Data-driven e-commerce

Data has always been a part of online marketing and the data-first model should be no stranger to online marketers. Data-first means instead of running campaigns or strategies and then looking at the data to determine success or not, you look at the data first to decide on the direction of your search engine marketing strategy. Traditionally this has meant keyword and competition analyses, but things are quickly changing.

One of the driving forces in this change is Google’s introduction of RankBrain to their search engine results algorithm in late 2015. RankBrain is their machine-learning artificial intelligence (AI) system. Early this year it was intercepting 15% of all search queries to guess on what the user’s true search intent was and what websites they would like to see (this is an oversimplification, but even Google themselves have admitted they don’t completely know how RankBrain works). By September it was affecting 100% of all search queries. Links, and keywords/content are all still important to your online stores SEO but their importance is may be dwindling as we head into this age of AI. RankBrain very quickly has become the third most important ranking signal behind the former two.

Although Google has stated you can’t optimize for RankBrain, as it’s constantly learning and evolving, the reality is you can. RankBrain is all about the user and creating a top-notch user experience is becoming the new search engine optimization. And now, using data to drive this experience is more important than ever.

One of the top user signals Google looks at is engagement. Google’s free Analytics platform provides some great page data as a default report. It’s a great starting point to see what pages, products, departments, etc. are drawing the most attention. And which ones aren’t. Factors such as bounce and exit rates help you see where you are losing customers. For example, you may have a product or page that brings in a high percentage of organic traffic, but it also has a high exit rate. In this case you have a high click-through rate (CTR), a high engagement rate for that page, but a low overall engagement rate for those potential customers. The key is to figure who these users are, which ones are worth optimizing for and what can be done to keep them shopping.

To begin defining your audience it helps to use Enhanced Ecommerce. Enhanced Ecommerce allows a more advanced break-down of the shopping process, including a graphical checkout funnel to help isolate possible barriers in the conversion path. More importantly, it will enable you to create targeted user segments based on specific shopping behavior.

User segments in Analytics gives you the ability to filter traffic by actions taken on your website. Some good e-commerce segments to get started with include separating out customers who have added items to their cart but never checked out, users who purchase in a single session versus users who take multiple visits to complete a transaction and users who have completed X number of transactions over a certain time period. These can be further broken down into sub-segments by dividing them up into regions, demographics, device type and more.

Using these segments, you can then begin to build audiences. Audiences are very powerful, as they’re users who have visited your site and fit a segment you’ve previously identified as key. There’s even a smart audience feature which uses machine learning to create audiences that are more likely to convert. Audiences have a wide range of use, but traditionally they have been used for remarketing PPC campaigns and A/B content testing.

As Google looks more in-depth at user experience for ranking, sites are using this data to deliver customized on-site content. The latter is a more advanced use, utilizing Analytics APIs (Application Program Interface), it sends audience information back to your e-commerce platform to displaying more specific cross-sale products or site ads for example.

With the recent release of User Explorer containing the previously hidden Client ID’s (the anonymous code that Analytics assigns each site visitor), you can now trace the path of individual users and see their actions for each visit. It’s not yet exportable to custom reports, but you can filter by segment to get a better idea of the paths different audiences take to conversion (or why they’re exiting the sales funnel early).

While user experience data is one (large) part of the equation, keywords and content still play a big role. Using Google’s Search Console you can get data on the queries that are driving traffic to your site and where pages are getting impressions in the search results. Connecting this data to Analytics and cross-referencing with your segments will allow you to identify and create tailored content and shopping experiences for your higher converting audiences. Third-party tools such as SEMRush or Moz will also allow you to see impression and keyword data for your competitors to see what you may be missing or get ideas, for both organic and paid.

This only touches the surface on what data driven strategies can do for your e-commerce business. While learning Google Analytics can seem like a daunting task to some site owners, as it’s interface isn’t necessarily inviting, new tools like Google Data Studio can take allthis data and display it in graphically informative dashboards that will help highlight your site’s next big change. Because if there’s one constant in search engine marketing, it’s change.