As a result of a recurring audit, New Look identified a voluminous number of out of stock (OOS) product pages appearing for their seasonal queries in the SERPs. These pages had low click through rates and ran the risk of damaging the domain's competitiveness. Forward3D was assigned to sample some of the OOS pages and provide suggestions on how to treat them
Popular product pages can achieve a high ranking for specific keywords, which becomes problematic when an item becomes out of stock and yet continues to outrank items that are available.
Since most products pages dynamically go out of stock and then have new inventory come in, it can be difficult to measure the impact that these pages have on overall revenue. Based on a specific set of test URLs from New Look in the ‘dresses’ category and a week’s revenue figures however, we identified that product pages that were out of stock for at least one day in a week had organic bounce rates 11.98% higher on average than in-stock pages. Additionally, these pages had a conversion rate 6.66% lower than in-stock products, and resulted in an 11% overall drop in revenue per page across the whole site.
For a retailer the scale of New Look, the number of OOS pages can quickly mount and a coherent strategy is required that maintains SEO standards without compromising on customer experience.
After considering all the options available and given the scale of New Look’s website, we decided that redirecting out of stock items to relevant category (sometimes referred as “PLP”s or “Grid”) pages would provide the largest overall benefit. We collected a list of 4,200 OOS pages and provided a map to redirect these to 150 category landing pages that we believed would provide the best user experience.
After 3 days, a significant uplift in revenues and visits had already been measured, along with improved keyword rankings, which reached its peak one week after the initial redirections were put in place. After analysing over 10,000 keywords, Forward3D measured the number of keywords OOS product pages ranked for to have dropped by 15%, largely replaced by the category page URLs.
The number of keywords that category pages started ranking for improved by 11.19%. These consisted of rankings that were passed along from the out of stock products, as well as exposure to new keywords.
Within 9 days, these category pages gained 7.14% more visits and revenue figures improved by 4.5%. Top target pages (those with > 3,500 OOS product pages redirected to them) have improved their average order value by 4.62% and their conversion rate by 12.6%.
Overall, the OOS redirections have also had a positive impact on pre-existing keyword rankings. 12% of the keywords tracked have now entered the top 10 results, compared to only 3% of keywords which dropped from the top 10 results.
66% of keywords that were already on the first page of results still improved in ranking by 1.64 positions. This compares to the remaining 18% of keywords that drifted in ranking (but still stayed on the first page) and dropped by only -1.38 positions on average.
New Look’s website improved its overall competitiveness, with the category pages (that were redirected to) outranking the OOS product URLs. Also, the number of keywords that these pages ranked for increased significantly, resulting in an uplift in both revenue and visits.
Due to general competition and fluctuation, rankings shifted towards long-tail keywords, counter-balancing the seasonal drift in rankings of generics.
Since previously sold products can be attributed with value other than their physical ownership (such as the generation of on-site product reviews, design and straightforward site navigation), how to most effectively deal with OOS product pages will depend on the scale of your website and the nature of your business.
Based on the data generated from this project however, when dealing with a large scale ecommerce site, it is highly effective from an SEO perspective to redirect OOS pages to relevant category or product pages with the aid of an overlay explaining the redirection to the user.