Citelift proof and progress: inspect the evidence

Before trusting an automated publishing tool, inspect what it produces, where its facts come from and what happens when a check fails. This page brings together Citelift-authored evidence, with dates and limitations. It is not an independent review or a collection of customer testimonials.

Status on 9 October 2026: Citelift has been listed on the Shopify App Store since 5 October 2026 (listing) and had no reviews yet. We have no clients and no merchant outcome case studies to publish. You can inspect our demonstrations, samples and research; these do not establish increased merchant traffic or sales.

Start with the question you need answered

Question	Inspect this evidence	What it cannot establish
What does the publishing workflow look like?	Editorial-control walkthrough, with dated screenshots from an owned QA store	A complete new end-to-end merchant run or customer results
Can a polished score hide a wrong product fact?	Product fact audit, exact synthetic inputs and a downloadable field-by-field worksheet	Overall factual accuracy of every generated article
What does an AI visibility observation mean?	Repeated-question study, with a reproducible aggregate bundle	Market share, universal AI rankings or a Citelift growth effect
Which SEO job does Citelift cover?	Shopify SEO app decision guide and current plan details	Superiority over every app or independent endorsement
Is our own website being discovered?	The dated baseline below	Shopify merchant results or proof that our software caused these observations

Inspect the product before trusting a claim

The walkthrough explains catalog selection, the brief, draft checks, preview, approval, scheduling and correction. Its screenshots are from an owned test store; its invented product examples are identified. A saved check result belongs to the draft revision that produced it, not automatically to later manual edits.

The article brief template shows what facts, intended reader and acceptance checks to prepare before drafting.

Owned QA store, captured 10 September 2026. Demonstration data, not a client store or a traffic result.

A complete same-task demonstration would retain one catalog snapshot, its actual generated draft, an observed failure, the correction, final checks and verified Shopify publication. That complete evidence chain remains unfinished. Existing screenshots and separate tests must not be stitched into a claim that one uninterrupted run succeeded.

Our fact-audit example gives you the inputs needed to repeat a narrower check yourself. The local writing scorer assesses structure and coverage; it is not a catalog-fact verifier. Use the source facts beside the draft instead of treating a high writing score as a factual certificate.

One real QA hold: the draft stayed unpublished

On 19 September 2026, we inspected one existing article on our owner-controlled fictional beauty QA store: “Soft Wash Shampoo vs Citrus Silk Body Wash vs Pink Clay Wash-Off Mask vs Soft Focus Eye Cream: which one you actually need.” This is a read-only observation of one persisted draft, not a new generation experiment or a customer result.

Download the captured draft, current catalog fields and recorded checks. The bundle includes the capture date and a SHA-256 fingerprint of the retained draft HTML. Catalog fields are the current capture, not a recovered snapshot of what the writer saw originally. Recorded gates are historical results; we did not rerun them against this capture.

Stage inspected	Recorded evidence	What this establishes
Catalog	Four named QA products, current descriptions and photo-permission settings	The product context available at capture time; not historical writer input
Draft	Retained article HTML and its fingerprint	A specific inspectable saved draft rather than a stitched set of screenshots
Recorded checks	A failed media check requires a verified brand-aligned generated image	The saved reason for the hold; other passing checks are not universal quality proof
Publication readiness	Article status held; readyAt is absent	The article has not been marked ready for the publication scheduler
Correction and final publication	Not completed in this run	We cannot claim this article passed a final review or reached the signed-out storefront

The next necessary correction is verified media, not bypassing the hold. The saved failure says no permitted hero photo was available. A new run needs a permitted source photo, a generated image that passes fidelity verification, the required relevant video, and fresh checks tied to the corrected revision. Until those steps and Shopify delivery are observed, this remains a failure-state walkthrough. It does not complete the catalog-to-publication evidence chain described above.

Our own discovery baseline: 18 September 2026

These are observations for citelift.app, our SaaS website. Research, editorial work and technical operations include human and agent work outside the merchant product. This is not a demonstration that Citelift automatically grew its own site.

Source and reporting window	Observation	Interpretation and coverage
Google Search Console, Web, 9–15 September 2026, Pacific reporting dates	142 impressions; 0 clicks; CTR 0%; average position 60.8	Previous complete 2–8 September window had 0 impressions. Percentage growth and a meaningful prior position comparison are undefined
Google indexing report, last updated 14 September	54 indexed of 83 known pages	A lagged report population, not the current sitemap denominator. Two priority URLs inspected on 18 September were indexed; this does not certify every URL
Bing, 10–16 September source dates	0 search clicks, 0 search impressions, 0 sampled AI citations	Previous comparison window unavailable. Bing AI reporting covers its defined Microsoft surfaces, not all assistants
Consumer ChatGPT panel, captured 18 September	0 Citelift name matches and 0 direct owned-domain link answers in 5 completed answers	Five fixed unbranded questions, one fresh temporary conversation each; High setting, exact model identity not exposed
Consumer Claude panel, captured 18 September	0 Citelift name matches and 0 direct owned-domain link answers in 5 completed answers	Same five questions, fresh incognito conversations; displayed Opus 5, High setting

The two consumer panels are small, not weighted by actual search volume, and are not causal experiments. Each platform was unchanged from the five matching questions collected on 17 September. A question that explicitly names Citelift belongs to a separate brand-comparison exercise and is excluded from these discovery counts.

Google's returned query rows account for 108 of the 142 impressions; privacy and reporting limits leave 34 unassigned. Query rows and property totals are different reporting populations. We do not infer a Google AI Overview versus AI Mode split: that breakdown was unavailable in the property. We also do not substitute anonymous website loads for customers or attributed orders for subscription revenue.

The Google AI reporting guide explains the available surfaces and why their measurements should remain separate.

This is a dated snapshot, not a live dashboard. We retain source exports privately and publish aggregate observations. Later evidence should carry its own capture and reporting dates rather than silently replacing this baseline.

What a future merchant case study must contain

Before calling an example a customer result, we need an actual consenting merchant, an eligible approved installation, a frozen baseline, an intervention log and a complete follow-up window. Search clicks, AI observations, attributed orders and operator time remain separate measures. Seasonal changes, promotions and other SEO work must be disclosed.

Download the blank three-pilot evidence worksheet. Its slots are not enrolled merchants. Keep completed records and consent privately; never upload customer data to this public template.

Participation, commercial terms and permission to publish a case study are separate decisions. A positive review is never a condition. We will publish failures and insufficient evidence where those are the actual results; we will not turn three prospective pilot places into three customers.

Evaluate us before merchant results exist

Use the manual AI-answer worksheet or blog SEO audit to inspect your own inputs without connecting a Shopify store. Those deterministic tools do not certify search eligibility or guarantee citations.

Use the Shopify content audit worksheet to inspect existing articles before adding another tool or page.

For a demonstration or to express interest in a future pilot, email support@citelift.app. This opens your email app; it does not enroll you, promise an installation date or start billing. We will confirm availability and scope before any pilot begins.

Shopify documents public-app distribution requirements and development-store limitations. Neither a development-store test nor this page is an App Store approval badge.