Is this Searchengineland article optimal for site architecture?

[phatmike10]

Hi, I’m going through a re-education process at the moment as I haven’t done any site architecture devilment for a few years now. I read the following article in searchengineland which has a diagram on how to structure a site for SEO:

Strategic SEO & Website Design - How Website Structure & Information Architecture Will Rocket You To Your Business Goals

It’s a traditional silo structure but I’m not sure it’s an optimal way of structuring a site in the real world. We all know it’s incredibly difficult to get links to category pages. I would say that it would be sensible to add the subtopic level (subcategory) to a sitewide CSS menu instead of adding it in a silo. So for example if you hover over the “bags” link, links to “handbags”, “shopping bags”, etc appear.

This means the main categories and subcategories are on the same link level and receive sitewide link juice. The downside is you split link equity across more pages but the benefit is any link juice on any page of the site is passed onto category pages and the subcategories. It is also one less click step for the user. Plus it means you can add links to product pages from the main category (covering products from all subcategories of that category) as well as products only being linked from subcategories. I’m guessing in the author’s diagram when the user clicks on “bags”, there aren’t any products listed but instead a list of subcategories. Otherwise, if there were products listed in the “bags” category page, the amount of link juice passed to the subcategory pages would be greatly reduced if there are many products listed.

I just wanted to know whether you agree with my logic or whether I’m barking up the wrong tree!

[joshz]

The logic is spot on in my opinion.

[realityhack]

Your website should be designed to help the user do whatever it is that you do.
I sell widgets. My website is 100% about helping people purchase them. Sure you make sure that within that framework you don't violate any SEO rules, and you can add information sections, etc.
But fundamentally if you get 10,000 visitors at .5% conversion or half that at 2% conversion which makes you more money?
Which is likely to build more long term customers?
Which is more likely to get 1 person a month to link to you building long term links without trying?

Your navigation should match your product and how users want to interact with it. After you have that, worry about SEO.

[phatmike10]

Agreed, I would look at content strategy before analysing the site and architecture in terms of SEO. If any SEO results in a poor user experience then it isn’t really SEO these days as user experience is becoming a large factor in gaining SERPs.

In terms of my original post, I was keen to understand if that kind of approach to architecture in that scenario is really optimal for a site selling products or whether my suggestion would be more optimal for that scenario.

[realityhack]

Optimal architecture can vary greatly with what product you are selling. In clothing for example you will want to group by men's boy's etc as people are used to doing in the store. Making the navigation as invisible as possible. In selling art you might need to be able to browse either by artist, or type of piece, or price level as your starting point, which means 3 top level navs.
For clothing color may be on the page resulting in 6 skus on one page. For vacuum cleaners you might have strictly 1 sku per page.

Start by thinking about how people shop for your product. How do retailers organize that section (could be right or wrong but think about it), other web retailers (ie. is there something people expect - again could be wright or wrong but you need to know).
Ask other people how they would go about it, or have them surf existing sites while you watch.

Get a good book or 5 on interface design. Check out alert box.

You will have warring desires to keep the number of choices per page reasonable, and the number of clicks to target low. The balance can sometimes be hard to achieve.

[phatmike10]

Exactly that makes a lot of sense. In fact, that leads me onto how many products per category page you should have and how to make category pages more engaging and to increase natural linkability. Deciding how many products should be in a category page largely depends on the products being sold. The answer can be found from understanding your market, but this can be done with minimal cost such as looking at the competition and spotting trends. Once you have an idea of how many products should appear per category page, you might want to split test to see if it works.

In terms of making category pages more engaging and increasing desire for links to them. A content strategy of mine is to provide helpful consumer guides which relate to my products. I have a section on the site where users can access these guides but I was also thinking of adding links and teasers to these guides next to the product links.

I think this might produce results for those who sell products where there’s a moderate degree of decision making involved in the buying process. Is anyone else doing anything similar?