is desperately trying to always play it safe and give zero incentive to potential spammers, even if this screws some users over in the short term. Bing and DDG are a lot better these days if you're looking for something highly specific, weird or uncommon, while Google is surprisingly fine for common, even trivial/mindless searches.




Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: