New data shows that a more than 62% of webpages on e-commerce sites are modified each month

Just as merchandisers change the brick-and-mortar shop window regularly to reflect the latest trends and new items, so do online store webpages. But, how much do they really change and how does this impact your SEM campaigns. Here’s the data.

As you may have read in a previous post, I had data that e-commerce websites are always changing and that can have a major impact on paid search campaigns. In response to that post, some of you asked whether that previous data was really typical. So, I analyzed other sites and now present data from 3 e-commerce websites – all major, successful retailers of apparel and accessories, but with quite different websites and models. For those of you who cannot wait for the answer, yes, all three websites keep changing rather rapidly – nearly 60% of pages were added or deleted, or had their menus or filters changed every month. So, although every website is unique, change is typical for e-commerce sites and your website probably fits to one of these below.

First, let’s revisit the underlying issue: why is the website changing so much? The main reasons are that your website marketing team is constantly designing new landing pages with new product categories and revised filters to better show your product offer, plus your product lines keep changing to reflect new brands or you are sold out of a particular color. That’s great news — your keywords capturing a shopper’s search intent for “on sale” or a color, size, feature or brand, for example, may now be able to land on a more specific product set that better reflects the shopper’s search intent. And that results in delivering a better shopper experience and higher conversion.


The Experiment and Methodology

To capture change the data on each of the 3 websites, our bots visited every page on each website 4 times at monthly intervals.

We classified each page of each website in four groups — three representing change and one for no change:

  1. The page was newly added to the website, for example, a Spring sale page or a brand page for a newly carried designer.
  1. The page had been deleted since the last view, for example, the Fendi handbag page or the prom dress page disappeared.
  1. The page had been changed — the page was still there, but the menus or facets/filters had changed, for example, a new filter for material had been added, or that the women’s cashmere sweaters now have items in red, but not yellow or that the shoe brand filter now doesn’t include Jimmy Choo.
  1. The page was unchanged since the last view.

Note that the analysis focused on product category webpages and excluded differences solely due to which particular merchandise items were displayed on a page.

For the example in my previous blog post, we looked at the changes over a 2-week period for a major apparel e-commerce retailer. Remarkably, only 38% of the webpages were unchanged over the two weeks. 10% were deleted and 22% were newly added. And over half were updated (group #3 above) – the item category name or URL did not change, but a menu or facet/filter within the category page did.

BUT is that typical? People asked us for more data. 

And here it is: data from 3 quite different classes of websites. All are major, very successful apparel retailers, but vary in the range of product categories they carry and their specialty. These differences are reflected in the number of distinct pages on each of the category: the “small site” has about 400, “medium site” about 1000, and “large site” nearly 10,000 webpages. For each site, we sampled four times over three months, resulting in data for 3 consecutive periods, each about a month in duration from late summer to early autumn of 2016 (explicitly excluding the Black Friday and holiday periods since we expected that sales and specials for those periods would skew the results).


New Data: 3 E-Commerce Websites

Not surprisingly, the websites are not the same! But it is interesting to note that each website is relatively consistent with their own changes – that is, each had a relatively similar amount of changes from month to month. Although there are clear variations between each of the periods, each website made roughly the same number and types of changes each month.

And the surprising bottom line: for the 3 monthly periods of the 3 retailers, more than two-thirds of the webpages were changed in all but 1 of the 9 periods.


Nice Data, But What Does It Mean for Your SEM Campaigns? Some Things that You Should Do

What does this finding suggest for campaign manager? Here are a few thoughts:

  1. You need to frequently validate the landing pages for all of your active keywords in a campaign – every month! Even for a retailer with a small website, an average of 17% of the pages were removed from the site and 20% were added every month. In the holiday season, it will probably be even larger. Not validating your landing pages may result in the dreaded “404-Not Found” redirection that clearly results in a waste of the paid search budget and is not the shopping experience your brand wants.
  1. Are you making full use of the filters on your destination landing pages? If you do, the big benefit is that you really do show your shopper the “red cashmere cardigans” you carry. And that is particularly important on small-screen mobile devices. But the validity of those pages can change frequently as your inventory changes. Thus, you need to validate that the filters and filter combinations are still available. For example, if the red cashmere cardigans are out of stock, you do not want to show a page with no items. When items are out of stock, either pause the keyword until there is inventory again or change the destination to show a larger category, for example, the cashmere cardigans you do have.
  1. Reconsider your negative keywords frequently. If a webpage was removed because of a change of product line, then adding a negative keyword may be warranted. For example, if a brand page was deleted because you no longer carry that brand, a negative keyword will prevent the shopper experience of showing the ad, getting the click and disappointing the shopper.
  1. And certainly consider adding more long-tail keywords based on new products/categories and item variations that you do carry. When a shopper searches for blue suede shoes, you want to show the blue suede shoes that you carry. Being responsive to what the shopper searched for does result in a better experience as well as more conversions.

Yes, that is a lot of work and doing it manually can be time consuming. That is why many managers now don’t do these tasks as frequently as their website and campaign require. And that is why more and more campaign managers are looking at automated tools to do the work for them.

At Vioby, we develop solutions for retailers and brands that extend the personalized shopper experience to the SEM touchpoint – improving engagement, conversion and user satisfaction. For each shopper, our automated marketing technology determines and validates the most appropriate, relevant destination landing page on the e-commerce website, using AI algorithms with semantic understanding and machine learning. For more information visit or send us a message.