Silo Structure - Huge Project

[andreint]

Hey guys,

so, I’m in the process of starting one huge project in an evergreen vertical.

I’m primarily looking to build an authority site without focusing on off-page SEO (@Ryuzaki journal style), but I need to structure it properly because it’s going to have tons of content (probably 500k+ words within 6 months). Magazine style.

I have some experience with silos, and they worked great for me with smaller projects in easier niches, but this will be the first time I’m ‘shooting for the stars’, so I’d like to hear your thoughts/suggestions/fixes on my thought process.

Here’s the basic layout

#1 - Main Categories
#2 - Sub-Categories
#3 - Tags (?) - Noindex, just for easier content management

I’m thinking about using standard, physical silos + virtual (content) silos

so, basically:

URL.com/category/sub-category/post
+
inner links funneling people from my pillar posts all the way down throughout the silo.

Content structure:

Each subcategory will have 1 pillar post (T1) covering every single angle of that specific topic (tags). Those pillar posts will be targeting the most difficult kw’s and phrases and I’ll use them for outreach and manual link placements.

Pillar posts will funnel people to my money pages/educational articles (T2) via contextual links.

There’ll also be a lot of social, clickbaity content (T3) for leaking and getting shares/mentions.

My first concern:

- I’m thinking about using pages as silos, instead of wp categories. So, my ‘category’ page (URL.com/category) would be a custom page showing my pillar articles/subcategories + relevant intro text.
- I’m also thinking about using my pillar post(s) as a sub-category.

URL.com/category/sub-category would be my pillar post, and every other piece of content that’s been filed under that sub-category (tag) would look like this: URL.com/category/sub-category/posturl

Thoughts?

My second concern/question:

- How would you link within the silos? Should I consider my main category as a silo, or is a sub-category more viable?


I definitely have a lot more questions, but this is all I can think of at the moment.

Please let me know what do you think about this layout and let’s brainstorm the shit out of this. I’m sure it’ll be helpful to a lot of other builders as well.

Thanks

 

[Ryuzaki]

This is exactly how I do it (down to the T-labels!), with a few differences.

I don't use tags. I considered it to make a giant index for users, but in the end I'm going to let them use the search function. The reason is... where do I stop tagging? I could have 50, 100, 500, 1k. What warrants a tag that my categories and sub-categories aren't covering? And then consider the search function too. My own site will become way too large for tags to continue to be feasible to the level of specificity they are designed for.

URL.com/category/sub-category/post
​

That's how I'm doing it. A lot of people have fetishized the Backlinko style of URL.com/post, but they forget how tiny his site is. That also offers nothing for user experience in terms of chopping the URL to jump back to sub-cat's, etc.

Taking this URL structure and adding in Breadcrumbs is more than enough to achieve a Virtual Silo. Google essentially displays this fact in the SERPs when done right (VERY few do, helps you stand out visually too).

The thing then is, you'll have T1, T2, and T3 things that all are related... are you not going to interlink them even when it makes sense, just to not break your Physical Silo? I interlink them but try to do it in a specific direction. I don't think the Physical Silo has much to do with SEO any more honestly. The Virtual style takes care of that. Then it seems there's a Relevancy Silo, if I'd even call it that. More like a mini-net of related content linking to the "top post in the silo."

If you think about it... in the grand scheme of the entire internet, NOBODY builds silos. Google has to abandon that flat-file build from the 90's in order to do their job almost 20 years later. It's all about interlinking in terms of relevancy. All these sidebar, menu, and footer links destroy physical silo's anyways. That's another reason Google weights contextual links higher than supplementary links.

The idea of having category or sub-category pages act as the silo-head with a PAGE is sound to me. You can even have rolling posts on them if you want, below the content. It's all a matter of using the template customizations provided by most of the big CMS's. But you mentioned using a Page over a Post for this. I'd be careful and test this out, because I don't think you can use a page and then assign a post as a child of a page. Pages can only birth pages, Posts can only have posts as parents... although I'm sure there's trickery you could employ with templates to make it happen.

How would I link within the silo's? I mentioned above I would break that convention. What happens when you have a T3 clickbait get 100 referring domains and 100k social signals and it starts ranking #1 for it's key term? Are you NOT going to link it to push juice and traffic to the T2 money page that it's directly related to? I would, even though they aren't in the same silo. That's where the virtual silo comes into play, and relevancy of the content trumps physical silo's now. It literally replaces that idea so that silo's can even cross domains and ownership.

 

[c4yrslf12]

How would you link within the silos?

I agree with @Ryuzaki. You've got most if it spot on. To answer this question, link naturally. Don't worry about it. If you have your breadcrumbs set and basic links throughout the silo already, it's a non-issue. I would suggest, however, less links in your T1's than your T2's and T3's. If you're going to curate link juice, do it from the T3 up to the top.

In the end, silos are all about relevance of content. Tight knit content held closely will rank, regardless of a non-perfect pyramid structure.

 

[The Kloser]

Then it seems there's a Relevancy Silo, if I'd even call it that. More like a mini-net of related content linking to the "top post in the silo."

I'm testing this as we speak with a set of cluster of articles.

On a side note, I know that mind maps help with the whole initial setup and planning. What about when things get moving pass that and you have a team of writers producing, how you guys track all the content been produced to make sure the initial (inter-linking, silos, CAT topics etc...) planning is been followed and executed properly from both ends?

 

[andreint]

Thanks guys for pitching in!

But you mentioned using a Page over a Post for this. I'd be careful and test this out, because I don't think you can use a page and then assign a post as a child of a page. Pages can only birth pages, Posts can only have posts as parents... although I'm sure there's trickery you could employ with templates to make it happen

Good point. I totally forgot about this part. If someone knows a good and tested (!) way to do this, please share

I'll try to keep the relevancy factor to the maximum by showing only silo-related stuff in the sidebar and related posts section, so the juice will naturally flow within the silo.

I'll probably end up noindex/nofollowing all my footer links except to the 'about' page. This is something I do all the time, since I reserve this section for secondary stuff, like contact, disclosure, tos, disclaimer, etc.

 

[Andrewkar]

I'll probably end up noindex/nofollowing all my footer links except to the 'about' page. This is something I do all the time, since I reserve this section for secondary stuff, like contact, disclosure, tos, disclaimer, etc.

I don't think you should do this actually. You can use footer links to show Google that you are a BIG BRAND by including links to "for press" page, "career" page etc.

Also, is there any strong reason that you must have pages instead of posts? When it comes to WP and site structure it's just unnecessary complication IMHO. Unless you have something special... I don't know what evil things you have in mind

 

[Andrewkar]

Here is a nice looking plugin that will turn your WP site, it's post and pages into static content, that is just regular html and css pages. I don't know if this is what you want, but this plugin should make your website a lot faster and secure than regular WP. However, rotating ads etc. and in general ad management will be a bit more difficult. Also forms, comment system etc. won't work. You would have to create separate page templates for contact forms, and other interactive stuff with right code hard coded.
https://wordpress.org/plugins/simply-static/

 

[andreint]

You can use footer links to show Google that you are a BIG BRAND by including links to "for press" page, "career" page, etc.

Yeah, I'll have a lot of similar stuff in my footer, but I really don't want to rank those pages (write for us, advertise here, etc.). Those will be there purely for users who intend to use them.

Contact forms, disclaimers, privacy policy ... the sole purpose of those pages is to appear as a brand in the eyes of a visitor. They don't need to appear in SE index/be crawlable. That'd be a sitewide loss of link juice.

Also, is there any strong reason that you must have pages instead of posts?

I'll be using posts 99% of the time, but my pillar articles will be pages (need to think this through a bit more), for the purpose of siloing.

Thanks for the plugin - I'll definitely check it out

 

[Andrewkar]

Yeah, I'll have a lot of similar stuff in my footer, but I really don't want to rank those pages (write for us, advertise here, etc.). Those will be there purely for users who intend to use them.

I don't have any solid proof, but I think that those pages, especially careers and for press etc. (the ones big sites have as well) might add a tiny bit to the overall website's SEO score. Just my experience.

Regarding SILO, personally I would go with what @Ryuzaki have said. Also, you can place content on category pages if you want, not all themes will suport that but it can be done by some changes to the templates. There is also a way to remove "category" from the URL (or replace it), it will probably slow down website a bit, and could make some problems while updating WP, but it's something you might be interested in.

 

[Ryuzaki]

I'm 99% sure that page rank juice still flows through a no-follow (and a no-index), based on experience and Googler's having said it at some point. The only difference with no-follow is that since you aren't endorsing the destination website, the juice never makes it there. But it still leaves your page... and slips off into the void.

Otherwise you'd have the entire SEO community no-following every single outbound link trying to preserve juice. Logically it makes sense that it has to be that way, even with no-index. Without this, you'd have the world's most broken and inaccurate link-graph ever. Google collects all of this information, not just what is shown in the index.

My approach for these pages is to find reasons here and there on each page to cycle the sitewide juice back out to particular pages I want to rank. That could be example posts, "Have You Seen:", etc.

Google has made it impossible to save, preserve, and plug holes in the dam with the rest of the algo. If you do, you lose in some fashion. The best thing you can do is let it flow freely but make sure you direct it to go where you want.

 

[Sutra]

@Ryuzaki

Your post just got me thinking...

Users on my site like to click images. So I have all my images linked to a relevant category on my site in order to increase pageviews/ad revenue. I figured that if I left all those image links as "dofollow" that I'd be funnelling tons of juice to the category pages and would seem unnatural. So I nofollow most of the image links (occasionally I leave some as dofollow just to spread the juice around). But since you said that nofollows do still pass juice, am I throwing tons of it away because it's nofollow? Would it be better to set all those image links to dofollow?

 

[Ryuzaki]

I would dofollow them, @Trankuility. You're passing it up to category pages and it will cycle right back down. Of course there's a dampening effect and there's less left over on each step, but imagine your page is at a 0.50.

You link an image back to the category page which is now a 2.00 due to all of these links. Then only 75% survives. It links back to your post in a quick cycle and now your post is sitting at 0.50 + (2.00 x 75%) = 2.00.

Assuming Google knows to only run one iteration as a way to combat these "accidental" wheels, you still benefit greatly. Even with made up numbers it illustrates the point.

I'd watch out for a possible Panda issue of Page Rank Sculpting. Not sure how much of this you can get away with. You might do better to link each picture a different post that is best represented by the image. Or even use captions, which nobody can resist reading for some reason. It helps the skimming process I guess.

 

[Sutra]

@Ryuzaki

Bit confused about the last part. Are you saying that whether I dofollow, or nofollow, the image links, I'm at risk of being hit by Panda because it may appear like I'm sculpting for page rank? So in other words, best not to do it all?

 

[Steve Brownlie]

@Ryuzaki is spot on with how this works (assuming Matt Cutts described it accurately anyway but I'm sure they also have some sophisticated ways to count and look at things but I'm sure his answer adequately describes things). I don't like nofollowing any internal links in your case either because it looks more like you're up to something than if you have your images link to something relevant.

 

[Ryuzaki]

@Ryuzaki

Bit confused about the last part. Are you saying that whether I dofollow, or nofollow, the image links, I'm at risk of being hit by Panda because it may appear like I'm sculpting for page rank? So in other words, best not to do it all?

I shouldn't even have said it. I don't think you'll encounter that problem at any normal scale, honestly. It'd be the same as saying "be careful how much you interlink with contextual links, or breadcrumbs, or nav menus"... I overcomplicated it in my brain in a non-real-world way.

@Ryuzaki is spot on with how this works (assuming Matt Cutts described it accurately anyway but I'm sure they also have some sophisticated ways to count and look at things but I'm sure his answer adequately describes things). I don't like nofollowing any internal links in your case either because it looks more like you're up to something than if you have your images link to something relevant.

Someone else on the forum once brought up the point that it's possible that LESS page rank than normal exits a no-follow link. So by no-following some sitewide footer links, you are wasting some juice but also sending more through the rest of the links that you would have been otherwise. It's a great point but one I've never tried to measure.

I think still, even in that scenario, I'd keep them followed and interlink them to the right money pages.

 

[Sutra]

Ah, ok. Thank you.

@Ryuzaki

One last thing. Total noob question but can't seem to find the answer in google. If I go into the text editor and remove just the nofollow text and leave the rest so it ends up as:

rel=""

Is that ok? I'm hoping you say yes, haha. Because I've been going through and updating this way.

 

[Handel]

On a side note, I know that mind maps help with the whole initial setup and planning. What about when things get moving pass that and you have a team of writers producing, how you guys track all the content been produced to make sure the initial (inter-linking, silos, CAT topics etc...) planning is been followed and executed properly from both ends?

Exactly, this is where I'm struggling as well. I've started with mind maps, but it gets too complicated.

Curently I have a query in Excel which takes all the posts titles, categories, etc. straight from website (from mysql) and formats in a nice way. At least I get an overview of what's posted and where my new posts fit in.

But I'd be realy glad to hear about any better approaches.

 

[Ryuzaki]

rel=""

No clue, honestly. My guess is that it's fine. There's no such thing as dofollow anyways, it's just the absence of the nofollow tag, which is technically what you have there.

What you can do is continue on like that, and in the end do a rel="" search in your database and replace it with nothing. Basically a removal of empty relation tags.

 

[ddasilva]

Just my two cents - If it were me, I'd use a flat URL structure (I.E. domain.com/post-name/) There's probably going to be a point in time where you're going to want to restructure your navigation, categories, etc... And if you do, having a flat structure is going to give you much more flexibility than having domain.com/cat/sub-cat/post-name/.

And I know people may disagree with me on this, but the URL structure of your site doesn't define your silo. Your site structure does.

 

[andreint]

Thanks guys, really appreciate the feedback.

I'll test everything out and will definitely report back with the results. It's going to take a shitton of time to set everything up like I planned, but I'll make sure that I share my findings itt.

 

[Chumbawamba]

Top quality stuff this.

I'm 99% sure that page rank juice still flows through a no-follow (and a no-index), based on experience and Googler's having said it at some point. The only difference with no-follow is that since you aren't endorsing the destination website, the juice never makes it there. But it still leaves your page... and slips off into the void.

That's my understanding as well - and AFAIK, it's been that way since around 2011(?)ish.

Link sculpting definitely still works - but the way to do it is to remove unnecessary links altogether. A link on a page, even with nofollow, will send link 'juice' away.

I wanted to pick up on this part (physical silo):

URL.com/category/sub-category would be my pillar post, and every other piece of content that’s been filed under that sub-category (tag) would look like this: URL.com/category/sub-category/posturl

Is there a reason why people may prefer this, as opposed to having the T1/Pillar post as:
url.com/category/
With then T2 posts as:
url.com/category/sub-category/

Supporting content (blog posts etc) would then fall under URL.com/category/sub-category/posturl/ as above.

Just wanted to get thoughts on what people prefer. Does my method break virtual silos/breadcrumbs

Or, given that we're talking about (relevancy silo) > virtual silo > physical silo, does this point become relevant?

Just my two cents - If it were me, I'd use a flat URL structure (I.E. domain.com/post-name/) There's probably going to be a point in time where you're going to want to restructure your navigation, categories, etc... And if you do, having a flat structure is going to give you much more flexibility than having domain.com/cat/sub-cat/post-name/.

And I know people may disagree with me on this, but the URL structure of your site doesn't define your silo. Your site structure does.

Thanks guys

 

[Capital]

You know what would really help this thread? A link to a live site w/ an ideal silo structure.

I'll try to dig one up... in the meantime, I welcome other examples.

 

[chanilla]

I've wasted a shitload of time in my career nit picking all of this to hell and back. You know what made the biggest difference in my bank account? Doing it the wrong way today, and fixing it later.

The time spent analyzing this to death would probably be a decent dent in the actual work. Every time I "thought" I had the perfect site structure when starting a new project, a year or two in always lead to unplanned changes.

I'm with DDasilva on this one... What I do now is %postname% everything, and then once I have topical themes built out that are the top 3 to 5 cornerstones of the architecture, I then build pages that house everything. (links to the relevant posts, categories, case study loops, etc...)

Sure it's manual work, but the fit and finish is better than using stock categories and whatnot, even if you do add unique content.

 

[andreint]

First silo wrapped up

This should be enough for now, I'll probably expand this silo after I prepare the other 3