SearchPilot: FAQ content lifted traffic, FAQ schema did not

Most claims about FAQ content and SEO are observational: a site added FAQs, traffic went up sometime after, and the two get linked. SearchPilot, a controlled SEO A/B-testing company formerly known as Distilled ODN, ran two actual split tests that isolate the FAQ variable itself. One tested the content; the other tested the schema markup around it. The results point in different directions, and SearchPilot’s own published case studies, not our analysis, are the source for both.

What each test controlled for and found#

Test one: adding FAQ content to footer copy. SearchPilot added page-specific FAQ content below the existing footer copy on a set of ecommerce product listing pages, addressing category-level questions. The control group kept the existing footer copy with no FAQ content added; product grids, filters and internal linking stayed identical across both groups. The test delivered a positive result at the 95% credible interval, with an estimated 9.7% increase in organic clicks. SearchPilot attributes this to the added content letting the page match a wider set of long-tail and informational queries it wasn’t previously targeting.

Test two: removing FAQ schema markup, keeping the content. In a separate test on a different ecommerce customer’s product listing pages, SearchPilot stripped only the itemprop, itemscope and itemtype microdata attributes from an existing, valid inline FAQ block, leaving the visible FAQ text untouched. Removing that markup produced no statistically significant impact on organic traffic in either direction. SearchPilot’s own reading: since Google stopped showing FAQ rich results for most sites, the schema wrapper no longer does anything measurable to organic performance, positive or negative, even though it remains a valid Schema.org type.

Neither test discloses an exact run length. Both published case studies describe the change, the control/variant setup and the statistical result, but not the calendar dates the test itself ran for. That is a real gap in what SearchPilot has made public, not something this piece is filling in.

Why a split test outweighs a before-and-after comparison#

The comparison happens at the same time, not across time. A before-and-after read on a single site can’t separate the effect of the change from whatever else happened in the same window: a core update, a seasonal traffic swing, a sitewide template change. SearchPilot instead runs a control group and a variant group concurrently on pages of the same site, so both groups experience the same external conditions and only the tested change differs between them.

The result carries an explicit confidence level. The FAQ content test isn’t reported as “traffic went up 9.7%”; it’s reported as a positive result at the 95% credible interval, a statement about how likely the observed lift is to be real rather than noise. The schema removal test is reported the same way, as a null result: no statistically significant difference, not “we didn’t check.” That distinction, a designed test with a stated confidence level versus an anecdote with a percentage attached, is the entire reason these two numbers are worth citing and Deloitte-style before-and-after retail figures need the caveats our page speed coverage already gives them.

A null result is still a result. Most published SEO case studies only report wins. SearchPilot publishing an inconclusive test, FAQ schema removal moving nothing, is itself evidence: it means the FAQ content’s traffic contribution in test one is coming from the text Google can read and match to queries, not from the structured data markup around it.

Why we care#

FAQPage markup lost its Google rich result on 7 May 2026, as we covered separately. That removal was a policy change: Google stopped showing a feature. These two SearchPilot tests answer a different question the policy change alone doesn’t: whether the underlying FAQ content and its markup were doing anything for organic traffic in the first place. The content evidently was, by a measured 9.7%; the markup, on its own, evidently was not. For anyone deciding whether to keep writing FAQ copy now that it earns no snippet, that split is the actual answer: keep writing the content for the queries it can still match, and treat the schema as optional plumbing rather than a traffic lever.