ZZashOne Builds Join the pilot
Capped pilot · Up to 5 Shopify merchants

See which Shopify orders came from AI without guessing.

A weekly attribution report that connects observable ChatGPT, Perplexity, Gemini and Claude discovery signals to Shopify orders and revenue.

Unknown If there is no referrer, UTM or customer declaration, the source stays unknown.

Join the pilot →See a sample weekly report ↗

No payment or installation happens when you apply.

ZASHONE BUILDS

Weekly AI-attribution report

FICTIONAL DATA
Evidenced AI revenue£1,462+£524 this week
AI-assisted orders17+5 this week
Share of revenue4.6%+1.4 pts
Unknown-source orders96Left unknown
Open the full sample report →

THE PROBLEM

AI is becoming a discovery channel. Most Shopify reports cannot show its value.

Customers increasingly use AI assistants to research products, compare options and decide what to buy. By the time they reach a store, that journey may appear as Direct traffic or be credited to another channel.

Some tools fill the gap with inference. ZashOne Builds takes a stricter approach: observed, declared or unknown.

  • Did an AI assistant contribute to this order?
  • Which AI source was involved?
  • How much revenue has observable AI discovery produced?
  • Which pages and products are AI-referred customers landing on?
  • Is the signal strong enough to justify more content, SEO or product-data work?

WHAT THE PILOT DELIVERS

One clear report each week

During the pilot, we combine the evidence already available in your customer journey with a short post-purchase question. The first version is delivered as a hands-on service, not an automated app.

01

AI-assisted orders and revenue

The number and value of orders linked to observable AI-referral evidence, split by source where the evidence supports it.

02

The proof behind each number

Referrer, UTM or customer declaration, shown separately so you can judge the strength of the evidence.

03

Landing pages and products

The pages AI-referred customers first reached and the products they later bought.

04

Survey health

Response rate, declared discovery sources and the share of orders where the source remains unknown.

05

Practical next steps

What the evidence may justify doing next, such as improving a page repeatedly reached from AI, checking a rising source over another week, or holding off because the sample is still too small.

See a sample weekly report ↗Fictional data, provided only to show the format.

FROM SIGNAL TO DECISION

Turn a hidden signal into a decision

01

Treat AI discovery as a measurable channel

Put an evidenced revenue number next to Meta, Google, email and organic search.

02

Decide whether to invest

See whether AI-referred revenue is large or consistent enough to justify work on content, product data or pages that AI assistants cite.

03

Improve the right pages

Find the landing pages and products that repeatedly appear in evidenced AI journeys.

04

Track the trend early

Watch the observable share of AI-assisted orders over time instead of waiting for the channel to become obvious.

05

Know when not to act

A small sample or missing evidence is reported plainly, so weak data does not become a confident budget recommendation.

THE EVIDENCE STANDARD

Honest attribution has three labels

Observed

A recognised AI-domain referrer or campaign UTM is captured and persisted to the Shopify order.

Declared

The customer explicitly names an AI assistant in the post-purchase question. Self-reported evidence stays separate.

Unknown

No referrer, UTM or customer declaration exists. The order remains Unknown.

WHAT WE WILL NOT DO

  • Re-label Direct traffic as AI traffic
  • Use fingerprinting to invent a source
  • Assign a named AI assistant from probability alone
  • Hide technical vs self-reported evidence
  • Claim every AI-assisted journey can be recovered
“Unknown is a valid result. We would rather leave a gap than fill it with a guess.”

HOW THE PILOT WORKS

A capped, hands-on pilot

We are onboarding up to 5 established Shopify merchants. The pilot is intentionally small so the work can be done carefully before we decide what should be automated.

  1. 1

    Apply

    Tell us about your store, acquisition mix and current attribution workflow.

  2. 2

    20-minute fit call

    We check whether the problem is real for your team, whether the current pilot can support your store and whether you are willing to share the read-only order and revenue data needed for the report.

  3. 3

    Agree the scope

    We confirm data access, the post-purchase question, report format, timing and pilot terms before anything is installed or paid.

  4. 4

    Receive a weekly report

    The first version is prepared manually using the agreed evidence sources and delivered each week.

  5. 5

    Review what changed

    We compare the evidence with your current reporting and decide whether it is useful enough to continue.

WHO THE PILOT IS FOR

Useful only when the fit is real

A GOOD FIT

  • A live Shopify store trading for at least 12 months
  • Roughly £500k-£15m in annual online revenue
  • Several meaningful acquisition channels
  • A founder or ecommerce, growth or marketing lead close to reporting and budget decisions
  • Willingness to use a short post-purchase discovery question, review order-level evidence and share the read-only order and revenue data needed for the report

PROBABLY NOT YET

  • Pre-revenue or very early stores
  • Single-channel businesses
  • Enterprise processes that make a small pilot impractical
  • Teams looking for a tool that assigns a source to every order

PILOT APPLICATIONS OPEN

Apply for the pilot

Up to 5 pilot places are available. Applications are reviewed individually. If there is a fit, we will arrange a short call and confirm the scope before moving forward.

Places availableApplying does not reserve a paid place, install anything or commit you to the pilot.
Which acquisition channels materially contribute to sales?
In the last 90 days, have two systems disagreed about source or revenue, or has a meaningful order had no useful source?
Do you personally own or materially influence attribution and reporting decisions?

Your application is sent securely to ophir@zashonebuilds.com. No third-party form service receives your information.

FAQ

Questions, answered plainly

Is this a finished Shopify app?+

No. This is an early, hands-on pilot. The weekly report is prepared manually while we test which evidence and decisions are useful enough to automate.

Can you identify every order that involved an AI assistant?+

No. If a journey arrives without a usable referrer or UTM and the customer does not declare the source, it remains Unknown.

What counts as AI-referral evidence?+

A recognised AI-domain referrer, an explicit campaign UTM or a customer declaration collected after purchase. The report keeps these evidence types separate.

Do you use fingerprinting or probabilistic attribution?+

No. We do not assign a named source from fingerprinting or probability.

What access will you need?+

The pilot requires read-only access to the order and revenue data needed to connect evidence to purchases and prepare the report. We verify willingness during the fit call, before agreeing any scope, and use the minimum access needed.

Does applying cost anything?+

No. Applying is only a request to join the pilot. If there is a fit, the scope, duration and any pilot price are confirmed before you decide whether to proceed.

What happens if the pilot is full?+

You can join the waitlist. We will contact you if capacity opens or another cohort is planned. We will not promise a start date we cannot support.

Will you publish my store's data?+

No. Pilot reporting is private to the merchant unless separate written permission is given to use an anonymised or named example.

UP TO 5 PILOT PLACES

Find out whether AI discovery is producing revenue you can actually evidence.

Apply for the capped ZashOne Builds pilot. No guessing, no automatic installation and no payment when you apply.

Apply for the pilot →