Thoughts from the SearchPilot team

It has been hard to miss the buzz around the latest natural language generation (NLG) capabilities recently, as the OpenAI group released machine learning models trained on ever-larger amounts of written content. This has culminated, at the time of writing, in the beta release of an API to GPT-3. In my opinion, the latest ML models can now create useful content that is worth testing for SEO and user benefits.

If you asked a UX professional whether adding upsell items to a checkout process results in more sales or more abandoned carts, they’d tell you to test it — second guessing users is seen as foolish. Yet for many companies, SEO has no testing at all, just endless reams of best practice and hand waving. Our senior consultant, Dominic Woodman, had the opportunity to change role for a year and got the chance to treat SEO differently, running over 50 SEO A/B tests across a range of different websites.

Across most verticals, websites commonly will have text below the fold that predominantly exists to target search engines, not users. We’ve all seen it and we all know hardly anybody reads it other than Googlebot. Unsurprisingly, John Mueller has previously advised against including this kind of text. The comment, shared in a webmaster hangout, was in reference to e-commerce websites, but it applies more broadly: Many e-commerce websites optimize their categories by adding a big chunk of text under the product listings.

As much as we raise alarm about the constant change in Google SERPs, us SEOs love new things with FAQ schema being the latest example. Some are reporting how much they love it, others are saying they’re losing clicks because of it. Here’s a handful of the Twitter buzz: Does winning FAQ snippets improve click-through rates? Naturally, we wanted to test it using SearchPilot to find out for ourselves. In the vertical we have been testing in, commercial pages winning FAQ snippets have quickly started to dominate the landscape.

One of the important features of SearchPilot is taking care of the statistics and measurement automatically. Any time you set up a test, the platform will automatically: Create a forecast of expected traffic to the group of pages you’re testing on Create statistically similar buckets of control and variant pages Measure the organic traffic to the two buckets Calculate whether any difference in traffic is statistically significant When we first built the tool, we used a measurement model that was based on the paper Inferring Causal Impact Using Bayesian Structural Time Series, which was published by Google.

)-- The SearchPilot platform is most famous for SEO A/B testing and more recently full-funnel testing. But fewer people are familiar with one of the other main features; the ability to act as a meta-CMS and change pretty much anything you want in the HTML of your site, without help from your development team or writing tickets. SearchPilot is platform independent, sitting between your website servers and website visitors, similar to a Content Delivery Network (CDN), as shown in the below diagram.

In How Split Testing Is Changing Consulting, Will sums up why high priority SEO changes linger in developer backlogs, and how we’re addressing these issues with our SearchPilot platform that allows us to test and roll out these recommendations without using our clients’ developer resources: we can substantiate best practices like H1 changes, alterations to internal links, and rendering content with and without Javascript. Let’s get started with three tests you should try to see if you can increase organic traffic to your site.

Until now it’s not been possible to measure the impact of SEO and CRO at the same time. Today we’re proud to announce a new SearchPilot feature that we’re calling full funnel testing. SearchPilot launched with a focus on SEO testing. You have probably thought about this by comparing it to tools like Optimizely that allow you to do CRO testing. If you want to know more about how SEO testing works and how it’s different to CRO, you can read more in this post on what is SEO testing.

Without a structured testing program, our experience shows that it’s very likely that most SEO efforts are at best taking two steps forward and one step back by routinely deploying changes that make things worse. hbspt.forms.create({ portalId: "7378679", formId: "9fa7f5d6-e043-4119-a3fc-4c3d4580b0f2" }); This is true even when the thinking behind a change is solid, is based on correct data, and is part of a well-thought-out strategy. The problem is not that all the changes are bad in theory - it’s that many changes come with inevitable trade-offs, and without testing, it’s impossible to tell whether multiple small downsides outweigh a single large upside or vice versa.

One of the most common questions I get about our SEO A/B testing platform is “why do you measure the results using search traffic instead of rankings?”. Three main reasons: Clickthrough rate (CTR) is a huge factor in how much organic traffic you get, and many SEO tests change how your site looks in the search results, and hence affect CTR. No position tracking tool can measure this Rank tracking can never cover the full tail of key phrases and will miss or misrepresent cases where the effect is different in the head vs in the long tail