Costs
Part of On-page SEO: a focused business guide for 2027
Case studies: what Rakuten and Monster India actually changed on-page
Rakuten Recipe expanded Recipe markup and Monster India piloted JobPosting structured data, with both detailed in Google's own official case study library.
What to take away
- Rakuten Recipe expanded its Recipe structured data through CMS templates, then validated the output before traffic moved.
- Monster India piloted JobPosting markup on part of its Indian inventory, checked the results, then scaled to the rest.
- Both accounts are Google-hosted success stories, so treat the traffic figures as publisher-attributed, not measured by an independent party.
- Copy the mechanism, not the resultsame page type, same supported markup, same visible-to-machine agreement.
- Keep the evidence record next to the decision so someone else can reproduce your reasoning.
On-page SEO case studies are useful for one thing: seeing which lever a team actually pulled. The two named below, Rakuten Recipe and Monster India, sit in Google's own case study library.
Read them for the intervention, not the percentage. Neither figure was audited by an outside party.
Rakuten Recipe: Recipe markup built into the CMS
Rakuten Recipe runs user-submitted recipes, so thousands of pages repeat the same fields: ingredients, cook time, yield and rating. Google's case study describes the team expanding the Recipe structured data across that page type.
Recipe markup fields and validation
- recipeIngredient
- recipeInstructions
- cookTime
- recipeYield
- aggregateRating
- Rich Results Test
- Schema Markup Validator
The template shipped fields Google documents for Recipe, among them recipeIngredient, recipeInstructions, cookTime, recipeYield and aggregateRating. Rakuten pushed the markup through its content management system rather than editing pages by hand, then validated the output.
Google's Rich Results Test and the Schema Markup Validator are the two tools for that check. Markup that fails either one never reaches the index as a rich result, so Rakuten checked its template against Google's requirements before rollout. Search Console then showed which pages were eligible and which were not.
Rakuten and Google attribute the reported gains in search traffic and session duration. Ask what else shipped that quarter, because a template release rarely travels alone.
The case study page reports a percentage for search traffic and for session duration; read the exact figure there. It is the Rakuten Recipe structured data case study in Google Search Central's case studies library, and the publication date sits at the top of that page.
Monster India: JobPosting markup, piloted then scaled
Monster India lists job openings, a page type where Google supports JobPosting structured data and surfaces listings in Google for Jobs. Google's JobPosting documentation requires fields such as title, description, datePosted and validThrough. It also requires hiringOrganization and jobLocation.
Pilot then scale JobPosting markup
- Pilot JobPosting markup on part of Indian inventory
- Check the results
- Scale across the rest
- Use Indexing API for crawling
The case study says the team piloted JobPosting markup on part of its Indian inventory, checked the results, then scaled across the rest. Google also documents the Indexing API as a way to ask for crawling of job posting URLs instead of waiting on the normal schedule.
The pilot is the transferable decision. A job board has millions of near-identical pages, so a markup error multiplies fast. Running it on a slice first means the fix is cheap.
Monster attributes the reported change in organic job-detail traffic and applications to that work. Job listings expire, so a stale posting that stays marked up as open is a bad result for the searcher and a risk for the site.
The case study page reports a figure for organic job-detail traffic and for applications; read both there and treat them as publisher-attributed. It sits in the same Google Search Central case studies library, under the Monster India Google for Jobs entry, with the publication date at the top.
What Rakuten Recipe and Monster India have in common
Both companies marked up information that was already visible on the page. Neither invented content to satisfy a rich result. That is the line Google draws, and crossing it is what gets markup ignored or penalized.
Both also changed a template, not a page. One page fixed by hand teaches you nothing about the other ten thousand.
A worked transfer test
Say you run an events site. Your event pages carry dates, venues, ticket status and prices, all visible. Pick twenty events, not two thousand.
| Step | What you do | What you record |
|---|---|---|
| 1 | Pick 20 live event pages that already show dates, venue, ticket status and price. | The URL list and the pull date. |
| 2 | Record the current baseline per URL: impressions, clicks, average position in Search Console. | The baseline table and the 28-day window. |
| 3 | Add Event markup to the page template only, matching the visible fields. | The field mapping. |
| 4 | Validate all 20 URLs in the Rich Results Test and the Schema Markup Validator. | Pass or fail per URL, plus the error text. |
| 5 | Re-check the same window after rollout, then apply the scale-or-stop rule you wrote down first. | The before and after numbers, and the call. |
Worked transfer test for events site
- Baselinesnapshot impressions, clicks, ticket actions
- Interventionadd Event markup to template
- Check with Rich Results Test
- Measure cohort against held-back group
- Decidescale, fix, or stop
Set the scale-or-stop rule before you see the numbers. Otherwise you will rationalize whatever came back.
Audit the published claims
The FTC advertising substantiation policy requires a reasonable basis before a company disseminates an objective advertising claim. That applies to claims you publish about your own SEO work.
The FTC guidance for marketers using reviews covers fake feedback, selective requests, conditioned incentives and undisclosed relationships. Read it before you quote a vendor's case study as proof.
For anything jurisdiction-specific, including how these rules apply to your business, ask a qualified lawyer rather than a marketing blog.
Common questions
Do these case studies prove structured data works?
No. They show two companies that marked up visible information on a supported page type and reported gains. Google's own structured data documentation is the authority on what is currently eligible. Both case studies sit in Google Search Central's case studies library, and each page shows its publication date.
Why does the pilot matter more than the result?
The result depends on demand, seasonality and everything else that shipped. The pilot is a decision you can copy on Monday.
What if my page type has no supported markup?
Then the transferable part is the method: fix the template, check visible content against the machine-readable version in the Rich Results Test, and measure a bounded group before scaling.
