Services Case Studies DT 1.5 Get Started About Try DT 1.5 free Log in
Methodology

From finding to verified change.

DT 1.5 audits a public storefront, keeps the supporting evidence, drafts only supported fixes, waits for merchant approval, and verifies eligible Shopify writes against the public result.

Evidence first

A finding is not a finished task.

DT separates what it observed, what it recommends, what a merchant approved, what a provider accepted, and what the public storefront shows afterward.

01

Public-site evidence

The audit reads reachable HTML, titles, descriptions, headings, canonical URLs, social metadata, structured data, robots.txt, crawler rules, and sitemap signals. It records the URL and scan time. It is not Googlebot and cannot see a search engine's private systems.

02

Source labels

Storefront observations, connected-provider data, merchant-confirmed facts, inferred context, and unknown values stay separate. An inference cannot silently replace a merchant-confirmed fact.

03

Supported fixes only

A recommendation becomes an executable draft only when its target, current value, proposed value, destination, and supported write surface are known. Everything else remains guidance or a review-ready patch.

04

No outcome shortcut

Passing a technical check means that specific implementation is present. It does not prove rankings, traffic, AI citations, sales, or future crawler behavior.

Control loop

Six states. No skipped steps.

STEP 01

Audit

Read the current public store and save the evidence.

STEP 02

Draft

Prepare a bounded fix from known store facts.

STEP 03

Approve

Lock the exact before and proposed after state.

STEP 04

Bind

Match the audited hostname to one authorized store.

STEP 05

Execute

Recheck authority and current state before the write.

STEP 06

Verify

Read the affected public surface and record the result.

Verification is specific. “Verified” means the expected public value was found on the checked surface after an eligible write. A blocked read, delayed theme render, provider error, conflicting value, or unsupported surface is labeled separately rather than reported as success.
Boundaries

What the method does not prove.

001

Search and AI visibility

Technical access, entity clarity, and structured content can be checked. Search engines and AI providers still decide what they crawl, index, rank, quote, or cite. Connected AI-answer scans are directional snapshots, not guaranteed consumer-app results.

no ranking guaranteeno citation guarantee
002

Shopify execution

The current automated executor is limited to eligible homepage metadata and complete sitewide Organization or WebSite data on the exact connected store. Catalog-scale and unsupported changes remain review-ready patches.

exact hostnamesupported surfaces only
003

Advertising

Planning, account preflight, policy checks, and paused creation depend on the connected provider, permissions, tracking, business facts, and safety limits. Sending a request is not treated as a completed provider action; read-back is required where supported.

provider dependentapproval gated
Publication standard

No study without the study.

MGO Data will publish research only after the collection date, sample, method, definitions, limitations, findings, and update history are ready. Planned topics are not results, and client data is not public evidence without permission.

Current status: no merchant-outcome or reliability study is being presented as complete. The public Safety Gauntlet reports only named synthetic rule fixtures, their current pass or fail state, and their limitations.
See the workflow

Start with the evidence.

Run one live, read-only scan before creating an account or connecting Shopify.

Audit my Shopify store