**TL;DR.** Mercadito de los Andes, a Bogotá-based marketplace for Colombian cocoa, produce, and gift bundles, migrated from three separate WooCommerce sites to a single Ordiko account with three storefronts. In 90 days they reduced platform costs 88%, improved LCP 62%, and grew organic traffic 4x.

## Background

Mercadito de los Andes was founded in 2019 by Diana Restrepo to connect Andean smallholder cocoa farmers and produce growers with urban Colombian customers. By 2025 the business operated three distinct storefronts:

- `cocoa.mercadito-de-los-andes.example.com` — single-origin chocolate bars and cocoa powder.
- `frescos.mercadito-de-los-andes.example.com` — weekly fresh produce subscription boxes.
- `regalos.mercadito-de-los-andes.example.com` — curated gift bundles combining both.

Each ran on its own WooCommerce install on Kinsta managed hosting, with separate Yoast Premium licenses, separate WooCommerce Subscriptions licenses, separate Wordfence security, and a part-time WordPress developer billing 6–10 hours per month per site.

## The challenge

By early 2026 the operation hit four hard walls:

1. **TCO**: Three Kinsta plans ($90/mo each) + plugin licenses ($120/mo combined) + dev retainer ($600/mo) + Cloudflare Pro ($20/mo) = ~$420/mo before fees.
2. **Performance**: LCP on the cocoa site averaged 3.4s on mobile due to plugin sprawl. INP routinely exceeded 400ms on filter clicks.
3. **SEO**: Google deprecated Product rich results for products missing `hasMerchantReturnPolicy` (rolled out enforcement Q1 2026). Mercadito lost rich snippets on ~85% of their catalog overnight. The schema plugin they used hadn't shipped 2026 fields.
4. **Operations**: Three admin dashboards, three inventory systems, three review queues. Order fulfillment errors crept up to 2.3%.

## Why Ordiko

Diana evaluated Shopify Plus (priced out at ~$2,000/mo for multi-store), BigCommerce Pro ($399/mo with revenue caps), and Ordiko ($49/mo Growth plan).

Ordiko's deciding factors:

- **Unlimited stores on one account**. Three storefronts on a single $49/mo plan.
- **Native 2026 Product schema** including `hasMerchantReturnPolicy` and `shippingDetails`.
- **Spanish-default storefront**. The 8 Ordiko-supported locales include Spanish as the default — important for a Colombian audience.
- **AEO surfaces**. `llms.txt`, `llms-full.txt`, Markdown twins, and AI crawler allow rules built in. Diana was already losing AI search citations to international competitors.

## The migration

The migration ran over 14 days with parallel WooCommerce + Ordiko operation on subdomains.

**Days 1–3: Catalog import**

WooCommerce CSV export for all three sites (1,850 SKUs total). Ordiko's WooCommerce importer mapped variable products, preserved slugs, and ingested category trees. Total import time: 2 hours per site.

**Days 4–6: Schema configuration**

Diana's team configured store-level return policies (30-day return by mail, free for Colombian customers) and shipping policies (free over COP 200,000, three regional zones). Ordiko emitted `hasMerchantReturnPolicy` and `shippingDetails` on every PDP automatically. Rich Results Test confirmed eligibility on a sample of 20 PDPs.

**Days 7–9: Redirects and SEO parity**

The team imported WooCommerce permalink history as 301 redirects, configured blessed facets for high-volume search queries (cocoa by origin region, produce by season), and verified hreflang clusters for `es` and `en` (English secondary market).

**Days 10–12: Parallel QA**

The team tested the full purchase flow on each Ordiko subdomain. They identified two issues — a payment gateway misconfiguration and a missing locale string on the regalos cart — and fixed both before cutover.

**Day 13: DNS cutover**

DNS swapped at 03:00 COT local time. Three sitemaps submitted to Google Search Console, Bing Webmaster, and Yandex. IndexNow auto-fired on subsequent mutations.

**Day 14: Monitoring**

GSC Pages report showed 12 unexpected 404s (all from typo-redirect plugin entries on the old WP sites). Each was added as a 301 in `/seo/redirects` within 4 hours.

## Results

**Performance** (Real-user p75 from CrUX, measured 60 days post-migration):

| Metric | Before (WP avg) | After (Ordiko) | Delta |
| ------ | --------------- | -------------- | ----- |
| LCP    | 3.4s             | 1.3s           | -62%  |
| INP    | 410ms            | 145ms          | -65%  |
| CLS    | 0.18             | 0.04           | -78%  |
| TTFB   | 880ms            | 95ms (edge)    | -89%  |

**Organic traffic** (monthly sessions, Google Search Console):

- Month 0 (pre-migration): 12,000
- Month 1: 18,500 (recovery underway, some short-term dip)
- Month 2: 31,000
- Month 3: 48,000

**SEO surface coverage**:

| Surface                                       | Before              | After                       |
| --------------------------------------------- | ------------------- | --------------------------- |
| Product pages with valid 2026 schema           | ~15% (schema gaps)  | 100%                        |
| llms.txt published                             | No                  | Yes (auto-generated)        |
| Markdown twins                                 | No                  | Yes for all blog posts      |
| IndexNow integration                           | Manual (rarely used) | Auto on every mutation      |
| Per-locale hreflang correctness                | Partial             | Complete                    |

**Cost**:

| Line item                          | Before     | After |
| ---------------------------------- | ---------- | ----- |
| Hosting (3× Kinsta)                 | $270/mo    | $0    |
| Plugin licenses                    | $120/mo    | $0    |
| Developer retainer                 | $600/mo    | $0    |
| Cloudflare Pro                     | $20/mo     | $0    |
| **Ordiko Growth plan**              | $0         | **$49/mo** |
| **Total monthly platform cost**     | **$1,010/mo** | **$49/mo** (-95%) |

Net savings: ~$11,500 per year, redirected into marketing.

**Operational**:

- One admin dashboard for all three storefronts.
- One inventory system (with per-store visibility).
- Order fulfillment error rate dropped from 2.3% to 0.4%.

## What we learned

Diana's three takeaways:

1. **Multi-tenant beats multi-account.** Operating three WooCommerce sites separately was twice the work for the same revenue. One Ordiko account with three stores collapsed the operational burden.
2. **2026 Product schema matters more than we realized.** Losing rich snippets was a step-change in CTR. Getting them back via Ordiko's default schema emission was the single biggest organic recovery driver.
3. **AEO is not optional anymore.** AI search citations from Perplexity and ChatGPT now drive ~8% of qualified traffic. Sites without llms.txt and Markdown twins miss this category entirely.

## Quote

> "We were running three regional storefronts on WooCommerce — one for cocoa, one for fresh produce, one for our gift bundles. Hosting and plugin licenses were eating our margin and our team was spending more time patching WordPress than selling. Ordiko consolidated everything into a single dashboard with unlimited stores, schema that finally got us back into Google's rich results, and a Spanish-default storefront our customers love. Four months in, organic traffic is up 4x and we're paying a tenth of what we were."
>
> — **Diana Restrepo**, Founder, Mercadito de los Andes

---

*Mercadito de los Andes runs on Ordiko Growth at mercadito-de-los-andes.example.com (cocoa, frescos, regalos storefronts).*