Case Study
×
COLLECTIVE SHOES • NEW ZEALAND • 90-DAY ANALYSIS

Search Drove 17% of Revenue
with 30% Higher LTV

9,000 orders. Order-level attribution. Full customer journey tracking. Here's what we found.

90-Day Results
NZ$237K
Search-Attributed Revenue
+30%
LTV Lift
79:1
Revenue ROI
<1 day
Payback Period
USD $1,000/month investment • Order-level attribution • No projections
The Navigation Tax

She wants comfortable heels for work. You sell them. Ziera. $200. 45% margin. Best product in your catalog.

She'll never find them.

Not because they're hidden. Because your navigation makes her work for it. Click into Women's. Click into Heels. Scroll past 4 pages. Hope "arch support" is a filter. It's not.

7 clicks. 6 minutes. She leaves.

This is the navigation tax. Every catalog pays it. The bigger the catalog, the worse it gets.

Search removes the tax. She types "comfortable heels for work." Ziera shows up first. 47 seconds. Done.

That's what we tested.

The Setup

We installed semantic search. We tagged every order with attribution data. We tracked complete customer journeys from first search to purchase.

Two cohorts. Same store. Same period. Same products.

Search Cohort
Customers who used search before purchase
n = 1,539 orders
Browse Cohort
Customers who purchased without search
n = 7,461 orders
01

Revenue Attribution

Search captured 17% of total store revenue.

Metric
Value
Total Revenue
NZ$1.4M
Search-Attributed Revenue
NZ$237K
Search Revenue Share
17%
Search Orders
1,539 (17% of total)

Industry benchmark: Search typically drives 10-15% of e-commerce revenue. Collective is outperforming.

02

Order Economics

Search cohort: +9.1% AOV, +18% UPT.

Metric
Search
Browse
Lift
AOV (Average Order Value)
NZ$154.29
NZ$141.43
+9.1%
UPT (Units Per Transaction)
1.38
1.17
+18%
Multi-Item Order Rate
22%
11%
+100%
Why This Happens

Customer searches "navy loafers". Below the results she sees: leather care kit · matching belt · no-show socks.

She came for loafers. She leaves with three items.

Browse customers don't see contextual suggestions. They buy the loafers and leave.

03

Conversion Velocity

Median time to convert: 1.2 minutes.

Time to Purchase
% of Search Orders
< 1 minute
47%
1-5 minutes
32%
5-30 minutes
13%
> 30 minutes
8%

79% of search customers converted in under 5 minutes.

Why This Happens

She types "comfortable work shoes for standing all day".

The results understand intent, not just keywords. First result: cushioned insole, arch support, slip-resistant. She clicks. She buys. 47 seconds.

Browse customers click through 6 pages looking for the same thing.

04

Customer Lifetime Value

Search cohort LTV: +30% higher.

Metric
Search
Browse
Lift
LTV
NZ$309
NZ$237
+30%
Orders Per Customer
1.42
1.21
+17%
Repeat Purchase Rate
28%
17%
+65%
Time to Second Purchase
18.3 days
19.7 days
-7%
Why This Happens

She searched "wide fit sandals". Found them in 2 minutes. They fit perfectly.

Three weeks later, she needs winter boots. She doesn't browse. She searches "wide fit boots". Same store. Same result.

When the first purchase works, she comes back. 65% more often.

05

Margin Quality

Search cohort pays full price more often.

Metric
Search
Browse
Full-Price Penetration
42.7%
39.9%
Avg Discount Depth (when used)
12.8%
14.2%
Why This Happens

She searched "red heels for wedding". Found the exact pair. $179.

No hesitation. No discount code needed. This is what she came for.

Browse customer scrolls through 40 pairs. Not sure which one. Waits for the 20% off email.

Search customers find what they want. They don't need convincing.

06

Session Depth

Deeper sessions = higher AOV.

Session Depth
Orders
AOV
Lift
1 search
847
NZ$142.15
Baseline
2-3 searches
412
NZ$158.91
+11.8%
4-5 searches
189
NZ$171.34
+20.5%
6+ searches
91
NZ$198.67
+39.8%
Why This Happens

First search: "summer sandals". 30 results. Too broad.

She refines: flat heel. Then: leather. Then: wide fit.

Six searches later, she's looking at 4 perfect options. She buys two. $198.

Filtering isn't friction. It's confidence building.

07

Incremental Revenue

NZ$119K from SKUs with zero browse conversions.

Metric
Value
SKUs sold exclusively via search
655
Revenue from search-exclusive SKUs
NZ$119K
Top search-exclusive brand
Ziera

This is the most conservative incrementality metric. These 655 products generated zero revenue through browse navigation. Search was the only path to purchase.

Why This Happens

Ziera orthopedic heels. $200+. Page 4 of "Women's Shoes."

Nobody browses to page 4. But someone searching "heels with arch support"? Ziera shows up first.

655 products only sold through search. They were always in the catalog. Browse couldn't surface them.

08

ROI Analysis

Return on Investment
79:1
Revenue ROI
40:1
Conservative ROI
<1 day
Payback Period

Revenue ROI Calculation

Metric
Value
Atomz Investment
USD $1,000/month
Quarterly Investment
USD $3,000 (≈ NZ$5K)
Search-Attributed Revenue
NZ$237,000
Revenue ROI
79:1

Conservative ROI Calculation (incremental revenue only)

Metric
Value
Provably Incremental Revenue
NZ$119,000
Quarterly Investment
USD $3,000 (≈ NZ$5K)
Conservative ROI
40:1

Payback Period

Calculation
Result
Daily search revenue
NZ$2,633 (NZ$237K ÷ 90 days)
Monthly investment
USD $1,000 (≈ NZ$1,670)
Payback period
<1 day

Even using only incremental revenue (NZ$119K), payback occurs within 3 days.

Total Business Impact (90 Days)
NZ$237K
Search-Attributed Revenue
NZ$119K
Incremental Revenue
NZ$19.7K
AOV Lift Value
NZ$110K+
LTV Lift (Projected Annual)
09

Portfolio Extrapolation

What this pattern looks like across multiple brands.

Collective Shoes: NZ$1.4M quarterly revenue. 17% search attribution.

Portfolio Size
Quarterly GMV
Search Revenue Potential
5 brands
$7M
$1.19M
10 brands
$14M
$2.38M
25 brands
$35M
$5.95M

Assumes similar catalog complexity and customer intent patterns. Results vary by category and catalog depth. The bigger the catalog, the bigger the opportunity.

Managing Multiple Brands?
Let's talk portfolio pricing and build your own case study.
Start the Conversation →
Why Search Customers Outperform
This isn't magic. It's selection bias—in your favor.

Search customers know what they want. They just can't find it. Give them a way to ask, and they:

Find faster
47 seconds vs 6 minutes
Buy more
2× multi-item rate
Pay full price
42.7% vs 39.9%
Come back
65% higher repeat rate
10

Objections Addressed

"Is this actually incremental, or would they have bought anyway?"
Conservative answer: NZ$119K is provably incremental—655 SKUs with zero browse sales.

Full answer: The +9.1% AOV lift and +65% repeat rate lift are behavioral changes, not channel shifts. Same customers, different experience, better outcomes.
"Our catalogs are different."
We've audited 200+ stores. 78% fail basic discoverability requirements. Premium products buried. Long-tail invisible. Navigation designed for 50 SKUs serving 5,000.

The bigger your catalog, the worse this problem gets.
"Our customers don't search."
10-15% of e-commerce visitors use search. They convert at 2-3× the rate of browse visitors and generate higher AOV.

You're not missing search customers. You're losing them to friction—or to competitors who make it easier.
"We already have search."
Keyword search matches words. When she types "comfortable work shoes," keyword search looks for products with "comfortable" and "work" in the title.

That's not what she asked. She asked for comfort. Arch support. Cushioned insole. All-day wear.

The search bar is a text box. What matters is what happens after she hits enter.
11

Implementation

15 min
Setup Time
Zero
Engineering Required
Day 1
Time to First Results
Shopify
Platform
Statistical Validity
Analysis Period90 days
Total Orders9,000
Search Cohort1,539 orders
Browse Cohort7,461 orders
Attribution MethodOrder-level tags
Journey TrackingSession-based

Every metric traces to tagged transactions in Shopify. No projections. No modeled estimates. Actual orders.

Two customers. Same store. Same day. Same products.

One browses. Clicks around. Maybe buys. Maybe leaves.

One searches. Finds exactly what she wants. Buys. Comes back.

The search bar didn't change your products.
It changed how customers find them.

Let's Build Your Case Study
We'll show you exactly what's possible with your catalog:
Which SKUs are invisible to browse navigation
Revenue attribution potential at your scale
Specific queries customers are typing (and failing to match)
See Your Catalog's Potential →
15 minutes. No dev work. Limited implementation slots this quarter.