How NordikHome Cut Search Costs 5x and Boosted Discovery by 34% with AACsearch

"Switching to AACsearch was the easiest infrastructure decision we've made. We cut costs by 5x, our search is faster, and the built-in analytics showed us product gaps we didn't know we had." — Erik Lindström, CTO at NordikHome

Metadata

Company: NordikHome — Scandinavian furniture and home goods retailer
Industry: E-commerce
Size: 50,000+ products across 3 storefronts (SE, NO, DK), 200k+ monthly search queries
Previous solution: managed search providers Growth Plan (~$499/mo)
Use case: Product search on e-commerce storefronts (3 country-specific sites)
AACsearch since: January 2026

1. The Problem

NordikHome runs a multi-brand e-commerce operation with three localized storefronts serving Sweden, Norway, and Denmark. Each storefront had its own product catalog with country-specific pricing, availability, and promotions — but they all shared a single managed search providers Growth Plan subscription at $499/month.

As their catalog grew and seasonal traffic spikes became more pronounced, three critical issues emerged:

Unpredictable costs. managed search providers's per-search-operation billing meant every promotion, holiday campaign, or flash sale increased their bill significantly. During the December holiday season, their managed search providers bill spiked to over $1,200.

Limited analytics on the Growth plan. The managed search providers Growth plan included only basic query tracking. NordikHome couldn't see which searches returned zero results, which products were frequently searched but not found, or how search behavior varied across their three markets.

Complex multi-tenancy. Running separate managed search providers indices for each country required custom middleware. Adding a new market (e.g., Finland) meant provisioning a new managed search providers index and reconfiguring the middleware layer.

"We were spending more on search infrastructure than on hosting our entire storefront," says Erik Lindström, CTO at NordikHome. "And we still couldn't answer basic questions like 'why are customers leaving without buying?'"

2. The Solution

After evaluating self-hosted AACSearch, AACSearch Cloud, MeiliSearch, and traditional search backends, NordikHome chose AACsearch for three reasons:

Predictable flat pricing — $99/month for 50k documents with unlimited search operations
Built-in analytics — no additional cost for search analytics, zero-result tracking, and trend reporting across all markets
Scoped API tokens — each country's storefront gets isolated search collections with a single AACsearch account

The implementation was remarkably fast:

Migration: Exported existing products as JSON from managed search providers → bulk-uploaded to AACsearch index. Total time: 2 hours.
Widget: Dropped the HTML/JS search widget into their Next.js storefront, replacing managed search providers InstantSearch. Time: 1 hour per storefront.
Configuration: Set up typo tolerance, faceted search for categories/sizes/colors, and custom ranking rules prioritizing in-stock items. Time: 1 hour.
Scoped tokens: Created 3 tokens for SE, NO, and DK storefronts, each scoped to their country-specific product catalog.

"We were fully live in under 4 hours, including importing our complete product catalog," says Lindström. "The scoped tokens meant we didn't need any middleware — each storefront just uses its own token."

3. The Results

| Metric | Before (managed search providers) | After (AACsearch) | Improvement | | ------------------------- | --------------------------------- | ----------------- | --------------------- | | Monthly search cost | $499/mo ($1,200 peak) | $99/mo | 5x cost reduction | | Search latency (p95) | 45ms | 8ms | 5.6x faster | | DevOps time for search | 4h/week (monitoring) | 0h/week | 100% reduction | | Zero-result searches | Unknown | 12% → 3%* | 75% reduction | | Time to launch new market | 2-3 days | 15 minutes | 96% faster | | Product discovery rate | 22% | 29.5% | +34% improvement |

*After analytics insights showed top zero-result queries, NordikHome added synonym groups and new product tags, reducing no-result searches from 12% to 3%.

The bottom line: NordikHome saved $4,800+/year in direct search costs, eliminated all DevOps overhead, and improved product discovery by 34% — all in one afternoon of work.

4. Technical Details

Documents indexed: 52,000 products across 3 collections
Features used: Faceted search, typo tolerance, custom ranking, scoped API tokens, analytics dashboard
Integrations: Next.js storefront, custom API for catalog updates
Languages: Swedish, Norwegian, Danish, English (product data)
Migration path: managed search providers JSON export → AACsearch bulk import