Results From the Field

The Problem

Phytesia is Europe's leading producer of hardy orchids, with an aggressive expansion into the UK and German markets. Their product catalogue was large, specialist, and growing — but organic visibility was near zero. UK organic daily visits sat at 20. A half-time FTE was manually enriching product content one SKU at a time. The UK Home & Garden market requires page-1 Google rankings to drive any meaningful volume — and Phytesia wasn't there.

What We Did

MashnLearn integrated IBM Watson's NLP and text-mining stack directly into Phytesia's product catalogue workflow. The AI scraped, structured, and enriched product content at scale — identifying top-performing keywords, generating SEO-optimised descriptions in English, and creating category-specific product pages. We also built a PPC-optimised content layer and coordinated the Google Shopping feed. Implementation: 10 days.

Multi-brand catalogue enrichment directly inside Odoo. MashnLearn processed product data for three brands simultaneously — Liewood (Scandinavian baby), Jonzac (organic cosmetics), Banwood (children's bikes) — generating localised product descriptions and pushing them into the Odoo product master. The engagement demonstrated the ERP-native integration model that is now the core of our coaching engagements.

The Problem

Teatower runs its product master in Odoo — SKUs, pricing, stock, and seasonal blends, across retail, wholesale, and POS — but sells direct-to-consumer through Shopify. Keeping the storefront catalogue and collections in sync with Odoo by hand means every new blend or seasonal change gets entered twice, in two data models, in every language the storefront serves — a cost that compounds with every new SKU and every new market. Teatower also migrated its storefront from PrestaShop to Shopify, which carries a well-known risk: platform migrations are one of the most common ways an established site loses organic rankings, because URLs, metadata, and structured data get rebuilt and nobody notices what broke until traffic already dropped. A redirect is never as good as the original URL — even a well-matched one passes less signal than the page it replaces — so the goal wasn't "add redirects," it was minimizing how many were needed and how lossy each one was.

The manual cost, priced out: 2 hours of product copy plus 1 hour of translation for each of the 4 non-source languages — 6 hours per SKU for one market-ready, multilingual listing, or €150 at a €25/hour rate. Across the 314 products live on the storefront today, that's roughly 1,884 hours — €47,100 — of copywriting and translation, and it recurs every time a new blend launches.

What's Running

Odoo's product.template stays the operational source of truth. The Shopify storefront runs a curated direct-to-consumer subset of the Odoo catalogue, with collections structured around Teatower's product categories rather than hand-curated lists. SEO is maintained continuously, not just at launch — the July 2026 audit found and resolved 23 on-page SEO issues (title length, duplicate H1s, missing meta descriptions, image alt text) across the live storefront, and collection-level SEO titles and descriptions are kept current as the catalogue evolves.

Time to market is part of the pipeline, not an afterthought: every new product page and blog post is submitted to Bing and Yandex via IndexNow the moment it publishes, across all 5 locale variants, plus a social bookmarking pass (Pinterest, Flipboard, Reddit, Quora, Medium) that gives new pages inbound signals from day one instead of waiting on organic discovery. A new SKU goes from Odoo entry to indexed, multilingual, discoverable storefront listing in one pass — not a queue.

The PrestaShop→Shopify URL migration itself was AI-matched, not a blind redirect table: 441 legacy URLs were analysed, and 329 were mapped to their exact Shopify equivalent with high confidence — the rest routed to the closest available match. Even done carefully, that's still redirects, not identical URLs, which is exactly why the migration cost real ground regardless (below).

What This Proves

• Your ERP stays the operational source of truth — no forced re-platforming of the business to gain a modern storefront.
• SEO equity survives a platform migration instead of resetting to zero — the redirect map and the audit are the evidence, not a promise.
• Multilingual scales without linear headcount — five storefront languages maintained from one source catalogue, not five parallel content teams.

The alternative most teams default to: a generic Odoo-Shopify connector app that syncs data fields and stops there, leaving content, SEO, and collection curation as permanent manual work — or no integration at all, and a spreadsheet doing the job by hand. Both get more expensive with every SKU added.