How iRASSHAi turned recipes into their primary SEO channel
In 2025, 75% of iRASSHAi's non-branded search traffic came from recipe pages. The Paris Japanese grocery says that is how it stopped being a shop only its existing fans could find.
iRASSHAi describes itself in one sentence that explains the whole operation:
iRASSHAi is a concept fully dedicated to Japanese gastronomy, which opened in Paris in 2023. Under one roof we bring together restaurants (canteen and gastronomic), a coffee shop that turns into a temaki and saké bar, and a grocery store, on top of the e-commerce experience.
That grocery sells more than 1,000 Japanese products to a mostly French audience, which means it sells a lot of jars a shopper has never opened before.
How does iRASSHAi use Recipe Kit? It runs a French-language recipe library of 304 published recipes, many of them with video, each built around an ingredient the store stocks and linked to the product page for it. This teardown belongs to the Recipe Kit case study library, where we document how food brands turn recipes into search traffic and repeat carts.
The challenge
iRASSHAi launched in September 2023 with a wave of press, and for a while the only people who could find it were the people who already knew the name. That is a specific commercial problem, not a vanity one, and we will come back to how the team measured it.
The shape of the problem is the catalog. A grocery holding more than 1,000 references in miso, soy sauces, dashi, noodles, rice, matcha and yuzu and citrus is selling mostly to French cooks who have never met most of it. iRASSHAi is physically at 40 rue du Louvre in Paris, but on the e-commerce side it ships across the entire European Union, so the addressable audience is far larger than the one that has heard of the shop.
The miso collection alone runs to 33 products, split across white, red and flavoured miso and filterable by the Japanese prefecture that produced it (live, August 2026)
There is a second problem underneath the first, and that screenshot is it. Japanese ingredients are widely known but not widely understood. A French cook who arrives at 33 misos has no way to choose between them, and a shopper who has just bought their first one needs to be told what to do with it. The recipe does that job and turns an unfamiliar SKU into tonight's dinner.
iRASSHAi put real effort into those collection pages, but they were not what broke the ceiling.
We spoke with Alma Terras, e-commerce manager at iRASSHAi, about how the library gets made and what it changed.
Who actually writes the recipes?
Plenty of food brands outsource recipe development or lean on whatever their agency produces. iRASSHAi keeps it in the building, and then puts it through the kitchen.
As the e-commerce manager and the main person creating and handling SEO content, I'm usually the one writing the recipes, with the help of my assistants. Once a recipe is drafted, our restaurant chefs proofread it to make sure there are no mistakes and that it stays true to Japanese traditions.
That review step is the part worth copying. The person who understands the search opportunity writes the draft, and the people who cook the cuisine professionally sign it off before it publishes. Neither skill is sufficient on its own.
How do you pick which product a recipe is built around?
With more than a thousand references, choosing the hero ingredient for a recipe is a real decision rather than an obvious one. iRASSHAi weighs commercial reality against how the ingredient actually behaves in the dish.
We have more than 1,000 products, so picking the right one for a recipe isn't always easy. It's mostly driven by price point and stock, but we also look at how the ingredient is actually used in the dish. For example, we'd use an accessible 1L soy sauce to make ajitama (marinated soft-boiled egg), but a more premium soy sauce if it's meant as a light topping on sashimi.
Seasonality is a big driver too. We build recipes around what's happening: refreshing dishes with Japanese ingredients during a heatwave, festive ideas around Christmas, and so on. On those windows the results are really striking, because we're meeting a demand that's already there.
You can see both halves of that on the storefront. Cold soba noodles for summer is the heatwave play, and the Christmas gift sets are the December one. Recipe Kit's own data shows the cadence follows those windows rather than a flat calendar: the library added twenty recipes in one February and one in the following March.
A recipe that is really a product demonstration
The udon carbonara is a good example of the whole system in one page. It is a French-facing hook, an Italian dish rebuilt with Japanese pantry staples, and it ranks on the first page of Google in France for its own dish name.
It is also a shopping list. The recipe calls for 400g of fresh udon, which links to iRASSHAi's own udon noodles, and two teaspoons of white miso, which links to its own miso blanc. A cook who wants to make the dish can add both to the cart without leaving the page.
The udon carbonara recipe card. The two ingredients iRASSHAi stocks carry an add button and link to their product pages, and Tout ajouter au panier adds the set in one go (live, August 2026)
That dish is also where the wider ambition shows up, because the barrier iRASSHAi is working against is the belief that this cuisine is difficult, not a lack of appetite.
Japanese cuisine and ingredients are well known, but for beginners or people less familiar with them, they can feel overwhelming, and there's a misconception that they require a lot of expertise. That's true for some recipes, but many others are quite accessible.
We also want to make more room for Japanese ingredients in French and European kitchens: you can cook with miso without making a strictly Japanese dish, our miso carbonara being a good example.
A miso carbonara asks a French cook to change one ingredient in a dish they already make, rather than to learn a new cuisine before dinner. That is a much easier way to sell a jar of miso.
What do you do that the big French recipe sites do not?
This is the question we most wanted answered, because iRASSHAi competes for recipe searches against publishers with far more domain authority and far more staff.
A couple of things set us apart from the big French recipe sites. First, our chefs validate recipes, and some are the ones we do in our restaurants. What you get is genuinely faithful to Japanese cooking rather than an approximation. That's real product and culinary expertise behind each page: with miso, for instance, there are many kinds and we add content to help people choose well (white miso works beautifully with sweets and fish, red miso is better for broths and meat, yuzu-flavoured miso is amazing in a vinaigrette, and so on).
Second, we're deliberate about mobile UX. On most recipe sites you scroll through endless text before you reach the actual recipe, text that exists purely for SEO. We put the ingredients and instructions first, and the cultural context and SEO content come afterwards. The reader gets what they came for immediately.
The second half of that runs against the received wisdom. The standard recipe-site pattern buries the recipe under a thousand words written for a crawler. iRASSHAi puts the recipe at the top and the context underneath, and still ranks.
It holds up on inspection. Load the udon carbonara on a phone and there is nothing above the recipe card at all: no preamble, no personal essay, no stack of ads. The title, the rating, the portion and timing controls, four lines of introduction, the photo, then the ingredients. There is even a cook mode toggle that strips the page back to instructions and ingredients only.
The same recipe on a phone. The card begins directly under the header, and the ingredient list is a single scroll away (live, August 2026)
For a brand, the trade is easy: the shopper who came for a dish gets the dish, and the ingredient links sit right next to it.
Do the restaurants and the storefront feed each other?
iRASSHAi runs Biwan, Shokudo and Kissaba around the same address as the grocery. We wanted to know whether the menus and the recipe library are one operation or two.
We are indeed trying to bring all of our restaurant recipes online, especially the ones that are home-cook friendly and those that are truly iconic in Japanese cuisine. Of course we keep a few secrets, but our main goal is to make Japanese cooking easier and easier to bring into daily life.
Sourcing is a big part of this too. We import directly from Japan, find and select our producers ourselves, and try to bring in products you simply can't find anywhere else in France. It's the same for our restaurant range: imported from Japan and hand-picked by our chefs. So the storefront and the restaurants really do feed each other.
Direct importing is also what makes the recipe library defensible. If a recipe is built on a producer nobody else in France carries, only iRASSHAi can fully satisfy it.
How Recipe Kit supports the system
iRASSHAi has run its recipe library on Recipe Kit since February 2023, about three and a half years as of writing. The shoppable recipe cards are what let an ingredient line link straight to the product page, and the structured recipe schema is part of what helps those pages surface in search.
The published library sits at 304 recipes, growing at roughly nine new recipes a month over the past year. Recipe Kit's analytics make that pace measurable rather than a hunch. Over the trailing ninety days, more than 3% of recipe views ended in an add to cart. Across the 236 Recipe Kit stores with meaningful recipe traffic in that same window, only five ran a higher rate.
The traffic concentrates the way a healthy library does. The ten most-viewed recipes carry more than 40% of all recipe views, so a handful of hero dishes quietly do the heavy merchandising. The most-viewed recipe of the last ninety days is the fluffy Japanese pancakes, followed by sushi rice, homemade gyoza, ajitama and omurice, each tied to ingredients sitting in the catalog.
One more thing worth naming, because the team raised it unprompted. iRASSHAi sells into Germany, the Netherlands, Spain and Italy as well as France, and translation was a real cost line. After a recent Recipe Kit update moved recipe content into a metaobject system that works with Shopify's native Translate & Adapt, and started emitting translated recipe schema on translated storefronts, the brand switched tools. In its words, that cut translation costs significantly while keeping strong indexing on the other markets.
What is actually selling the pantry products?
We asked the team to step outside recipes entirely and say what moves the needle on selling Japanese pantry goods in France right now. Recipes turned out to be the thread running through all of it, but they are not the whole answer.
Email is one of our biggest channels. We use it to tell the full story around a product: sometimes recipe ideas, sometimes the history and tradition behind an ingredient, sometimes the specific Japanese producer it comes from. The better people understand a product, the more likely they are to buy it.
On the retail side, tastings really do make a difference. If you've never tried different kinds of soy sauce, you can't understand how they differ or how they complement each other. The great thing about France is that a lot of people cook at home and love to go deep into ingredients and recipes.
The most transferable idea in the whole interview is what iRASSHAi does with that content afterwards, because the recipe does not stay on the website.
We try to make our content follow the customer everywhere: online recipes, free printed cards in store, the back labels on the products themselves, and detailed product pages. It's all the same idea, teaching Japanese cooking and pointing back to the ingredient, wherever someone meets us.
Each of those surfaces does the same job under a different constraint. The label has the least room, so it gets the anecdote.
Because we import everything ourselves, we also put a lot of work into the back labels of our products. On top of the legal information, we add anecdotes, recipe ideas and context to help people actually use the ingredient. And we go even further on the product pages online, where we can add much more content than a label allows.
The printed card is the neatest loop of the lot, because it carries a shopper from a dish they have just seen to the shelf the ingredients are sitting on.
The recipe content also lives in our physical stores. We print recipe cards that customers can pick up for free, each one flagging the ingredients they'll find on our shelves. It closes the loop nicely: someone discovers a dish in store, takes the card home, and knows exactly which products to buy, whether in store or online.
Online the same loop runs on automation, because at this catalogue size nothing else scales.
Online, every recipe has the ingredient list with add-to-cart buttons that link straight to the matching product. We also use featured collections and a plugin that recommends products based on each blog post's tags. With around 300 recipes, hundreds of products and a small team, we can't match them manually, so we're constantly pushing the automation further.
The homepage carries the same idea in a third format, a shelf of recipe videos where each one is tagged with the product it is built on.
Recipe videos on the iRASSHAi homepage, each one carrying the product it is built around and its price, from koshihikari rice to shiitake dashi (live, August 2026)
What did recipes actually change?
Back to the branded ceiling, and the clearest account of it is Alma Terras's own.
We launched in September 2023 with a big spike in traffic thanks to the PR around the opening. But at that point, essentially 100% of our traffic was branded: the only people finding us were those who had already heard of iRASSHAi. The real challenge was to start existing for people who had never heard our name.
So we worked to get every page indexed by Google and, little by little, climb the rankings. We put a lot of effort into our collection pages (matcha, miso, Japanese noodles, etc.), but recipes are what really moved the needle and helped us break past that branded ceiling.
Then she put a number on it.
In 2025, 75% of our non-branded traffic came from recipe pages.
That is the figure to take away from this story, and it is the reason we frame recipe content as an owned channel rather than a content project. Non-branded search is the traffic a brand does not already have. For iRASSHAi, three quarters of it now arrives through recipes, on pages the brand owns outright, at a cost that does not reset every month.
The team was candid about what the route there took.
The hard parts weren't glamorous. At the beginning, we underestimated how well the products would sell and ran into frequent stock-outs, which is painful when restocking directly from Japan takes around three months. Getting fully indexed took several months, and there's no shortcut for that.
Both of those are worth hearing before starting. Indexing a new domain takes as long as it takes, whatever the budget, and a recipe that works creates demand for a shelf that has to be stocked to meet it.
Lessons for pantry and grocery brands
A short list of the moves that travel well from iRASSHAi to any other pantry or grocery brand:
- Have the cuisine expert sign off the SEO writer's draft. iRASSHAi's e-commerce lead writes the recipes and the restaurant chefs proofread them for accuracy before publishing. Search skill and culinary credibility rarely live in the same person, and the recipe needs both.
- Publish into demand that already exists. Seasonality drives a lot of the calendar here, from cooling dishes in a heatwave to festive ideas in December. iRASSHAi reports the strongest results in those windows precisely because it is meeting demand rather than trying to create it.
- Pick the SKU on price point, stock and role in the dish. An accessible 1L soy sauce for a marinade, a premium bottle for a finishing drizzle. The recipe should feature the product a shopper will realistically buy for that job.
- Put the recipe first and the SEO copy last. iRASSHAi deliberately inverts the usual recipe-blog layout on mobile, ingredients and method at the top, cultural context below. It still ranks, and the ingredient links sit where the reader actually is.
- Take the recipe off the website. The same content runs on free in-store cards, on product back labels and on product pages. The recipe should meet the customer wherever they encounter the brand, not just on the website.
- Give recipes a real navigation slot. RECETTES sits in the main navigation beside ESHOP and RESTAURANTS, not three clicks deep under a blog. That tells both shoppers and search engines the channel is resourced on purpose.
For the deeper thesis on why this works, see the recipe content marketing guide, the rented versus owned land framework, and the explainer on recipe schema markup.
More pantry brand stories from the case study library: how asiastreetfood built a non-English recipe SEO moat, how Melissa's Produce gives every vegetable a reason to sell, how Cabot Creamery publishes direct to its fans, and how Bold Bean Co makes a jar of beans dinner.
What comes next
The obvious threat to any recipe-led traffic strategy is that the click disappears. iRASSHAi is already planning around it rather than waiting to be surprised.
We know recipe traffic is exposed to zero-click search, so we're already thinking about how to keep the product-to-recipe link strong even if fewer people land on the page itself.
It is the right question. If Google's AI Overviews answer the recipe before the reader arrives, the value shifts to the surfaces the brand still controls: the in-store card, the back label, the product page, the email. iRASSHAi already runs all four, which is a reasonable definition of being ready.
Turn your catalog into a library of product demonstrations
Publish structured recipes on Shopify, connect each dish with the products used to make it, and show shoppers what your products can actually do.
