So firstly, thank you so much for joining me in my first ever MozCon appearance. I’ve thoroughly enjoyed it and fingers crossed I’ll be back soon!
When you start planning presentations, you honestly feel that half hour is more than plenty of time – but I could talk about cool shit like this for hours!
I couldn’t squeeze anymore in to my MozCon slot without making throwing the schedule out, but I wanted to leave you with a just a couple more cool ways custom extraction can offer some super CRO value and insights. Oh, I also said I’d wrote a thing – and this is it; let’s get to it!
Looking for my MozCon slides?
If you missed this year’s MozCon, or just want your hands on a copy of my slides, you can download the PDF version by clicking here.
Find patterns amongst products that don’t sell
In the world of ecommerce, there’s always a handful of products to be found that receive healthy levels of traffic but for one reason or another, sales are poor or non-existent.
Using custom extraction, you can pull all of the product’s key data sets at scale. For example:
- Product reviews
- Price
- Stock levels
- Titles and product descriptions
- Product image(s)
You can then pair this data with your basic behavioral stats such as exit rate, bounce rate, users, sessions etc and start to see if you’ve any patterns between items with high traffic levels and poor conversions.
Once you’ve this wealth of data and sniffed out a handful of some potential trends, you can proactively work in stripping products of these counterproductive points and directly improve conversion etc.
Could it be that a large number of products that don’t sell have poorly written descriptions, reviews of 3 or less, poor product images, low stock, poor product title? Once you’ve your magic formula – you can get to work!
Identify how many reviews and what ratings typically drive the strongest sales
Wouldn’t it be great if we could identify what review ratings and the minimum quantity of said reviews drives the most sales and greatest conversions amongst your products? Well, by crawling the items and extracting their aggregate rating along with the quantity of ratings, you have the first half of what you need.
Secondly, you’ll need to extract your sales performance data and pair this to your review data you’ve just extracted to get the other piece of the puzzle. However…
Be careful…
Depending on the size of the ecommerce site in question here, aggregate review scores can change all of the time. It’s important to use only fresh sales data to make your data as accurate as possible. This will reduce the chances of pulling sales stats for items that may have had a different score a month or so prior
You might be surprised as to what this reveals…
Mr Andi Jarvis of Eximo Marketing here in the UK has some incredible insight in to the power and perception of product reviews. Specifically around what actually makes us choose one product over another when it comes to aggregate review scores.
In summary, more people are likely to purchase an item that has a lower aggregate rating but has more reviews than a higher rated item with fewer reviews. Andi presented this talk at last year’s (2018) Searchlove and you can find the slides here
Check to see whether there’s a GTM/GA snippet on every page
Custom extraction isn’t just about lifting data from the front end of a web page. Providing it’s in the rendered HTML, you can get at it!
Here’s a story I sorely remember and won’t be forgetting any time soon…
A client of mine had just migrated to a new website and at the time I was quietly confident that all SEO migration considerations were done and completed well.
However, 24 hours later, traffic quite literally nose-dived, hard. Now, as an SEO your first instinct is to fire up a bunch of SEO tools and get digging in to the detail. Screaming Frog, Search Console, server logs, the works and nothing appeared to be out of place. I was literally on the verge of losing my mind and felt my professional kudos and integrity was shot.
After 24 hours of frantically and exhaustively trying to figure out what the fuck was killing SEO for this site and find that ‘magic bullet fix’ I’d ran out of ideas. I had to change my perspective.
Parking SEO for a while, I Instead decided that I would check for the presence of a GTM container on every one of the site’s pages. However to take this a step further, I wanted the container ID too and here’s what I discovered…
Every single page on the site had a GTM container, but PDPs (Product Detail Pages) was using the staging container, which had a different GTM ID. This was immediately switched to the correct ID and within 24 hours, traffic not only returned to normal, but was actually up 10%!
Custom extraction is this case made finding this issue possible, which actually wasn’t an SEO issue at all.
The lesson learned here is that not all migration issues are always down to SEO and custom extraction can help shed light on these edge-case areas.
Bonus tip – cache warming using a SEO crawler
it’s absolutely no secret that page speed has a direct impact on CRO and sales performance.
Anyway, back to my point here. Many websites leverage web page caching to slash load times and improve page speed. Nothing new there. However, caches can have their downsides; one of which is getting a freshly launched or amended page cached successfully on to a CDN server and ready to serve cached pages to visitors globally.
Using a cron job and an SEO crawler, sitemaps containing key URLs can be scheduled to run on a daily basis or as you see fit to ensure the cache on the CDN is always fresh and fast. This can help to improve site speed on key ecommerce pages such as product and category URLs from a global standpoint, a great way to further improve CRO and site engagement.
Fili Wiese from Searchbrothers detailed this clever strategy at this year’s (April 2019) Brighton SEO. Deepcrawl have dropped a cool re-cap on Fili’s talk titled, “Why I adore Sitemaps, an ex-Google engineer’s love story” and you can check it out here.
Enjoy my presentation and want to get in touch?
Whether you’ve questions around my topic, ecommerce SEO/CRO or want to discuss how we could potentially work together; I’m all ears! Please get in touch using my contact form, or reach out to me on Twitter or LinkedIn.
Looking forward to hearing from you and thanks so much for taking the time out to check out my post! 🙂
Comments are closed.