Repair broken shopping links, align product and policy facts, and establish a reliable search and AI-visibility baseline.
Live storefront evidence collected November 6. Ahrefs comparison: October 23 → November 6. Earlier storefront comparison: October 1 findings in the October 2 revised report. English-language, primarily US retail and wholesale context.
Fix the shopping paths and conflicting facts first
LegacySips has a useful foundation: real farmer stories, product-specific flavor and brewing information, crawlable pages and consistent canonicals. The immediate gaps are concrete: three broken tea submenu links, one broken featured story, inconsistent shipping thresholds, an unfinished footer returns page, and missing product structured data.
Highest-value next step: repair the three submenu destinations and consolidate shipping and return information. These are observable buyer and crawler problems. Conversion or ranking improvement must be measured after the fixes.
43/43
sitemap HTML pages returned 200
4
confirmed broken link destinations
0/20
product pages with extracted JSON-LD
0
Ahrefs AI citations in both snapshots
Three wins available now: repair shared navigation; replace the stale policy page; fix the featured Resources story. Three larger gaps: Product/Offer markup, trustworthy authority signals, and current first-party measurement.
No composite score is issued: a comparable prior SEO/GEO scoring rubric was not located, and current first-party and consumer AI-panel evidence is incomplete. Counts above are measurements with defined denominators, not grades.
What changed since the earlier review
Topic
October 1 / revised October 2
November 6
Shipping
Blank shipping section; $50 claim conflicted with one tested basket.
Partial: formal policy now explains $18 eligibility and green-coffee exclusions. Storefront still advertises $50 and bundle copy $25.
Returns
Address placeholder and generic instructions.
Unchanged on footer route. Separate formal policy is more specific, so the two now conflict.
Quantity clarity
Weights farther down descriptions.
Partial: tea selectors now include weights/servings; roasted coffee still uses Discovery without weight in selector.
Review counts / card excerpts
Repeated 127 counts and concatenated Description fragments.
Both remain visible. No review-source reconciliation obtained.
Not retested. Prior confirmation is preserved; current results are unavailable, not zero.
Checkout behavior
One basket and destination worked; no payment placed.
Not retested. No current claim about shipping configuration, subscriptions or payment completion.
Newly observed findings are not automatically regressions: the earlier report did not establish their previous state. No intervening full SEO/GEO report was located in the sources searched.
Measured visibility remains limited
Ahrefs measure
October 23
November 6
Change / interpretation
Organic keywords in top 100
0
0
No tracked keywords in this vendor dataset
Estimated monthly organic visits
0
0
Modeled estimate, not GA4 traffic
Live backlinks
943
926
−17 (−1.8%)
Live referring domains
841
825
−16 (−1.9%)
Domain Rating
1.3
1.1
−0.2 points on Ahrefs’ logarithmic scale
AI citation links / cited pages
0 / 0
0 / 0
All seven requested platforms and aggregate
Organic traffic value
Unavailable
Unavailable
API returned null; not converted to $0
Source: Ahrefs Site Explorer. Domain scope includes subdomains; organic totals are all-country. These are snapshots, not totals accumulated over the last two weeks. Public search also surfaced LegacySips pages; a zero keyword estimate does not establish that Google has indexed no pages.
0 in the one extracted JSON-LD block; not a schema eligibility test
Findings and recommended repairs
Status compares with the located earlier report where possible. Owners and dates are proposed, not accepted assignments. P1 = address first; P2 = next cycle; P3 = monitor or conditional work.
Observed: Essentials Collection, Rare Teas and Full Set point to /collections/ handles that returned 404 in Chrome. The corresponding /products/ bundle pages return 200. These links occur in the shared navigation.
Earlier evidence: Not assessed in the located October report.
Action: Point the submenu to the intended existing bundle product pages, or create substantive collections if that is the intended shopping route.
Definition of done: All three navigation choices reach the correct live offer on desktop and mobile; a crawl finds zero 404 destinations for these links.
Limits: No ranking or conversion effect measured.
Proposed owner: Store developer · Effort: Small · Target: November 9, 2026
LS02 Shipping messages disagree across shopping surfaces P1 · partial
Observed: Home and collection cards say free shipping over $50, including green coffee cards. All three tea-bag bundles say over $25. Individual tea bags say over $18. The formal policy says $18 or more after discounts for eligible US tea, roasted coffee and pods, with green coffee excluded and separate mixed-basket charges.
Earlier evidence: October 1 storefront review found a blank shipping section and a $50 claim inconsistent with one tested $20 basket.
Action: Approve the operating rules, then use one shared summary linked to the formal policy across cards, bundles, product pages and the footer. Clarify the exact-$18 boundary and green-coffee exception.
Definition of done: Every displayed threshold and exclusion agrees. An owner-authorized rate matrix verifies tea, roasted coffee, green coffee and mixed baskets at $17.99/$18/$18.01 and representative destinations.
Limits: The formal policy is now substantive; configured shipping rates and checkout behavior were not retested. Policy text is not proof of backend rules.
Proposed owner: Mbugua + store developer · Effort: Small · Target: November 9, 2026
LS03 Footer returns page still has a placeholder and stale instructions P1 · unchanged
Observed: /pages/shipping-and-returns still contains [INSERT RETURN ADDRESS], contact@ as the return route, and a promise to send a label. /policies/refund-policy instead names sales@, gives an address, makes eligible change-of-mind shipping the customer’s responsibility, and explicitly excludes tea/coffee change-of-mind returns.
Earlier evidence: October 1 review documented the same return-address placeholder and generic return text.
Action: Replace the duplicated footer page with an approved concise summary and links to the current formal policies, or synchronize its content. Have the owner approve the commercial terms.
Definition of done: No placeholder remains. Footer and formal policy give the same contact route, eligibility, cost responsibility and instructions.
Limits: This is a content-consistency finding, not a legal assessment of the return terms.
Proposed owner: Mbugua + store developer · Effort: Small · Target: November 9, 2026
LS04 Repeated review counts still need product-level substantiation P1 · unchanged
Observed: Three homepage featured cards display 127 reviews each; the collection cards repeat the same count. The sampled purple and bundle pages did not expose corresponding review evidence in the collected content.
Earlier evidence: October 1 review identified the same three featured products with five stars and 127 reviews.
Action: Trace the displayed rating/count to a real review source and the correct product. Remove the count or rating where it cannot be supported.
Definition of done: Each displayed review figure reconciles to accessible product reviews; no static placeholder count or unsupported aggregate rating is emitted.
Limits: The audit does not establish that reviews were fabricated; external review records were not available.
Proposed owner: Mbugua + store developer · Effort: Small · Target: November 9, 2026
LS05 Product structured data is missing from the inspected markup P1 · newly observed
Observed: No JSON-LD was extracted from any of the 20 sitemap product pages. The rendered purple-tea and Rare Teas bundle spot checks also had no JSON-LD or itemtype microdata. Only the Bridgerton article emitted JSON-LD among 43 sitemap HTML pages.
Earlier evidence: Not assessed in the located October report.
Action: Add accurate Product/Offer or appropriate variant markup through the product template. Include identifiers, price/currency, availability, URL and images that match the selected offer. Add organization identity centrally.
Definition of done: Representative tea, coffee, bundle and variant pages pass applicable Rich Results Test checks and match visible facts. Validate sold-out handling and avoid unsupported review markup.
Limits: Coverage combines raw HTML and browser DOM. Full RDFa parsing and Google rendering were not performed. Valid markup supports eligibility, not guaranteed rich results or AI citations.
Proposed owner: Store developer · Effort: Medium · Target: November 13, 2026
Observed: Missing descriptions: About, the standalone Resources page, Sip with LegacySips, Frontpage collection, All collection, News blog and Resources blog. The standalone Resources and Resources blog also share the title Resources – LegacySips.
Earlier evidence: Not assessed in the located October report.
Action: Write distinct descriptions for useful destination pages; decide whether empty or duplicate routes should be consolidated before writing metadata. Improve the generic homepage title to describe Kenyan tea and coffee.
Definition of done: Every retained indexable landing page has an accurate, distinctive title and description; duplicate resource intent is deliberately resolved.
Limits: A missing description is not an indexing block; Google may generate or rewrite snippets.
Proposed owner: Store developer · Effort: Small · Target: November 13, 2026
LS07 A sitemap landing page has no substantive main content P2 · newly observed
Observed: /pages/sip-with-legacysips returned 200 but had no H1 and no meaningful main-page copy in the collected HTML. Site navigation, footer and predictive search dominate its text.
Earlier evidence: Not assessed in the located October report.
Action: Confirm whether the page is an active campaign destination. Publish the intended content, or retire it with an appropriate destination after checking links and traffic.
Definition of done: The route has a clear purpose and substantive content, or a deliberate redirect/removal supported by link and analytics checks.
Limits: No ranking or conversion effect measured.
Proposed owner: Content lead + store developer · Effort: Small · Target: November 13, 2026
LS08 Featured Resources story is broken and has an unsupported relative date P1 · newly observed
Observed: The Resources page advertises a 70-year journey article with “1 day ago”; its Read Now destination returns 404. Five actual articles are present in the sitemap.
Earlier evidence: Not assessed in the located October report.
Action: Replace the featured card with an existing relevant article and its real publication date, or publish the intended story. Generate dates from article records.
Definition of done: Featured story returns 200, the card title matches its destination, and the date is factual.
Limits: No ranking or conversion effect measured.
Proposed owner: Store developer · Effort: Small · Target: November 9, 2026
LS09 Product quantities improved, but the coffee selector remains vague P2 · partial
Observed: Loose-leaf tea selectors now display weights, including purple tea 1 oz/2 oz and white tea 0.5 oz/1 oz. Tea-bag selectors show 5 servings. The roasted-coffee selector still says Discovery, while 6 oz appears in the description. Bundle counts are not clearly stated beside the price in the inspected pages.
Earlier evidence: October 1 review found tea and coffee physical quantities farther down the description.
Action: Keep the improved tea labels; add the coffee weight and each bundle’s exact bag count or total servings beside price and purchase controls, using approved catalog facts.
Definition of done: Buyers can identify physical quantity before expanding or reading the description on desktop and mobile.
Limits: No ranking or conversion effect measured.
Proposed owner: Merchandising + store developer · Effort: Small · Target: November 13, 2026
LS10 Product-card excerpts still expose formatting fragments P2 · unchanged
Observed: Homepage and tea collection excerpts concatenate marketing text and headings, including tea.Description and complex.Description.
Earlier evidence: October 1 review specifically flagged tea.Description and complex.Description fragments.
Action: Use a dedicated short card description or strip headings and preserve spaces when deriving excerpts.
Definition of done: All featured and collection cards show readable summaries without heading fragments, duplicated words or clipped formatting artifacts.
Limits: No ranking or conversion effect measured.
Proposed owner: Store developer · Effort: Small · Target: November 13, 2026
LS11 Claims need scoped evidence that people and search systems can verify P2 · newly observed
Observed: The homepage claims 100% sustainable. The process page claims Rainforest Alliance certification and pesticide testing. The purple-tea article discusses lower caffeine and its excerpt promotes health benefits. Visible source links or certificate scope were not found beside these claims.
Earlier evidence: Not assessed in the located October report.
Action: Document which farms/products, certificates and dates each claim covers. Link primary research for composition claims; use lot-specific test evidence for quantitative product claims. Replace broad health or universal claims with approved, supportable wording.
Definition of done: A reader can verify the scope and source of each material claim; unsupported claims are removed or qualified.
Limits: The claims are unverified in this audit, not proven false. No medical conclusions are drawn.
Proposed owner: Mbugua + content lead · Effort: Medium · Target: November 13, 2026
Observed: The Farmers page already names five growers and gives farm, elevation and experience details. Brewing content provides practical temperatures and times. These are useful foundations, but four of five articles lack JSON-LD and the sampled byline has no evident expertise profile in the inspected content.
Earlier evidence: Not assessed in the located October report.
Action: Connect each relevant product to its actual source farm, add reviewed practical answers, identify the author/reviewer and their real expertise, and use accurate Article/Breadcrumb markup. Prioritize existing content before adding more generic articles.
Definition of done: Two priority guides have source links, a responsible reviewer, direct answers and contextual links to appropriate products/farms.
Limits: No ranking or conversion effect measured.
Proposed owner: Content lead + Mbugua · Effort: Medium · Target: November 20, 2026
LS13 Wholesale page can answer buying questions more precisely P2 · opportunity
Observed: LegacySips provides a Faire route and an inquiry form, but the inspected main copy does not clearly state minimum quantities, sample process, pack formats or a response-time commitment. JusTea states a 1–3 business-day response and packaged/bulk options; Fettle lists 250g–5kg formats and export steps.
Earlier evidence: Not assessed in the located October report.
Action: Add owner-approved formats, minimums, sample availability, lead times, destination coverage and inquiry-response expectations. Qualify Faire benefits and eligibility; do not copy another supplier’s terms.
Definition of done: A wholesale buyer can evaluate fit and next steps without guessing. Measure qualified inquiries separately from retail sales.
Limits: These are two relevant content benchmarks, not proven ranking winners or an exhaustive competitor study.
Proposed owner: Mbugua + content lead · Effort: Medium · Target: November 20, 2026
LS14 Backlink totals overstate the useful authority signal P2 · newly observed
Observed: Ahrefs reports 825 live referring domains, down from 841 on October 23, and DR 1.1 versus 1.3. Fourteen of the 15 highest-DR domains in the current sample carry Ahrefs’ is_spam flag. The sample is biased toward highest DR and is not a whole-profile spam percentage.
Earlier evidence: Not assessed in the located October report.
Action: Review actual linking pages for useful trade/event/retailer references. Pursue genuine supplier, stockist, event and editorial coverage tied to specific products and farm evidence. Investigate suspicious links before any remediation.
Definition of done: Document the quality and relevance of priority links and earn a small number of verifiable relevant references. Track qualified referral visits and indexing, not link counts alone.
Limits: Vendor flags do not prove a Google penalty or that the owner purchased links. No automatic disavow recommendation; no causal explanation for DR changes established.
Proposed owner: SEO lead + Mbugua · Effort: Medium · Target: November 20, 2026
LS15 Current first-party search and revenue measurements remain unrefreshed P1 · not retested
Observed: The prior October 2 report confirmed that GA4 receives ecommerce events. Current GA4, Search Console, Merchant Center and Shopify conversion data were not accessed. LegacySips was absent from the three projects returned by the connected Ahrefs management endpoint.
Earlier evidence: October 2 authenticated review confirmed ecommerce event receipt but left complete journey and Shopify reconciliation unresolved.
Action: Use the existing GA4 integration, confirm Search Console property access, and reconcile order IDs and purchase events. Compare October 23–November 5 with October 9–22 using the same timezone and definitions. Separate brand/non-brand and AI referrals.
Definition of done: Save comparable 14-day clicks, impressions, landing pages, organic/AI sessions and matched orders, with processing lag and testing exclusions stated.
Limits: Unavailable current data is not zero. Do not add a duplicate pixel or relabel the existing integration absent.
Proposed owner: Analytics owner + SEO lead · Effort: Medium · Target: November 13, 2026
LS16 AI citation counts remain zero in the measured vendor corpus P2 · unchanged in vendor snapshots
Observed: Ahrefs returned zero citations and zero distinct cited pages across ChatGPT, Perplexity, Google AI Overviews, AI Mode, Gemini, Copilot and Grok on both October 23 and November 6. This is domain citation coverage, not unlinked brand-mention coverage.
Earlier evidence: October 23 historical Ahrefs snapshot: zero citations on every requested platform.
Action: After repairing factual and navigation defects, establish the fixed five-question panel below. Record mention, recommendation, linked citation and source separately across repeated runs. Track referral traffic independently.
Definition of done: A dated panel records exact prompts, platform, geography/language, answer sources and repeat count. Retest on November 20 and after sufficient indexing time.
Limits: No consumer ChatGPT/Perplexity/AI Mode prompt panel was executed. Zero in Ahrefs does not establish zero citations everywhere, zero mentions or zero customer traffic.
Proposed owner: SEO lead · Effort: Medium · Target: November 20, 2026
LS17 Discovery foundations pass, but indexing and field performance need verification P2 · partially verified
Observed: All 43 sitemap HTML pages returned 200 using direct fetches plus Chrome recovery. All have self-referencing canonicals; none exposed a noindex meta directive. robots.txt allows public content. Homepage at 390px and sampled desktop pages fit the viewport.
Earlier evidence: Not assessed in the located October report.
Action: Use Search Console to check representative URL indexing, chosen canonicals, sitemap processing and crawl problems. Obtain field Core Web Vitals before making speed claims. Test key mobile shopping interactions separately.
Definition of done: Representative pages have documented indexing states; field LCP/INP/CLS and date range are recorded, or insufficient data is stated.
Limits: Twenty direct requests returned local_rate_limited responses; those pages loaded 200 in Chrome. This is not evidence that Googlebot is blocked. Response headers, Googlebot logs and field performance were not fully audited.
Proposed owner: SEO lead + store developer · Effort: Medium · Target: November 20, 2026
LS18 AI discovery files exist, but do not establish visibility or transaction success P3 · verified foundation
Observed: /agents.md and /llms.txt return public Shopify agent instructions, and the agent-discovery sitemap is present. Their existence coexists with zero observed Ahrefs citations.
Earlier evidence: Not assessed in the located October report.
Action: Keep discovery information accurate. Treat supported commerce integrations as a separate workstream; prioritize reliable product, availability and policy facts before agent checkout experiments.
Definition of done: Future integration tests use advertised capabilities and supported buyer-review flows. Citation and conversion results are measured separately.
Limits: No live commerce tool calls, cart changes or payments were performed. llms.txt is not a Google AI inclusion requirement or a citation guarantee.
Proposed owner: Store developer + SEO lead · Effort: Small · Target: November 20, 2026
LS01–04, LS08: navigation, shipping, returns, review proof and featured story
Four broken destinations repaired; policy promises synchronized; review claims substantiated or removed.
November 13
Store developer + merchandising
LS05–10: product markup, metadata, page purpose, quantities and excerpts
Validate representative product templates, then recrawl all sitemap pages. Preserve correct canonicals.
November 13
Analytics owner + SEO lead
LS15: refresh first-party measurement
Comparable completed 14-day periods with consistent timezone; matched purchases and test-traffic exclusions.
November 20
Mbugua + content / SEO lead
LS11–14, LS16–18: evidence, wholesale answers, quality links and AI baseline
Two improved priority guides; documented source claims; five-prompt repeated panel and dated follow-up report.
Feedback loop: the implementation owner records the live URL and change date for each completed item. Kaizen retests the acceptance criteria and retains the finding ID. Review outcomes over 28–56 days where traffic permits; a two-week implementation check is too short to promise ranking or citation gains.
Dependency: Mbugua confirms commercial facts; the developer implements approved content and templates; analytics access enables outcome measurement. This audit changed no storefront settings, submitted no forms and created no recurring automation.
Content that helps buyers decide
The fastest content work is to strengthen existing pages around specific questions. Search volumes and competitor ranking gaps were not measured, so the topics below are intent hypotheses rather than volume-ranked keywords.
Intent / question
Existing destination
Recommended improvement
Buy Kenyan purple tea in the US
Purple loose leaf product + purple-tea article
Connect flavor, exact package size, brew instructions, caffeine caveat, origin and current delivery terms.
Whole-leaf tea bags versus loose leaf
Tea collection + relevant product pages
Compare formats, serving counts, preparation and price per serving once quantities are approved.
Kenyan AA green coffee for home roasting
Kiganjo green coffee product
Explain unroasted status, processing, origin, approved lot details, package sizes and shipping exclusions.
Wholesale Kenyan tea for cafes
Wholesale page
State approved minimums, formats, sample process, lead time and next-step response expectations.
How to brew Kenyan purple tea
Existing brewing guide
Add product-tested recipe details and a named reviewer; link to purple tea and explain adjustments.
Draft metadata for owner review
Homepage title: Kenyan Tea & Coffee from Family Farms | LegacySips
About description: Meet the family behind LegacySips and explore the Kenyan farms, growers and heritage behind our specialty tea and coffee.
Resources blog description: Explore Kenyan tea and coffee with LegacySips brewing guides, farmer stories and practical advice for choosing your next cup.
Draft FAQ answers based on current public facts
Does green coffee arrive roasted? No. LegacySips’ Kiganjo AA and Ironman Estate green coffees are unroasted beans intended for roasting before brewing.
Does free shipping apply to green coffee? The current formal shipping policy excludes green coffee from the domestic free-shipping offer. Green-coffee charges can still apply in mixed baskets. Link to the approved policy and verify configured rates before reusing this answer.
Can a cafe buy wholesale? LegacySips offers a wholesale inquiry form and a Faire ordering route. Use the inquiry form to discuss product formats and your business needs. Do not promise minimums, samples or lead times until the owner confirms them.
These are proposed drafts only. No site copy was changed.
AI visibility: separate access from results
Search-readable content, honest product facts and useful source material are the foundation. Google says its existing SEO practices apply to AI features; special AI markup or llms.txt is not required. Product markup can support shopping presentation, but neither schema nor crawler access guarantees a citation. Google AI-feature guidance.
Five-question panel for the next baseline
Where can I buy Kenyan purple tea in the United States?
Which companies supply Kenyan tea wholesale to US cafes?
Where can I buy Kenyan AA green coffee beans for home roasting?
What is the difference between Kenyan purple tea and green tea?
Which Kenyan tea gift sets ship in the United States?
Proposed method: US English, ChatGPT search, Perplexity and Google AI Mode; three independent runs per question per platform. Record exact wording, timestamp, product/model, location context, brand mention, recommendation, linked citation, cited URL and answer facts. Keep a separate branded diagnostic: “What is LegacySips and where does its tea come from?” No such consumer-answer panel was executed in this audit. Ahrefs domain-citation counts are the measured GEO result.
A bounded test
After the technical repairs, improve the purple-tea article and relevant product page with approved source facts, direct answers and farm links. Keep an unchanged relevant guide as an observational comparison, if suitable. Record indexing and baseline impressions first. Review repeated citations, non-brand clicks, qualified visits and matched purchases after 28–56 days. Low traffic, seasonality, platform changes and simultaneous store repairs limit causal inference. Revert any inaccurate or misleading copy immediately; do not declare failure from one missing citation.
Current platform changes worth separating from SEO
Shopify’s October 28 UCP /buy update supports a common cart/checkout permalink route. Its documentation says selling plans and certain custom fields remain unsupported there, so it is not an automatic replacement for subscription routes. This is a conditional commerce-integration opportunity, not evidence of higher rankings or AI visibility. Official October 28 announcement.
Google’s October 23 URL-structure documentation update clarifies word separation. LegacySips already uses hyphenated product handles; this is not a reason to rename established URLs. Google documentation update log.
Sources, coverage and limits
Completed: sitemap discovery; 43 HTML pages and agent text; metadata, H1, canonical and JSON-LD extraction; selected browser rendering; four suspect destination checks; robots and agent-file review; two historical Ahrefs snapshots; a 15-domain authority sample; two wholesale content benchmarks; earlier audit comparison.
Collection detail: initial direct requests returned local_rate_limited for 20 URLs. A slow retry did not recover them; Chrome loaded every affected sitemap page successfully. The final inventory uses those browser DOM captures, and retains the initial limitations locally. This does not diagnose bot blocking at the store. The 44 sitemap entries are distinct from the four additional broken-link targets and policy/host probes.
Unavailable / not executed: current GA4, Search Console and Merchant Center data; Shopify administration; field Core Web Vitals; full response-header audit; a complete accessibility audit; all external-link checks; transactional/subscription tests; Google Business Profile verification; consumer AI prompt panel; keyword-volume/content-gap study; exhaustive backlink review. No claim of full Google indexing or improved sales is made.
Source coverage: live public site observations and Ahrefs data retrieved November 6; October 1 storefront baseline in an October 2 revision; current Google documentation; Shopify October 28 announcement; JusTea and Fettle public wholesale pages. Platform research is a targeted check, not an exhaustive review of every AI provider, X discussion or industry study. No broad causal GEO claims are inferred from promotional material.
Publication: this report contains public-site findings and summarized vendor metrics. Raw API responses, browser captures and internal working files are retained locally and are not served by the public report. The report requests no indexing; this is not password protection.
SEO: visibility in ordinary search. GEO: visibility in AI-generated answers. JSON-LD: machine-readable facts embedded in a page. Canonical: the preferred URL declared by a page. H1: the main heading. DR: Ahrefs’ relative backlink-strength metric. GA4: Google Analytics 4. UCP: Universal Commerce Protocol. LCP, INP and CLS measure loading, responsiveness and layout movement; no field values were obtained here.