The business question

Better search technology.
But better business?

We invested in a more advanced search engine. Did it improve ecommerce performance enough to justify the investment?
Search qualityUser behaviorConversionRevenueROI
01 / Performance journey

Baseline → Drop → Recovery → Slight uplift

Performance recovered, but recovery is not the same as incremental value.

Legacy performance baseline
Phase 01

Legacy Search

€150K / week

Stable baseline · €7.8M annualized

Phase 02

Algolia Launch

€142K / week

Commercial decline

Phase 03

Optimization

€148K / week

Progressive recovery

Phase 04

Stabilized

€151.5K / week

Slight uplift

LegacyCVR 5.8%CTR 42%Zero results 9.2%Exit 18%
LaunchCVR 5.1%CTR 38%Zero results 7.1%Exit 20%
OptimizationCVR 5.6%CTR 43%Zero results 4.5%Exit 17%
StabilizedCVR 5.9%CTR 45%Zero results 3.8%Exit 16%
Optimization focusSynonymsRankingMerchandising rulesPopular queriesZero-result queriesAvailabilityMobile UX
Search performance / stabilized period
Illustrative case-study data
Search Usage Rate32%of sessions
Search CTR45%↑ 3 pts
Search Conversion5.9%↑ 0.1 pt
Zero Results Rate3.8%↓ 5.4 pts
Search Exit Rate16%↓ 2 pts
Revenue / Search Users€62Kper week
AOV — Search€86per order
Before

Legacy Search

Search CTR
42%
Search CVR
5.8%
Zero Results
9.2%
Search Exit
18%
Weekly Revenue
€150K
VS
After optimization

Algolia — Stabilized

Search CTR
45%
Search CVR
5.9%
Zero Results
3.8%
Search Exit
16%
Weekly Revenue
€151.5K
Search experienceClearly improved
Business performanceAlmost unchanged
02 / Product impact

Search quality clearly improved.

The implementation delivered product value. The remaining question is whether that value became incremental business impact.

Fewer zero-result searches9.2% → 3.8%

Users are more frequently presented with results.

Higher search CTR42% → 45%

Users interact slightly more with search results.

Better search analyticsMore visibility

Queries, clicks, conversions and null results can be analyzed systematically.

Official analytics overview
The analytical transition

Better search does not automatically mean more revenue.

Search Quality IndexEcommerce Revenue
LegacyLaunchOptimizationStabilized

The implementation improved several search-experience KPIs, but commercial impact remained limited.

03 / Causality check

Search users convert more. Did search cause it?

Search users7.4% Conversion Rate
€86 AOV€6.40 Revenue / User
Non-search users3.2% Conversion Rate
€78 AOV€2.90 Revenue / User
04 / Segment analysis

Where did Algolia perform better — and where did it struggle?

Global averages can hide the audiences and categories that still need work.

Device performanceMobile remains the main optimization opportunity.
05 / Search funnel

Where does search intent turn into revenue?

The largest drop-off reveals where optimization should focus next.

100,000Searches100%
45%
45,000Result Clicks45%
30%
13,500Product Engagements13.5%
38.5%
5,200Add to Cart5.2%
51%
2,650Transactions2.65%
06 / ROI analysis

Did the investment pay for itself?

A before-vs-after uplift is not the same as an attributable return.

Illustrative annual investment-€20KProject assumption, not official pricing
Apparent weekly difference+€1.5K€151.5K vs €150K baseline
Naive annualizationDo not use€1.5K × 52 does not establish causality
AcquisitionPromotionsSeasonalityPricingAvailabilityCustomer mixCampaignsUX changes
Controlled estimate+€250 / week

Estimated incremental contribution after controlling for traffic mix, seasonality and major commercial events.

Annualized revenue≈ €13K

Incremental revenue, not incremental profit.

Revenue-level difference-€7K

The illustrative €20K cost is not recovered.

Break-even analysis

Revenue is not margin.

At revenue level, €20K per year equals about €385 per week. With an illustrative 35% gross margin, generating €20K of margin would require about €57K in incremental annual revenue.

Revenue-only simplification
€385 / week
At 35% illustrative margin
≈ €1.1K / week
Algolia investment-€20K
Incremental revenue+€13K
Gross margin contribution< €13K
Net business caseNot recovered

Incremental Revenue ≠ Incremental Profit

07 / The real conclusion

One implementation. Three different answers.

ProductSuccess

Relevance improved, zero-result searches decreased and users interacted more with results.

PerformanceNeutral / slight improvement

Commercial performance recovered, but remained close to the legacy baseline.

Business caseNot yet proven

Measured incremental value does not currently compensate for the illustrative annual investment.

A better product experience and a positive financial ROI are two different questions.
08 / Recommendations

Keep learning before making a removal decision.

The data supports continued targeted optimization, not a simplistic “remove Algolia” conclusion.

01

Optimize high-volume queries

Prioritize queries combining high volume with low CTR or CVR.

02

Prioritize mobile search

Review result layout, filters, ranking, UX and product availability.

03

Reduce zero-result journeys

Continue tuning synonyms, alternatives, redirects and product matching.

04

Run controlled tests

Use A/B tests or controlled rollouts instead of relying only on before vs after.

05

Review ROI again

Reassess CVR, CTR, incremental revenue, margin contribution and cost after three months.

Final message

Better Search ≠ Better Business Performance

Search qualityUser experienceConversionIncremental revenueMarginROI
The goal of analytics is not to prove that an investment worked. It is to measure whether it created enough value to justify the decision.