Understand your data

Rasida says: “Many successful eCommerce brands miss or have a gap in understanding SEO performance by content group. To accomplish this, you must understand how your product pages are performing versus category pages and assess which part is brand and which part is non-brand. Instead, people focus on a very top-line view looking at seasonality and analysing data on a page-by-page basis. Doing so limits how you view your performance and what you’re able to feed into your strategy.”

How do you break down your different categories of pages? What other categories of pages are worth focusing on?

“The easiest way to break this down is by looking at the URL pattern in terms of setup. Lots of problems can derive from not having the correct setup. It’ll be great If you have the skill set to execute the setup in conjunction with the data team. 90% of eCommerce sites are product pages, meaning your product pages must be good. The remaining 10% will be your homepage, category, content, or editorial, which will become focal points as you move forward. Editorial will be on an association basis in terms of looking at your brand and branding. If you’re looking at editorial content you should avoid assessing it from a revenue perspective.”

What makes a great product page in terms of volume, type, calls to action, and other elements of SEO?

“A good product page can be generated from working closely with your copywriting team. It’s important to establish a close relationship with the copy that’s being produced and understand the tone of voice, brand, etc. You should also look at the search element and determine what a user would look for if they wanted product X. This could also feed into the metadata logic, assessing what order the brand is appearing in, the product total, and whether you’re naming things in line with what people are looking for.

You should avoid using terms that only resonate with a small number of people. Never assume that everyone will understand what you’re talking about, especially when you’re using niche terms. Take a step back and work with your copywriting team, better understand your brand, tell a story, and use data to target users who are looking for your product so you can show up and be associated.”

Has Google changed the way it’s using metadata?

“Regarding how you’re creating metadata, a lot of the time there’ll be a logic in your CMS and it’ll just pull it from your product entry. You should look at whether this is automated and whether you can play around with that, improve the click-through rate, etc. It’ll all be about understanding what’s happening in the process and testing to determine whether you can improve something.

If things don’t go as planned, you can apply your learnings going forward as part of a nice cycle to follow. It’s worth running tests on product pages every six weeks, not in terms of changes but in measuring performance. Six weeks gives you enough time to do post and pre-work one week to see if results fluctuate. When you’re working in eCommerce, there’s so much seasonality. You can experiment with shorter periods but there are too many variables that could influence that. Six weeks gives you enough data to understand whether an initiative works. If so, you can roll it out to the rest of the pages. If not, you can put a pause on the strategy.”

What’s an example of some analysis that produced a result you didn’t really anticipate, but that people can learn from?

“Making a simple change to a homepage meta title. This involved editing the title to incorporate express shipping as a USP. The changes were nothing out of the ordinary. The title had never been changed before, so it was worth experimenting with. We wanted to test the content ordering to see if the order of the message would be displayed in the meta title.

In eCommerce, shipping is one of the biggest factors that determine whether a visitor converts or not. Because of shipping, people often decide to go to a competitor at the last minute. A week after going live with the new title, our clicks were up 80%, and six weeks later up by 160%. These results remained steady for the whole year, which was incredible because all we did was add ‘express shipping’ in our meta title. From this, the homepage increased in click-through rate and kept bringing in those clicks for more than a year.”

How can you isolate that it’s definitely those particular keywords or phraseology that has made the difference? Are you able to do split testing or do you have to run it and compare results before and after?

“We measured about 12 to 16 months of data and then looked at the 30 days prior to that. No other changes took place on the site and there was no seasonality that we saw across the other marketing channels. Everything was really steady. The only change that had been made was this one, and it aligned exactly with when it went live. Of course, there will be external variables, but we isolated it as much as possible and looked at any wider impact it could have had.”

What software are you using to analyse? Are you just looking in Search Console or something else?

“Keep it simple with tests and analyses. These won’t need to be overcomplicated. It’s great to do split tests using an algorithm; look at 100 pages and compare them to another 100 pages for likeness. This can be a useful approach if you have the budget, team, or people to justify spending lots of time on this.

Look closely at what you have and what you can do. Use really simple tests, addressing the pre and post to look at external variables. You can then use the Search Console and GA to understand how it’s impacting KPIs like revenue. You can then look at how it’s impacting traffic.”

Where does SEO fit into editorial or is editorial more of a brand play?

“Editorial is interesting because, depending on your industry, there might be millions and millions invested into it. They’ll often be significantly more invested in this than marketing, so it plays a role in terms of branding. When you’re producing content online, you want to be reaching an audience so people can see the content you’re producing. That way you’ll be making use of the investment and will be more likely to see a return. However, when working for a brand it can be very difficult to gain any commercial value through content. You won’t only be competing with other eCommerce brands but also large entities with huge marketing/editorial budgets. If you’re unable to have a clear focus it can become difficult to get a return.

It’ll be more about the brand and using it through email, on-site, and on social media. If you’re looking to establish a hierarchy, and your product pages and sites are doing well, then you can look into editorial. If your site is not doing well, you should avoid investing your money there and instead look at the more transactional pages.”

What’s the importance of writing everything down from the previous year that didn’t work before forming your strategy for the coming year?

“As SEOs, we all experience the same problems. For example: not having enough tech or people not listening to you in the business to get budget sign-off. In the midst of this, it’s easy to become so overwhelmed with everything that you forget to reflect on something you’ve done that you probably shouldn’t repeat. You should focus on getting support from your team or manager to understand how you can do something better.

Marketing SEO doesn’t need to be overcomplicated. It’s about narrowing down to the things that will make a difference. That could be the way you communicate: whether you need stakeholder buy-in, who to speak with, whether you’ve got enough exposure, etc. These are questions you must constantly ask yourself - and get help. Asking for help means you’re good at your job and willing to embrace feedback to improve. That’s why writing down and reflecting is so important.”

How long do SEO strategies normally last and how often do you revisit and hone them?

“That largely depends on your business cycle. A lot of the time it’s annual, but in terms of longevity of your strategy, you should be looking closer to two, three, or four years ahead. When you assess performance at the end of the year, you should look at things to determine what works, what doesn’t, and what you should continue to do going forward.

Should you keep optimising content or carry on working with teams to better understand where you need to focus your time? It’s like having an ongoing strategy that’s open to yearly changes. We’re operating in fast-evolving markets where it’s important to adjust your strategy in line with changing market conditions. If your business vision has changed, then your strategy will need to change.”

What shouldn’t SEOs be doing in 2023? What’s seductive in terms of time, but ultimately counterproductive?

“Though it’s good to understand how algorithm changes impact your site, you shouldn’t get lost in this. You could end up thinking about the algorithms, creating panic and assuming you must change everything.

First, you need to truly understand what’s happened and how it’s impacted you. How big is the issue and to what extent has it impacted your commercial KPIs? If it hasn’t, then it’s not worth paying much attention to. Focus on vision, strategy, where you want your brand to be, and the story you’re telling. This will save you from panicking at the prospect of algorithm updates. If you can move away from this and focus more on the long term, when something happens, you can assess the risk and change your vision.”

Rasida Begum is Product Manager of Search and SEO at Yell, and you can find her by searching ‘Rasida Begum’ on Linkedin.