Evidence: A large Italian Affairs promotional modal and a separate cookie panel appear simultaneously on initial entry,
covering most of the resort hero.
Guest or team impact: The first-time visitor must clear two layers before evaluating the resort or booking action. No
conversion effect has been measured.
Built into the new site: Use one coordinated, accessible consent layer; place event promotion in a contextual module
or delay it until relevant engagement.
Acceptance test: In clean-session desktop and mobile QA, consent and promotion never stack over one another;
primary navigation remains discoverable; dialogs have labelled close controls, focus handling and keyboard dismissal
where appropriate.
https://www.andaraphuket.com/

Evidence: The Chandara introduction merges a resort-founder sentence into the estate description, then resumes with
a detached reference to Andara. The malformed wording is visible in the main introductory paragraph.
Guest or team impact: A prominent luxury-brand statement reads unfinished and makes the relationship between the
resort and estate harder to understand.
Built into the new site: Rewrite and proofread a concise introduction, then give Chandara a separate paragraph or card.
Acceptance test: Approved homepage copy passes editorial review in every launch language; Chandara facts and
founder story remain separate, grammatical and consistent with supplied source content.
https://www.andaraphuket.com/

Evidence: The visible homepage hero is photographic/video-led with global navigation and Book Now; no hero
headline or on-page date/guest selector was observed.
Guest or team impact: Visitors receive atmosphere before practical reassurance about location, accommodation choice
and the next booking step. This is a strategic opportunity, not proof of low performance.
Built into the new site: Keep cinematic imagery, add a concise positioning line and an unmistakable booking action. A
date-and-guest bar is an optional enhancement subject to SynXis validation and separate scope approval.
Acceptance test: In approved desktop and mobile designs, a first-time guest can identify Andara, its location/stay
proposition and the primary booking action without dismissing a promotion or scrolling. Any date/guest selector
successfully hands supported values to SynXis.
https://www.andaraphuket.com/

Detailed findings 2

Evidence: The introductory area combines resort positioning, sizes, bedroom ranges, estate information, dining,
wellness and founder history into one long paragraph beside a map.
Guest or team impact: Important selection facts are difficult to scan; unrelated messages compete within one reading
task.
Built into the new site: Separate a short positioning statement, three or four stay facts, accommodation routes, an
estate feature and supporting brand story.
Acceptance test: The approved homepage content model has discrete positioning, stay facts, accommodation choices
and estate/story sections; key facts can be found without reading the entire introduction.
https://www.andaraphuket.com/

Evidence: The villa body CTA uses gc.synxis.com/rez.aspx with shell/template values and a historic _ga query, while the
header and suites use be.synxis.com/?Hotel=27746.
Guest or team impact: Equivalent actions are implemented through different handoffs, increasing maintenance and QA
burden. This alone does not establish a broken booking engine.
Built into the new site: Standardize all booking actions on one confirmed current SynXis integration and remove
obsolete tracking values.
Acceptance test: All room-booking actions use the signed-off SynXis link/integration contract. No legacy route or
copied analytics identifier remains unless explicitly approved and tested.
https://www.andaraphuket.com/accommodation/villas/

Evidence: Terrace, Pool and Penthouse Suite buttons all point to the same hotel-only SynXis URL with Hotel=27746 and
no visible category parameter.
Guest or team impact: The guest has already selected a category but the outgoing URL does not communicate it.
Whether the engine recovers that intent by other means was not tested.
Built into the new site: Use supported SynXis room/category deep links if available and separately agreed. Otherwise
make the generic handoff clear so guests know they will select their stay in SynXis.
Acceptance test: For each suite card, compare the selected category against the SynXis result using supported
parameters. If category selection cannot be prefilled, the approved fallback accurately explains that the guest must
select it in SynXis.
https://www.andaraphuket.com/accommodation/suites/

Detailed findings 3

Evidence: The three suite cards contain long descriptions plus Book Now and View Gallery; category-specific floor area,
bedroom range and occupancy are not summarized in each card. Those facts are on a separate comparison page.
Guest or team impact: Guests must move between descriptions and comparison to resolve basic fit.
Built into the new site: Put verified bedroom range, guests, area, pool type and key differentiator directly into each
card; offer a deeper room-detail view.
Acceptance test: Each published suite card displays approved values for its required decision fields and links to
matching imagery/detail content. Check content parity against the accommodation matrix at desktop and mobile
breakpoints.
https://www.andaraphuket.com/accommodation/suites/

Evidence: Villas overview says 6–14 guests; comparison says 14 adults + 4 children. Suites overview says 2–8 guests;
comparison allows 8 adults + 2 children for Pool/Penthouse.
Guest or team impact: It is unclear whether overview maxima exclude children, apply only to some configurations or
have different rules. No capacity claim is declared false.
Built into the new site: Create one approved capacity model by unit/configuration and state adults, children, extra beds
and exceptions consistently.
Acceptance test: Overview, category/detail page, comparison and booking-help content use the same approved
adults/children/extra-bed capacity definitions for every offered configuration.
https://www.andaraphuket.com/accommodation/villas/
https://www.andaraphuket.com/accommodation/suites/
https://www.andaraphuket.com/room-comparison/

Evidence: Villa Amenities lists Hair Dryer, Bath Amenities and later Hair Dryer again.
Guest or team impact: Duplication looks unedited and adds noise to a practical checklist.
Built into the new site: Deduplicate the amenity model and use consistent labels across suites and villas.
Acceptance test: Render the villa amenities from the approved content model; no duplicate entries appear, and
equivalent amenities use consistent labels across accommodation types.
https://www.andaraphuket.com/accommodation/villas/

Detailed findings 4

Evidence: At CSS viewport 1180×757, the comparison container is 1085px wide with 1202px of content. The Signature
Villas column is visibly cut off; a horizontal scrollbar appears at the bottom.
Guest or team impact: A key category can be partly hidden until the visitor notices and uses the lower scrollbar.
Built into the new site: Use a responsive comparison pattern with a clear scroll affordance, sticky row labels, selectable
categories or a better-fitting desktop layout.
Acceptance test: At the agreed desktop/mobile test widths, every category and feature is reachable without accidental
page overflow. If horizontal comparison is intentional, its affordance is visible and category/row context remains
available while scrolling.
https://www.andaraphuket.com/room-comparison/

Evidence: DOM inspection found six separate tables, each with 12 rows and zero th elements. The accessibility tree
exposes generic containers rather than a coherent comparison table.
Guest or team impact: Assistive technology cannot reliably associate category/feature headers with values through
standard table semantics. A full screen-reader audit was not performed.
Built into the new site: Rebuild as an accessible table with caption and scoped headers, or equivalent accessible
comparison cards.
Acceptance test: Keyboard and screen-reader spot checks correctly announce table caption, category and feature
headers with values; automated accessibility checks find no missing comparison-header relationships.
https://www.andaraphuket.com/room-comparison/

Evidence: The Signature Villas category shows no Bathroom or Pool Size values while other category columns supply
those facts.
Guest or team impact: A blank cell does not explain whether the feature varies, is unknown or is unavailable.
Built into the new site: Populate approved values or explicitly label variation and provide a route to the relevant villa
configuration.
Acceptance test: All required comparison fields contain an approved value or an explicit variation/availability
explanation; no unexplained blank Signature Villa cells remain.
https://www.andaraphuket.com/room-comparison/

Detailed findings 5

Evidence: The inspected comparison area contains images and specifications but no category-specific detail or
reservation links. The global Book Now remains available.
Guest or team impact: After evaluating fit, guests must return to navigation or use a general booking action.
Built into the new site: Add View this stay and Check availability actions linked to the selected category, while keeping
booking on SynXis.
Acceptance test: Each comparison choice has a working, clearly labelled detail/booking continuation that refers to that
same accommodation type and follows the agreed SynXis contract.
https://www.andaraphuket.com/room-comparison/

Evidence: All, Terrace Suite and Pool Suite filters are LI elements with no role or tabindex; the active category is
represented by a CSS class. The accessibility tree exposes these as text.
Guest or team impact: Pointer users can filter the gallery, but the controls lack standard keyboard and assistive-
technology semantics.
Built into the new site: Implement buttons or an accessible tabs/filter pattern with focus behavior, selected state and
announced updates.
Acceptance test: Reach every gallery filter with Tab, operate with keyboard, see a visible focus indicator, hear selected
state and confirm the correct images update without losing focus.
https://www.andaraphuket.com/accommodation/room-gallery/?tab=pool_suite

Evidence: Pool Suite gallery correctly respects tab=pool_suite and shows four matching photographs, but accessible
link labels repeat only the category name rather than describing each room/view.
Guest or team impact: Guests using assistive technology receive little distinction between the images; sighted guests
also have limited room-scene context.
Built into the new site: Use concise scene captions and descriptive alternatives; keep the working category deep links
and add a contextual booking route.
Acceptance test: Gallery links/images expose approved scene descriptions rather than repeated category-only labels;
category query links still select the intended filter and all images belong to that category.
https://www.andaraphuket.com/accommodation/room-gallery/?tab=pool_suite

Detailed findings 6

Evidence: The homepage and wedding/wellness accessibility trees expose slide buttons named with internal labels
such as mainslide--home-1, PS4 SD, and photo--wedding-9.
Guest or team impact: Nonvisual navigation is cluttered with asset-production labels that do not explain the slide
content.
Built into the new site: Name controls meaningfully, label slide position, hide decorative filenames and validate
keyboard/pause behavior across all sliders.
Acceptance test: Inspect accessible names across sliders: no internal filenames or production labels are exposed.
Keyboard navigation, selected-slide announcements and motion controls satisfy the agreed accessibility baseline.
https://www.andaraphuket.com/

Evidence: The shared footer email input has an empty associated label, no aria-label and only the placeholder email
address; the accessibility tree shows an unnamed text field.
Guest or team impact: The purpose may not be consistently conveyed to assistive technology, especially once a value
replaces the placeholder.
Built into the new site: Provide a persistent, programmatically associated Email address label and a clear Subscribe
action.
Acceptance test: The newsletter input has a persistent programmatic Email address label; the action clearly identifies
subscription. Validate keyboard focus and announced purpose after text has been entered.
https://www.andaraphuket.com/contact/

Evidence: Property sales is displayed as propertysales@andaraphuket.com but href is
mailto:kenneth@andaraphuket.com. The CEO contact block displays ceo@andaraphuket.com but href is
mailto:gm@andaraphuket.com; a later CEO sign-off does link to ceo.
Guest or team impact: Clicking can address a message to a different mailbox than the guest reasonably expects. This
may be intentional routing but is not communicated.
Built into the new site: Confirm intended recipients with the property, then align displayed addresses, link destinations
and role labels.
Acceptance test: Compare every published contact label with its actual mailto destination against a client-approved
routing list. Clicking each opens the intended address; any role alias or forwarding destination is explained consistently.
https://www.andaraphuket.com/contact/

Detailed findings 7

Evidence: At the tested desktop width, the left contact column is narrow enough to wrap silk@andaraphuket.com and
propertysales@andaraphuket.com across multiple lines within the addresses.
Guest or team impact: Contact information is harder to read and the layout feels less polished at an ordinary desktop
width.
Built into the new site: Use a balanced contact-card grid with sufficient width; apply deliberate overflow/wrapping
rules and test long translated labels.
Acceptance test: At all agreed responsive widths, long email addresses and translated department labels stay readable
and do not overflow, truncate or wrap in an accidental single-character fragment.
https://www.andaraphuket.com/contact/

Evidence: Contact and wedding fields are transparent with fine white borders placed directly over detailed
monochrome photographs. No formal contrast ratio has been measured.
Guest or team impact: The photo competes with field boundaries and text, particularly on denser portions of the
image.
Built into the new site: Use a solid or predictably opaque form surface, stronger field boundaries and verified
text/control contrast.
Acceptance test: Measure text, field-boundary and focus-indicator contrast against the actual final backgrounds.
Repeat for focus, entered text, errors and disabled states; meet the agreed accessibility target.
https://www.andaraphuket.com/contact/

Evidence: DOM inspection found no autocomplete attributes on first name, last name, email, phone, address or
country inputs. Labels are otherwise associated on the main contact form.
Guest or team impact: Browsers have fewer explicit hints to help fill recognized personal details; mobile completion
may require additional typing. Mobile was not tested.
Built into the new site: Apply appropriate standard autocomplete tokens and preserve correctly associated labels.
Acceptance test: Inspect rendered form inputs for the agreed autocomplete tokens and suitable input types. Test
browser autofill on representative desktop and mobile browsers without sending a real enquiry.
https://www.andaraphuket.com/contact/

Detailed findings 8

Evidence: Title and phone are visibly marked required; Title choices are only Mr, Mrs and Ms. Address, country and
attribution questions are also present.
Guest or team impact: Guests who only need a short answer must navigate additional fields; required honorifics can
exclude preferred forms of address. Completion-rate effects are unmeasured.
Built into the new site: Remove or make honorific optional, offer preferred contact method and keep only information
needed for routing or response.
Acceptance test: Only approved essential fields are mandatory. Honorific is optional or removed; all intended guests
can complete the form without selecting an inaccurate title or unneeded contact channel.
https://www.andaraphuket.com/contact/

Evidence: The page lists six departmental contact routes, but the shared form has no visible
reason-for-contact/department selector.
Guest or team impact: Guests may be unsure which team receives a general form and what to include. Backend routing
was not inspected.
Built into the new site: Offer a small purpose selector with conditional fields and a clear response expectation; keep
direct contacts available.
Acceptance test: For each approved enquiry purpose, the form shows appropriate fields and communicates the
receiving team/next step. Use agreed test recipients in staging to verify routing before production enablement.
https://www.andaraphuket.com/contact/

Evidence: The public page emphasizes setting, service, cake and spa, followed by enquiry. No visible venue-capacity
table, indoor/rain alternative, package outline or venue-download link was found in the inspected page.
Guest or team impact: Couples cannot easily judge basic suitability before making contact.
Built into the new site: Add approved venue/capacity options, ceremony/reception formats, weather alternatives,
inclusions and a concise planning guide. All commercial details will be approved by Andara.
Acceptance test: Wedding content includes the agreed, client-approved suitability facts and downloadable material
where scoped. Capacities, weather alternatives and commercial statements match the approved source sheet.
https://www.andaraphuket.com/events/weddings/

Detailed findings 9

Evidence: The form marks Title, name, email, phone, country, date and free-text needs required; it has no explicit guest
count or flexible-date option.
Guest or team impact: Guests without a firm date lack an obvious way to express flexibility, while the team may need
follow-up to establish party size.
Built into the new site: Add approximate guest count and date flexibility; reduce mandatory fields and explain the next
step.
Acceptance test: An enquiry can express an approximate/flexible date and guest count; required fields and helper text
match the approved sales-team workflow. Confirm accessibility and staging receipt with consent.
https://www.andaraphuket.com/events/weddings/

Evidence: SILK page has descriptive dining copy, events and operating hours; no current food/drinks menu link was
present among the inspected page links.
Guest or team impact: Guests cannot easily check dishes or suitability before enquiring.
Built into the new site: Add accessible current menus, dietary guidance, venue details and a clear last-reviewed date.
Acceptance test: Menus and dietary information are available through clear restaurant-page actions, accessible on
desktop/mobile and connected to the approved update process; no expired document is linked at launch.
https://www.andaraphuket.com/dining/silk-restaurant/

Evidence: General reservation action is an inline email link; Italian Affairs uses Book Now to a WhatsApp URL; the
persistent site Book Now sends to hotel-room reservations.
Guest or team impact: The same booking language leads to different services, with no single restaurant-specific
primary action.
Built into the new site: Use Reserve a table or Enquire about this event labels, group phone/email/WhatsApp
consistently and keep room booking clearly identified.
Acceptance test: Each dining CTA unambiguously identifies table, event or room booking; destinations match that
purpose. Test the approved email/phone/WhatsApp actions without submitting a real reservation.
https://www.andaraphuket.com/dining/silk-restaurant/

Detailed findings 10

Evidence: The page describes Spa Haven, pool, gym hours/memberships and classes. In the inspected page links there
is no treatment menu, activity timetable or wellness enquiry CTA; only global/footer routes.
Guest or team impact: Interest in a treatment or class has no obvious direct continuation on the page.
Built into the new site: Add an approved spa menu, class schedule and relevant enquiry/booking route without
replacing the hotel booking engine.
Acceptance test: The scoped wellness page exposes its approved menu/schedule/enquiry routes; all routes work on
desktop/mobile and lead to the correct wellness contact or system.
https://www.andaraphuket.com/experiences/wellness/

Evidence: The map visibly displays This page can’t load Google Maps correctly and For development purposes only.
Selecting Phuket Old Town still updates its description and directions link.
Guest or team impact: The core spatial-planning interface is visibly degraded even though supporting attraction
selection remains functional. Exact configuration/billing cause is not established.
Built into the new site: Build a new destination component with validated map configuration and a useful static-map or
directions fallback.
Acceptance test: The new destination experience renders without Google Maps configuration dialogs or development
watermark on the production domain. Attraction selection and directions work; the fallback remains useful if the
embed cannot load.
https://www.andaraphuket.com/destination/

Evidence: Japanese offers show 50% summer/40% high-season savings and THB 13,450 starting price, 30% early-bird
and 20–25% long stay. English lists a different campaign set and up-to-15/20% savings. No judgment is made about
validity or market-specific intent.
Guest or team impact: Unexplained differences can create conflicting expectations across language versions and
booking handoffs.
Built into the new site: Validate market-specific offers with revenue management; use one governed offer model,
explicit locale/market eligibility and synchronized expiry/translation review.
Acceptance test: Revenue/content owners approve a per-language/per-market offer matrix. Live pages match that
matrix, including eligibility, dates, benefits and booking routes; intentional differences are clearly documented.
https://www.andaraphuket.com/jp/offers/
https://www.andaraphuket.com/offers/

Detailed findings 11

Evidence: Japanese navigation/content is localized, while header Book Now remains English and its URL is the same
hotel-only link with no locale parameter. The SynXis landing language was not checked.
Guest or team impact: There is no explicit handoff guarantee that the chosen language will continue through booking.
Built into the new site: Use translated action labels and a supported SynXis locale handoff; verify the full multilingual
journey.
Acceptance test: Choose Japanese, follow every booking CTA and verify the approved language handoff. Where SynXis
cannot preserve language automatically, the limitation and guest language choice are clear in the agreed flow.
https://www.andaraphuket.com/jp/offers/

Evidence: Discover More contains arrival 15 January and departure 16 January 2026, both in the past at review. The
link was inspected; engine handling of these dates was not tested.
Guest or team impact: The campaign intent may not match the availability search a guest expects.
Built into the new site: Use a shared booking-link component with validated parameters and an explicit generic-
property fallback.
Acceptance test: Inspect every launch CTA and exercise the intended date behavior in SynXis; no unintended past dates
remain.
https://www.andaraphuket.com/

Evidence: Great Savings on Offers contains arrival 6 May and departure 9 May 2023. This does not prove that SynXis
rejects the link.
Guest or team impact: An offer can carry obsolete search context into a high-intent booking step.
Built into the new site: Store booking configuration alongside each offer, with required owner and review date.
Acceptance test: The offer owner approves the destination, and the new CTA opens the expected property and valid
offer context.
https://www.andaraphuket.com/offers/

Detailed findings 12

Evidence: Sampled homepages use hreflang cn and the Chinese document uses lang cn-CN.
Guest or team impact: Language targeting and assistive pronunciation metadata are unreliable. Ranking loss was not
measured.
Built into the new site: Generate valid language metadata from one approved locale configuration, using the
appropriate zh language/script/region variant.
Acceptance test: Validate reciprocal alternates and document language for each launch language. Existing URL paths
need not change merely to repair metadata.
https://www.andaraphuket.com/
https://www.andaraphuket.com/cn/

Evidence: Sampled homepages use hreflang jp and the Japanese document uses lang jp-JP.
Guest or team impact: The Japanese version is not described correctly in language metadata.
Built into the new site: Use ja or the appropriate valid Japanese locale in the new language configuration.
Acceptance test: Check rendered lang and hreflang output, reciprocal references and canonical destinations.
https://www.andaraphuket.com/
https://www.andaraphuket.com/jp/

Evidence: The sampled page sitemap contains 38 URL entries but only 12 distinct locations. These are duplicate URL
records, not alternate-language annotations.
Guest or team impact: The generated search inventory is unnecessarily ambiguous and harder to validate. Indexing
harm is unproven.
Built into the new site: Generate a clean canonical URL inventory and validate sitemap output during migration.
Acceptance test: Each intended indexable page appears at its canonical URL without unintended duplicate records.
https://www.andaraphuket.com/page-sitemap.xml

Detailed findings 13

Evidence: The accommodation sitemap has 16 URL entries and four distinct locations, with suites repeated four times.
Guest or team impact: Room-page discovery output needs quality control alongside the new accommodation
structure.
Built into the new site: Include accommodation URLs in the same migration and sitemap validation process.
Acceptance test: Check all accommodation canonical URLs and ensure sitemap entries match the approved inventory.
https://www.andaraphuket.com/accommodation-sitemap.xml

Evidence: No /cn/, /jp/ or /ru/ location paths appear in the two sampled sitemaps. Other sitemaps and indexing
coverage were not exhaustively checked.
Guest or team impact: The full international discovery inventory remains uncertain.
Built into the new site: Reconcile all launch languages against generated sitemap coverage and Search Console.
Acceptance test: Every intended localized URL has an agreed discovery route and valid canonical/alternate
relationships.
https://www.andaraphuket.com/page-sitemap.xml
https://www.andaraphuket.com/accommodation-sitemap.xml

Evidence: The homepage title includes “5-Start”.
Guest or team impact: A visible brand-quality error can appear in browser or search presentation.
Built into the new site: Use a reviewed metadata field for every page, with prelaunch copy proofing.
Acceptance test: Inspect the actual title output and the preview text; no known typo remains.
https://www.andaraphuket.com/

Detailed findings 14

Evidence: The homepage social-preview image is a 115 by 120 pixel TripAdvisor logo.
Guest or team impact: Shared links miss the opportunity to communicate Andara’s setting and accommodation.
Built into the new site: Set property-led social images and page-specific social metadata, using approved photography.
Acceptance test: Preview the homepage and key campaign shares with correct image, title and description.
https://www.andaraphuket.com/

Evidence: Chinese and Japanese homepages have the same English title and description despite translated body
content.
Guest or team impact: The language experience becomes inconsistent at discovery.
Built into the new site: Provide translated metadata in the same publishing workflow as translated page copy.
Acceptance test: Native-language reviewers approve titles and descriptions, and output matches the selected
language.
https://www.andaraphuket.com/cn/
https://www.andaraphuket.com/jp/

Evidence: Signature Villas copy on the sampled Chinese and Japanese homepages remains English.
Guest or team impact: Guests switching language still encounter an incomplete accommodation story.
Built into the new site: Create shared content records with translation status and a launch-language completeness
checklist.
Acceptance test: Review every section and CTA in each agreed language before publication.
https://www.andaraphuket.com/cn/
https://www.andaraphuket.com/jp/

Detailed findings 15

Evidence: The Offers source contains no H1 or H2.
Guest or team impact: The page has less useful semantic structure for navigation and content understanding.
Built into the new site: Build campaign listing templates with a descriptive H1 and logical offer headings.
Acceptance test: Inspect rendered heading structure and navigate it with an accessibility tool.
https://www.andaraphuket.com/offers/

Evidence: Suites source contains no H1 and two Andara Suites H2s; duplication may reflect responsive templates.
Guest or team impact: The semantic structure is inconsistent; source duplication alone does not prove duplicated
visible content.
Built into the new site: Provide one clear page heading and correctly nested room-category headings in the new
template.
Acceptance test: Check desktop and mobile rendered/accessibility trees for an appropriate hierarchy without duplicate
announced headings.
https://www.andaraphuket.com/accommodation/suites/

Evidence: The sampled homepage uses one H1 followed by H3 section headings.
Guest or team impact: Section relationships are less explicit for non-visual navigation.
Built into the new site: Use a consistent heading hierarchy across all reusable sections.
Acceptance test: Check that section headings reflect the page structure and do not depend on font size alone.
https://www.andaraphuket.com/

Detailed findings 16

Evidence: Public output includes GTM, Meta PageView, legacy Universal Analytics and a GA4 tag in the GTM resource.
Presence does not prove accurate firing or duplicated events.
Guest or team impact: Marketing cannot judge attribution quality from installed-tag presence alone.
Built into the new site: Rebuild tracking from a documented event and tag-owner register, retaining only approved
destinations.
Acceptance test: Verify event receipt and deduplication in authorized accounts; no obsolete tags remain without a
documented purpose.
https://www.andaraphuket.com/

Evidence: The site hands guests to SynXis; confirmation events and revenue attribution were not inspected.
Guest or team impact: A booking click may be measurable while completed revenue remains unattributed or
fragmented.
Built into the new site: Include a tested measurement plan covering source site and supported SynXis instrumentation.
Acceptance test: Use an approved test path to reconcile booking IDs, value and currency; explicitly document any
engine limitation.
https://www.andaraphuket.com/

Evidence: The public consent callback shown logs preferences; tracking snippets appear in source. This is insufficient to
establish actual consent behavior or legal compliance.
Guest or team impact: Marketing and privacy owners need to know what runs before and after consent.
Built into the new site: Implement an approved consent approach and test it with the actual tracking stack.
Acceptance test: Test fresh, accepted, rejected and changed-preference sessions on the new site and relevant booking
flow.
https://www.andaraphuket.com/

Detailed findings 17

Evidence: Homepage source has 44 external script references and 24 stylesheet references. This count is not a page-
speed score.
Guest or team impact: Uncontrolled additions can compromise the new image-led experience.
Built into the new site: Keep a lean theme, document every third-party script and profile critical templates before
launch.
Acceptance test: Use repeatable device/network lab settings, inspect the waterfall and compare against agreed
budgets.
https://www.andaraphuket.com/

Evidence: Homepage markup contains 66 images and only one img srcset; JavaScript or picture-based variants were
not ruled out. Sample hero JPEGs are approximately 142 KB and 443 KB.
Guest or team impact: Large photography must remain sharp without wasting mobile bandwidth.
Built into the new site: Deliver responsive image sizes, suitable modern formats and intentional eager/lazy loading in
the new sections.
Acceptance test: Check downloaded image dimensions and bytes at agreed desktop and mobile sizes; the main visual
appears promptly without layout shift.
https://www.andaraphuket.com/

Evidence: The homepage viewport includes maximum-scale=1.0 and user-scalable=0. Actual mobile-browser
enforcement varies and was not tested.
Guest or team impact: Guests who need magnification may encounter avoidable restrictions.
Built into the new site: Use a responsive layout that permits zoom and reflow.
Acceptance test: Test zoom and text enlargement on agreed mobile browsers without loss of content or controls.
https://www.andaraphuket.com/

Detailed findings 18

Evidence: 53 of 66 homepage image elements have empty alt text, including decorative media and trackers. This is not
automatically 53 defects.
Guest or team impact: Useful property information may need descriptions, while decorative images should remain
quiet.
Built into the new site: Make meaningful-versus-decorative alt decisions part of the new image component and
handbook.
Acceptance test: Review each substantive image in context; descriptive images are announced appropriately and
decorative ones are ignored.
https://www.andaraphuket.com/

Evidence: Public assets identify WordPress/Divi/WPML, but no authenticated editor workflow was reviewed. Peem
explicitly requests independent campaign publishing.
Guest or team impact: Changing CMS without changing the workflow could preserve the same dependency on
developers.
Built into the new site: Provide bounded reusable templates, clear ownership, editor documentation and separately
billed whole-team training.
Acceptance test: The nominated editor completes the agreed campaign task independently using the new system.
https://www.andaraphuket.com/

Evidence: Headers show delivery-stack indicators, but hosting contracts, backups and support responsibilities were not
available.
Guest or team impact: Marketing needs a clear route when the website, form or booking handoff stops behaving as
expected.
Built into the new site: Deliver an ownership register, explicit support options, recoverable releases and a restore
procedure.
Acceptance test: Andara can identify the responsible party, access its accounts and demonstrate the agreed recovery
process.
https://www.andaraphuket.com/