Planeta Huerto has sold organic food, cosmetics, and home products online for 15 years, and has been part of the Carrefour group since 2018. Today, 85% of its sales come from Spain and Portugal, with a growing share from the UK, Germany, Italy, and France.
That growth had outgrown the platform behind it, a single homemade system built to run both the storefront and the back office. Pablo Sánchez, Planeta Huerto’s co-founder and chief executive, rebuilt the whole operation on Shopify in three phases, without losing a day of sales along the way.
- Grew revenue 30% year-on-year once the new platform stabilised.
- Runs a catalogue of up to 485,000 references, 15,000 of them its own, on a single Shopify setup.
- Cut €16,000 a month in infrastructure costs by moving off self-hosted infrastructure.
The challenge: 25 engineers running one legacy platform
Planeta Huerto built its own platform in 2011, when Shopify barely existed in Spain. Over 15 years, that system grew to run everything: the storefront, purchasing, and logistics, all in one proprietary build.
Keeping it running took a technical team of more than 25 engineers, with high salaries and constant turnover. “The business kept moving faster than the technology,” Sánchez said. Because the platform was entirely custom, no outside partner could easily step in when demand on that team peaked.
Layered on top was a catalogue problem few other retailers had. Planeta Huerto’s own warehouse holds around 15,000 products, but once its Mirakl marketplace integration is counted, the assortment customers can buy rises to roughly 485,000 references. Customers routinely build hybrid baskets mixing the two, splitting a single order into two shipments. “There wasn’t a case like ours,” Sánchez said, “first because of the sheer number of products, and then, on top of that, the extra complexity of running a marketplace.”
The solution: a phased migration with no deadline
Planeta Huerto’s first move wasn’t Shopify at all. It split its ERP out to Odoo, a standard market platform, separating purchasing and logistics from the storefront before touching ecommerce at all.
We’re a retail company, not a technology company. We needed technology our sales and product teams could run themselves, without needing technical resources.
For the ecommerce rebuild, Planeta Huerto worked with Nitsnets, a Shopify partner. Having just been through one difficult ERP cutover the year before, the two agreed no fixed deadline and moved in three phases instead: an international storefront first, with nothing to lose since it carried no marketplace complexity, then a B2B portal built entirely to Shopify’s own standard. “I didn’t change a single thing about Shopify’s standard setup,” Sánchez said, “and it’s worked brilliantly.”
The hardest phase came last: migrating its Spain and Portugal stores to Shopify while building Mirakl fully into the new setup, through custom apps and a middleware layer. Not losing a single day of sales was the one condition Planeta Huerto set from the start, and it held.
The result: lower costs and steady revenue growth
A year on, Planeta Huerto runs without a dedicated technology team at all. The budget that once covered more than 25 engineers, plus roughly €16,000 a month in pure infrastructure cost, now goes to new products and closer contact with customers. Recurring small bugs, once a constant, are gone too.
Black Friday planning looks nothing like it used to. Every item on the agenda for this year’s preparation meeting was a business question, not one about server capacity, a shift Sánchez credits to handing the infrastructure over entirely.
You have 100 units of energy, and if you spend 20 of them worrying about IT, that’s 20 taken away from the business.
Day to day, the team also runs price changes and sales analysis through Shopify’s AI Toolkit connected to Claude Code, in plain language, with no code required. Sánchez runs these queries himself.
Revenue grew 30% year-on-year, and Sánchez says conversion is at its highest point yet. Planeta Huerto has also started preparing its catalogue for AI shopping agents, enriching and normalising product data so a model can find clear answers about what it sells, and it is already seeing more traffic arrive that way.