Schema.org shipped a release at the end of February and almost nobody talked about it. We almost didn't notice ourselves; the changelog was buried under a half-dozen unrelated updates and the Google Search Central blog covered exactly one of the changes. We went through it last week with three of our local-business clients in mind. Most of it is noise. Three pieces are worth your time.

Here's the short version, in priority order.

LocalBusiness now wants areaServed with structure

The biggest change for small businesses is that areaServed is now explicitly typed for geographic objects rather than free-text strings. If you were doing "areaServed": "Atlanta, GA and surrounding metro", that still parses, but Google's enrichment of the local pack favors structured values now. The recommended shape uses an AdministrativeArea or City nested object, optionally with a GeoCircle for service-area businesses that work within a radius.

What to do this month: if you're a service-area business (plumber, electrician, mobile vet, consultant who travels), update your LocalBusiness markup to include either a list of City objects or a GeoCircle with geoMidpoint and geoRadius. We've seen early movers pick up extra impression share in the local map results for adjacent-suburb queries, which is exactly the long-tail traffic these businesses can't get any other way.

A minimal example for an Atlanta-area HVAC company would set areaServed to an array of City objects covering Roswell, Sandy Springs, Alpharetta, Marietta, and the city of Atlanta itself, each with a containedInPlace back to the State of Georgia. Five lines of JSON-LD. Real traffic upside.

Review markup got stricter

Schema.org tightened the requirements around Review and AggregateRateing in a way that doesn't break old markup but does change what gets eligibility for review snippets. The headline change: a Review now requires both author with a typed Person or Organization and a datePublished. If you've been generating reviews dynamically from a third-party platform and stripping the author name for privacy reasons, Google will start ignoring those reviews for rich-result purposes.

The practical fix is small but it requires a sweep through your review markup. If you're pulling reviews into the page from a service like BirdEye or Podium, make sure the author field is populated (first name plus last initial is fine) and that you're emitting an ISO-formatted datePublished. We had to fix this on six client sites in March. None of them lost rankings, but the star ratings disappeared from their SERP listings until the next crawl.

Also: itemReviewed is now strictly required when the review is for a service rather than a product. The old habit of putting the review JSON-LD on a page and assuming the entire page was being reviewed will no longer fly. Be explicit. Point itemReviewed at the specific service, product, or organization the review covers.

FAQ schema is mostly dead, except for one thing

This is the change everybody got half right. Google quietly removed FAQ rich results from generic site eligibility in late 2023 and only ever kept them for "authoritative government and health sites." That's still true. What's new in the Q1 schema.org update is that FAQPage markup is now considered useful again, but in a different way: AI-driven SERP features (the AI Overviews, Google's about this result panels, and the generative search experience) are explicitly consuming FAQPage structured data even when no visible rich result appears.

In other words: you won't get a stars-and-questions block on your SERP listing, but the AI Overview that's eating your traffic is now more likely to cite your FAQ content if it's marked up properly. Given that AI Overviews are otherwise terrible news for SMB sites (see our piece from last week), this is one of the few levers we have to be a cited source instead of a forgotten link.

Keep your FAQ markup. Add it to FAQ pages, support pages, and the bottom of service pages. Make sure each question is a real question your customers ask, not keyword-stuffed nonsense. The model is reading these now in a way Googlebot never quite did.

The rest of the release

We read the rest of the changelog so you don't have to. Most of it is plumbing: a new MedicalCondition subtype here, refinements to EducationalOccupationalProgram there, two new properties on Event that almost nobody uses. There's a deprecation notice on a handful of older properties that have been redundant for years; if your CMS is auto-generating markup that includes provider on a WebSite object, that's harmless but no longer recommended.

Nothing in the rest of the release moves the needle for a normal small-business site. If you've audited your markup against the three changes above, you're current.

The one habit that beats every update

Run your sitemap through Google's Rich Results Test once a quarter. Not the Schema Markup Validator — that just checks syntactic validity. The Rich Results Test tells you what Google's actual parser thinks about your markup right now, with the current rules applied. We do this on every active client site on the first business day of every quarter. It takes about an hour. It has caught regressions we'd never have noticed otherwise, including a Shopify theme update that quietly broke product markup across 400 SKUs.

If you do nothing else with this article, put the quarterly Rich Results audit on your calendar. The schema.org updates will keep coming. Most won't matter. The ones that do will show up in that audit before they show up in your traffic chart.