Episode 57: E-commerce categorisation and faceted navigation SEO with Kristina Azarenko

Play this episode:

Or get it on:

What's in this episode?

Product categorisation and faceted navigation may seem straight-forward for users but it's a minefield for search engines. Without prior planning, it's easy to wreck your chances of building a competitive e-commerce platform. In this episode, Mark Williams-Cook and SEO and e-commerce expert Kristina Azarenko explore what you need to be thinking about, planning and implementing to get the best possible chance of improving your organic traffic.

Show notes

SEO Challenge course -

The framework for building indexable facets based on particular filters -

Transcription MC: Welcome to episode 57 of the Search with Candour podcast. My name is Mark Williams-Cook and today we are going to be joined by technical and e-commerce specialist Kristina Azarenka to talk about one of the areas that causes a lot of e-commerce sites the greatest pain which is categories faceted navigation and pagination. If that doesn't mean a lot to you, stick around because we'll start from the ground up and if this is something you deal with on a daily basis, I think it's gonna be good to listen as well. Kristina, thank you so much for taking the time to join us, really appreciate it. Why don't you take a moment to introduce yourself, your background and tell people a little bit about you, what you do.

KA: Yeah, I'm also excited to join this podcast. I love podcasts and love what Mark is sharing, so I'm excited to be here. I'm an e-commerce technical SEO consultant, based in Toronto, Canada and I help businesses get traffic and get the sales they deserve from Google traffic. I've been in digital marketing and SEO for about ten years and I've worked with many companies, when I was working in house and agencies, but my true passion lies in healthy online stores and this is very timely in all these current situations.

MC: Yeah I thought this was a particularly good subject. So we've had a few episodes recently about obviously coronavirus, covid19 stuff and talking about digitally what's happening, how businesses are changing and one thing that's come out through that is lots of businesses are either upping their e-commerce game, investing more in their SEO for e-commerce, and there’s a whole load of SMBs, small medium businesses, enterprises that didn't have an e-commerce offering before, and are now scrambling to get one.

So I thought this would be a really good chance to speak to someone like Kristina, who's this is her bread and butter, this technical SEO for ecomm sites and I wanted to pick up on her suggestion actually, to talk about this categorisation faceted navigation which is a scary technical end of e-comm in a way. It requires a bit of planning, it's one of my favorite areas for SEO because normally there's a lot of improvements that can be made on most sites that you get involved with. So Kristina I guess if we address those businesses that are starting out now, with an e-commerce site - so they've decided, they've got their products and they didn't necessarily sell online before or maybe, like some of the clients we've had are looking to move onto a platform because they're currently selling on say like an eBay shop. In terms of categorisation, so in terms of we've got this product range, how should they decide what goes where? What's the process? What are the tools? How do they validate it? What kind of process should they start thinking about in terms of how they categorise their products on their ecomm site?

KA: Yeah that's a great question and actually, I also wanted to mention that as you said, ecommerce is booming right now because people who didn't sell online, they were just operating offline now have to switch and in one of the talks I had this statistic and people were fascinated by it. It said that by 2040, about 95 percent of all the sales will be done online. I feel that this 2040 year is already coming because everybody has to migrate their efforts online.

If you are just starting, first of all you should decide which platform to choose and this is a completely huge topic, but I'd say that start with something small, if you have never had an online shop, start with maybe Shopify or WooCommerce and when it comes to understanding which categories you should create; first of all, you should already have some understanding of your customers because if you were operating offline, you can analyse the demand and you can understand what people will search for potentially, but one good thing and it's actually quite a fast thing to do is to look at your competitors, it never hurts. It never hurts to look at your competitors websites or for some websites for partial competitors and see what they have as categories, especially those that you know are doing really well online - this never hurts. So you can have a number of your products and then decide which categories to create but based on these products.

What I would also say is usually people think only about the types of products, if we for an example - baby gifts, random example but baby gifts; you can think about okay maybe baby gifts for girls, baby gifts for boys etc, this is pretty straightforward but I would also suggest thinking about the use cases of your products. So in this example this would be, baby shower gifts or gifts for a first birthday - this might not be a really good example right now but you know, this is the idea - how you look at your products, not only in terms of their type, but also which use cases can be seen for those parts.

MC: So this might be we find as we're using tools to mine data, say like Google Suggest, then we're seeing what people are searching - that they might just be searching by the actual function of the product because they don't necessarily know a brand or they don't know the name of the product they need, they just have a problem they need solving. So you're saying we maybe need to create categories to cater for them as well because I guess if we don't, it comes down to if you don't have the page with the content you want, you aren’t gonna rank for that, so we need to actually think about creating those categories around that as well, that’s what you're saying yeah?

KA: Yeah, yeah exactly because if you have, for example, one category for all it works and can see yeah, these products can be used like this, these products can be used like this but people don't know it and when they are looking for these use cases for all these types of products, there are millions, millions of other websites who have those pages, so your website, your store won't be as relevant.

Here's one more example and I just remembered about this a really quick way to come up with some suggestions as well. For example if you go to Google and look for something which is quite general, for example swing chair, and if you go to images at the top you will see different types of swing chairs that can exist - so if you're selling chairs, again not a very good example…

MC: No, I think a lot of people will be in the garden for summer. So I've actually just done this while you've been talking - so I've googled swing chairs, I've gone to images and we've got indoor, outdoor, garden, living room, wooden, hanging, so it's a really good example. So we're saying you know there are going to be people searching for indoor swing chairs, outdoor swing chairs, garden swings chairs, living room swing chairs, so we need to think about that.

KA: I’ve actually got a swing chair.

MC: Is it indoor or outdoor or a garden one?

KA: So I live in a condo, so it's actually outdoor - in summer it's on the balcony but in the winter, it's an indoor chair - so it's convertible.

MC: That's a new search term there, convertible swing chair. So you've opened yourself up to a can of worms here Kristina because you've dived in and you've mentioned about Shopify and WooCommerce already and one question I wanted to ask you was - so there's a problem on some content management systems where if we add the same product to multiple categories, so this is obviously quite a common thing that you might have - so let me see if I can think of example - yes, so say we take our swing chair example and we have outdoor swing chairs as a category and we also have garden swing chairs as a category - now you would expect an outdoor swing chair maybe to be in both of those categories, now with some ecommerce platforms the way they're structuring URLs is they'll have you know sitecom/ the category names or something like outdoor swing chairs and then the product name and then obviously we get a duplicate URL sometimes created which would be you know forward slash then it'll be garden swing chairs and then the product name. So essentially if we're adding the same product into two categories, we might get duplicate URLs.

Now I know and we won't get into the faceted navigation stuff now, but Shopify I believe handles this by making a product URL - I think they call collections and uses canonical tags so it does create all of the different URLs, WooCommerce I believe for stuff like variations, it just creates new URLs for every variation you have and it leaves you to tidy that up. So my general question here is, you've mentioned Shopify, WooCommerce, are there any particular pros and cons that you could recommend for small businesses/ why would they choose one over the other? Or if you go for WooCommerce, here's a couple of things that you should look out for.

KA: Yeah sure. So this is actually one of the most common issues when it comes to picking content for online stores and this is really weird that all the other popular e-commerce systems that don't really handle this out of the box will allow, so I'd say that if again if you just starting, if you don't want to put a lot of effort and putting all the canonicals in, I’d say just go with a flat structure; website slash and then the product name versus having category name in the URL and as you said, in Shopify you can disable it but also what you should do, no matter which system you use, is to make sure that all your internal links are also at the same as you decided. So for example if you have a product URL without the category, all the internal links should also have these product URLs without the category because I have seen so many situations when, for example, there's a Shopify store and they are handling the product URLs real well, they have just a flat structure but the collections pages have links to these product URLs with collection, product etc - so this is the wrong link and you should make sure that this link is the flat structure like your product has.

You can change it, I believe in the liquid files; it's not really hard but this is a very important thing to realise that your internal structure should have all the URLs which are canonical versions, preferred versions.

MC: So that's something that's much easier to do at the planning and build stage isn't it, because if you go live with the default collection URLs and they get linked to internally and some maybe end up indexed, because as we know Google says canonical is a hint not a directive so they could get indexed, there's like a big perhaps tidying up job you've got to do there afterwards right?

KA: Yeah I see that. I'm a big fan actually if you can do something right from the beginning, do this right from the beginning because otherwise this is this is going to be expensive to redo this things because you will need to estimate them, you will need to have more development time and this is going to be like more costly and I'm for more cost efficient or cost efficient strategies. One thing also, just to getting back to the platforms and this setting; for example in Magento, as far as I remember, in Magento 2 and I think in Magento 1 as well, there is an option to also make your product URLs flat, which means that you won't have any category there because otherwise you'll have as we were talking, you’d duplicate URLs for products, the same product with different URLs.

But also there are some extensions Magento has, lots of extensions, and I actually used to work for a company who developed extensions, so that's why I know a lot of these functionalities that can be added. So there was an extension called Unique Product URL and it automatically configured it to automatically choose only one category the product belongs to as a canonical. So for example, if we have a swing chair in the garden category, then swing chair indoor and swing chair, I don't know, in furniture - you can choose the longest or the shortest path as a canonical, which means that in this case, garden or indoor.

MC: That's really interesting that whichever platform you choose it's likely that you're going to need to tweak things to make it efficient in the long term and to give you the best chance of ranking well and as you say that's kind of surprising because all of these systems sell themselves as you know the e-commerce platform that SEO is built in, you don't need to worry about it, but actually it turns out that all of them have their quirks and they're not necessarily as good as they can be out of the box.

KA: Yeah.

MC: So faceted navigation is something I want to talk to you about already so that I mean we've already talked for like 15 minutes about just categories which is cool, but faceted navigation is probably the more complex subject. Do you want to give just an intro maybe for people who haven't come across that term before, as to what faceted navigation is and why it is an important consideration?

KA: Yeah sure. So when you come to an online store, the best example that I can think about is of course clothes, so when you come to an online store and you're looking for something, this is a huge category - for example women's clothes, but you want to drill down to actually find what you need without spending hours looking for all the different pages. So you choose the colour - for example I want to go with sneakers, I want black sneakers, I choose black colour, then I know my own size so I choose my size, so this helps me to drill down into the specific set of products I'm interested in. This happens with the help of different sorting and filters options, what you usually see in the sidebar or sometimes above the products - when you can choose again size, color, brand all this different stuff which helps you save time and find products much faster.

From the users perspective this is an ideal theme when you want to find something particular, but when it comes to SEO, the problem is all these variations is that in most cases it creates different URLs. So when I add colour there is a different URL, there's some parameter added to the URL. When I add size, a new parameter is added to the URL and if all those URLs are indexed, it happens a lot, so when they are all indexed it means that your website has so many pages, which are not really of high value and wastes basically the same products all over again.

MC: And this is gonna be, even if you've got three or four things you can filter by and each of those have three or four filters, that's a huge number of pages, you know we're talking immediately in the thousands of variations of pages that are really similar. So what's the general, we know that we've got canonical as one thing in our toolbox that we can use to combat when we've got two pages that are really similar and maybe with the same intent. What's your general approach to faceted navigation? What are the best practice rules you go by when you're looking at ecommerce sites?

KA: Yeah so for different ways to handle it and I also wanted to give one more idea why this is bad. The thing is, Google evaluates your website in general, so on the base on the basis of all the pages that you have, so if see ninety five pages of your website are you cookies or think content - this happens in my life and when I joined one of the clients that 95 percent of index pages were all duplicates - this means that your website doesn't look really go to Google. So yeah it means that you need to clean those up and when it comes to best ways to handle it, my best go-to thing is canonicalisation just because it consolidates all the other signals for the pages. So for example, if one page gets a link, an external link from some website, your website will benefit from it and if this page which got the link has a canonical pointing to another page, your website still benefits, but if you use no index or something like that these signals won't consolidate.

But the problem with canonicals is that if your website has millions and millions of pages it can be really costly, in terms of crawl budget. Crawl budget is the rate in which Google can crawl your website and if it comes to your website and spends, I would say wastes all the time on each page with the canonical, which is actually not really a valuable page but it still stands this time, it won't get to the real important pages like product pages, especially if your stop is very dynamic and you have new products coming in stock. This might just hurt your website. So in this case I see like if you have millions of pages then you can use URL parameters tool in Google search console, but be very careful because this tool is,well, you can do bad things with this - I mean you should understand what you are doing and you should read everything carefully when you choosing options. So there is an option to choose a URL parameter and say that it doesn't really change or its swords to the products and then choose that Google should not crawl these pages.

MC: So this is the tool equivalent of a chainsaw, in that it's very powerful but you need to make sure you don't end up chopping off things that aren't meant to be chopped off.

KA: Yeah, exactly. Also Google has this red notice like, make sure that you understand what you are doing before actually using this tool and this is really right because I think that they keep so many examples when people just were using it without real understanding, so that's good that they really reinforced it.

One more thing that you can do is to use no index. It doesn't necessarily save the crawl budget because the page still needs to be crawled, but if you have no index it means that at some point the crawl rate will be really really low, so we will be able to prioritise the more important pages and won't spend lots of the crawl budget for crawling these current class pages. The downside is that again, the signals from external links, from internal links, are not consolidated and you're just losing those if you have them come in on no indexed pages.

MC: So would you say, if it's possible, a good approach might be to access filters on a page that you don't want to indexed, maybe through JavaScript - a way that Google is really going to struggle to get to those URLs or maybe use URL fragments or anything after the hash that Google is just gonna ignore anyway, would that be a valid approach?

KA: Yeah, so I would say in an ideal world, it's better to use aJax for these filters, so that you don't even create any new URL. People can benefit from these sortings and they can drill down to the products they need, but no new URLs are created. At the same time some filters can become indexable facets which means that they will be indexed by Google and they will have value. So should I come into specifics right now or are we going to talk about this later?

MC: Well I guess the question I have that I think would be really good to go over, would be over when we draw the line between when something should and shouldn't be indexed. so if we have like a website say that just sells running shoes,we do our keyword research and we say okay there's lots of people searching for like red running shoes, we'll just say there is I don't know how many there are, but we say that's like a genuine search term that people are looking for, so we decide okay that's maybe gonna be an indexable subcategory - it’s going to be red running shoes and then, of course, we've got our normal filters which might be things like to sort by or filter by price but then there's this, for me, this gray area in the middle which might be something like size for instance. So size to me normally falls into like a non indexable filter category, but then there's the argument of well are there people searching for say, size 9 running shoes and if so, should we not create a page for them. So how do you decide where to draw that line of when something just becomes too specific that it's not worth creating a page, indexable URL for? And it should just be a filter that's not indexed?

KA: Yeah so first of all I would say that I would come from a place of deciding, so when you have all the filters not indexable and then you decide which should be indexable versus when you have all the indexable and design which should not be indexable. So what I did - I have a talk about this and it's a really interesting question because what you can do is analyse, first of all, do your keyword research because it really helps a lot and as you said for some websites size might make sense, sometimes when price, but for some websites these would be just garbage pages, so it's really important to see what works in your specific situation.

Then I'd say if you already have these pages, you can look into Google Analytics, you can see which filters people are using, how they use them, so it can give you a good understanding of what your customers are interested in. Then also use internal search; you should definitely have it, have it configured in Google Analytics - it just takes five seconds - and use search terms. If people are looking for black sneaker size ten and many people are looking for this, create a separate page for this and make sure that this page is indexable, so that people can find it through Google.

I have an approach for one of the clients. I did all the research and I recommended some categories but they had a huge number of filters, so creating one category by one was just I don't know it took two years, so I came up with this framework where I chose particular filters based on keyword research, I choose particular filters these were, if i remember correctly - type, size, and colour or something like that and those filters would become indexable facets - this was for t-shirts. If they have enough products in there when you filter, if they have enough products, then this is a separate indexable URL.

So to give you an example; there is an overall category - t-shirts - then there is an indexable filter which is colour; so we'll have yellow t-shirts for example. In this category yellow t-shirts have 9 products or more, then this is a standalone facet page which we index and which people can find through Google. If this page has two products, then we have a canonical pointing to a parent page, which is going to be t-shirts in this case.

MC: So this is like a framework then that will grow with the site maybe. So as you add more products as well these pages can open up and then become indexable.

KA: Exactly it helps to solve this issue with too many faceted pages which are not really valuable or having pages which are unique, which have real value for users, have them indexable, but to have it all at a scale.

MC: So I think it'd be really nice to finish up on a question actually around numbers of products and that's to do with pagination. So we have probably all seen pagination before, which is when you load a category on an ecommerce site and it's showing you products one to fifty, and then you scroll to the bottom and then it's like you're on page one of 142 or something. So I'll be really interested to hear your thoughts on what you should do if you've got a top-level category that someone can browse to and it's got thousands of products in and what is the best approach to that because obviously, we've got options like making a view-all page, just for BOTS or for users and bots, we've got seeing people given advice about no indexing so far down into paginations, we've got Googlebot supporting lazy load now. What's your preferred approach when you've got that situation, we've got a cat with lots and lots of products in?

KA: Yeah I think that this might be a popular opinion. So do you remember that there was it used to be rel=next/prev.

MC: Yes and I was telling people to use it, right up until the point where Google said they hadn't even been using it themselves for quite a few months.

KA: Yes exactly. Well it's still valuable because some other search engines use it but in terms we talked about Google. So at first, I was freaking out and then I was like it's interesting because nobody actually, I mean in terms of users or in terms of SEOs, nobody actually noticed that something was wrong while Google wasn't using it. So we were implementing, we were recommending to use it and nothing was wrong, everything was fine and then we realised, well not realised, but Google told us, it wasn't even an announcement, that it's not using it anymore and everything was fine.

So my opinion is that I'd say keep the pagination indexed, so don't no index it because eventually - well a better explanation - so each page of the pagination has links to the products and if you have many of these products it's really valuable that Google should discover those products from internal links and internal links also have carried signals from one page to another. So it's really valuable to have Google go through these links and to actually see these links. I'm gonna index pages, even if you have no index follow, as we know at some point Google treats no index follow pages as many external follow pages, so it means you just cut off this part of your website, that's why I say that keep the pagination pages indexed, don't canonicalise to the main page to the main category page don't know index them.

When it comes to view all, this might also be a good option but at the same time, as I mentioned before, you should have internal links the same as your canonical tag. So if inside of your website you link to category without view all and then you canonicalise view all category, it gives conflicting signals to Google. You're basically saying my view all page should become canonical, but at the same time my canonical page should be without view all. So this is also something not really good. So I’d go to your Google search console to see how Google perceives these pages and in most situations you won't even see issues with that. like we didn't see those issues before we actually knew that rel=next/prev was discontinued.

MC: It's a particularly interesting case I find because I think last time, I don't know if has been updated but I read the Google docs on pagination and view all, their recommendation was their research had showed that users tend to prefer landing on a view-all page rather than a view 10 products, and I think that's anecdotally been backed up when I've watched user testing of people on ecommerce sites. One of the first things a lot of them do is change their “how many” on a page, to like the maximum amounts because it's an easier UX experience scrolling and sliding rather than clicking. So they've recommended using the canonical tag back to the view all page because they said users prefer that experience and the overall latency is lower than if they clicked through and reloaded paginated sets. However, that to me obviously has a logical cutoff because if you have a pageant page set of fifty or a hundred pages, no user, well it's unlikely that a user’s going to sit and click through every single one of them, so it doesn't make sense. Like you're to try and load thousands of products onto a single page because that's gonna become a very kind of heavy page then. I guess we're just kind of circling around the classic SEO answer of “it depends.” I think the important takeaway there is there's nothing inherently wrong with pagination, if you've got it and it's working for you and it's working for your users, you don't need to worry about trying to move everything onto a view all.

Have you got any preferred ecommerce platform that you'd recommend to startup businesses? We talked about WooCommerce, Shopify and Magento, is there anyone you think handles all these things particularly well? Is there anyone you'd normally go to? I know we kind of covered that earlier with their quirks, I'm just trying to tease out from everyone, I did from Luke as well, if you have got a preferred platform you do try and get people on?

KA: Well the best answer here is it depends.

MC: It depends.

KA: Yeah, it depends on how tech-savvy you are and how much budget you have. so I'll give a few ideas that might help people choose a platform in their particular situation.

So when it comes to Shopify, this is best for people who are not really tech savvy and that's fine because they make things pretty easy, less flexible - much less flexible, but really easy for people to start and as far as I know they've now expanded their free trial, which is amazing because they are trying to help those people who need to start their online stores quickly.

MC: Is it three months now?

KA: Yes, I believe it’s 90 days, so 3 months. So there you can just get up and running pretty fast, they have many templates, so it's pretty easy if you want to start quickly to start with them.

Then when it comes to WooCommerce, this is a good platform but you should already have some technical knowledge because it's word press, so if you're good with WordPress that's great for you, but you should have some technical knowledge, how to do stuff there because it's pretty flexible, it also has plug ins, so you can get everything best from all the platforms. It's flexible but it's at the same time it's not that expensive.

But if you want to have a really big store and you have developers, if you want maximum customisation and numerous different features go with Magento, because it's yeah, it’s quite expensive, but it has many different features, many different extensions and you can customise your show so much and also you can customise and improve your efficiency as an admin of this store because there's so many things that can help you do this. But you will definitely need a developer, so if you don't have a developer, if you don't have money at this point for a developer, don't start with Magento.

MC: Does a number of products make much difference to you? I mean the impression just from again anecdotally I've seen with Magento, is that it tends to be for sites with more products. Have you seen it running on small sites, as in you know that have got a couple hundred products?

KA: Yeah I’ve seen stores like that, but I see that if you have just yeah a few hundred a couple of hundred products or if you have like ten products, Magento is not the answer, it just just won't be efficient for you it will be costing, so started with this in this case with Shopify or WooCommerce.

MC: Brilliant Kristina, thank you so much for giving up some of your time to speak to us about e-commerce SEO, categories, navigation, faceted navigation, canonicalisation, all the big words that make us sound super clever. It's been really interesting talking to you and we can find out more about Kristina at I believe.

KA: Yes,

MC: And I know because again I neglected to mention at the beginning of the show, so a bit like Luke Kathy, Kristina just edged her way onto my radar because she posts on LinkedIn a lot and I was saying to her, just before we started recording, actually I organically started to notice her posts because I saw she was putting out a lot of good information and I believe you run a SEO challenge course, is that right? Do you want to tell everyone about that just before we finish?

KA: Yeah, that's right. So I'm a teacher by education and I'm an SEO by profession, so I think that this teacher part is coming out of me, that's why I decided to create and as your challenge course, I created it last year. and this is for people who want to learn SEO from the beginning to intermediate level, we talk a lot about a lot of stuff, there is very practical stuff and yeah I just have it now on demand for all the people who are bored and keen on learning something new.

MC: And they can find that at Marketing Syrup as well.

KA: Yes, so it’s at and you will see “SEO challenge”

MC: Cool, brilliant. So I hope people check that out as well, thank you again Kristina for joining us. Our next episode is going to be on Monday the 27th of April. Of course as usual, you can get these show notes and links to everything we've talked about, as well as a full transcription, at and I hope you'll tune in next week.

More from the blog