Website audit · glicowings.co.id
Beautiful on screen.
But the path to the ice cream breaks.
A 6-layer audit of PT Glico Wings Indonesia’s official website: what a visitor actually experiences, what it costs the brand, and what to fix first.
22 October 2026 · first audit (baseline)
Tested as an anonymous visitor, desktop and phone
Website audit · glicowings.co.id
Beautiful on screen.
But the path to the ice cream breaks.
A 6-layer audit of PT Glico Wings Indonesia’s official website: what a visitor actually experiences, what it costs the brand, and what to fix first.
22 October 2026 · first audit (baseline)
Tested as an anonymous visitor, desktop and phone
In one minute
The site is fast and on-brand, but it doesn’t do the brand’s main job: help people find a flavour and buy it
| Layer | Today |
|---|---|
| 0 · Strategy & goal | Poor |
| 1 · Functionality | Poor |
| 2 · Technical quality | Poor |
| 3 · Usability (UX) | Poor |
| 4 · Content | Poor |
| 5 · Look & feel | Fair |
| Cross-cutting · Privacy law (UU PDP) | Fail |
| Cross-cutting · Accessibility | Poor |
Server response
0.15 s
fast — the foundation is fine
Homepage download
3.0 MB
one banner alone is 1.44 MB
Product images with alt text
0 / 22
invisible to screen readers and Google Images
Latest news on the site
2021
activities stop in October 2021
Scale: Fail · Poor · Fair · Good. No Blocker was found — the main flows work somewhere — but 13 findings are Major.
A visitor’s journey today
“I saw a Frostbite ad on Instagram. Where can I buy one, and is it halal?”
1 · Opens the site on her phone
A red loading screen, then a mascot in a circle and the line “move your cursor” — on a phone with no cursor.
Findings #19, #4, #23
2 · Looks for Frostbite
Brands come one at a time in a carousel. Frostbite is the second slide; nothing tells her that.
Findings #21
3 · Picks a flavour
12 small round icons with no names. She has to tap them one by one to read what they are.
Findings #22, #13
4 · Tries to buy
A tiny “Dapatkan di” link (11 px) leads to one Blibli store where the first items show “Stok habis”.
Findings #3
5 · Asks about halal
No ingredients, no halal logo, no contact number, no e-mail. The question has nowhere to go.
Findings #6, #5
Every step adds friction, and the last two end in a dead end. Meanwhile Meta Pixel and GA4 have been tracking her since step 1 — without asking (#1).
Illustrative scenario built only from what we observed on 22 Oct 2026; “#” numbers refer to the findings list at the end of this deck.
What to fix first
Three things matter before any visual work
1 · Ask before tracking
Meta Pixel and GA4 run from the first second, with no consent banner and no privacy policy. Under Indonesia’s Personal Data Protection law (UU PDP) this is a legal risk for the company.
2 · Open the path to the product
Show products in the first screen, fix the footer link that opens a blank page, and give “buy” a clear button with more than one store.
3 · Answer the basic questions
Contact details, ingredients, nutrition, allergens, BPOM number and halal status are missing — the things parents and careful buyers check first.
Findings #1, #2–4 and #21–22, #5–6.
How we audited
Bottom-up: if a button doesn’t work, its colour doesn’t matter yet
| # | Layer | Key question |
|---|---|---|
| 0 | Strategy | Who is it for, and what is the one action we want? |
| 1 | Functionality | Do the main flows work from start to finish? |
| 2 | Technical | Fast, secure, stable, findable by search engines? |
| 3 | Usability | Easy for the real audience, on the devices they use? |
| 4 | Content | Accurate, clear, current, complete? |
| 5 | Look & feel | Does the design support the message and trust? |
Severity
a main flow fails completely
hurts many visitors, or legal / security risk
annoying, but there is a way around it
an opportunity
Plus five cross-cutting checks on every layer: privacy law, data integrity, accessibility, measurement, operations.
Layer 0 · Strategy & goal
The site hasn’t chosen the one thing it wants visitors to do
Audience is implied, not stated
Copy like “Hangout bareng temen” and “auto sat set” recipes speaks to young people and families — consistent, but never written down as a goal.
No main action
The homepage offers “Lihat Detail” and “Our History”. Buying only appears deep in product pages, small, and to one store.
Measured, but not for sales
GA4 and Meta Pixel are installed, but we could not see any event for clicks to a store, so the site’s contribution to sales is unknown.
Recommendation
Pick one goal — a click on “Buy” per flavour — and track it as a GA4 conversion.
Finding #17.
Layer 1 · Functionality
Two of the paths to a product end in the wrong place
A blank page in the footer
“Our Products” in the footer links to /products/wakuwaku (status 200, 0 bytes). The homepage cards use /products/waku-waku.
Buying ends in “out of stock”
“Dapatkan di” redirects to the Glico Wings flagship store on Blibli; on 22 Oct the first product row showed “Stok habis”.
What works
The homepage loads with a clean console. /about, /contact and /privacy-policy return 404 — they simply don’t exist.
Findings #2, #3. Not Blocker: the header menu still reaches the products.
Layer 2 · Technical quality
The server is fast, but the pages are heavy and unprotected
Heavy
3.0 MB downloaded on the homepage; a single JPEG banner is 1.44 MB. On a mid-range phone on 4G that is seconds of waiting.
Unprotected
No HSTS, CSP, X-Frame-Options or X-Content-Type-Options. Session cookie without Secure. jQuery 3.3.1 (2018) with known CVEs. The header reveals Apache/2.4.58 (Ubuntu).
Hard to find on Google
No meta descriptions or Open Graph tags, /sitemap.xml returns 404, and product names are filled in by JavaScript (“{{ subProductActive.title_1 }}” in the raw HTML).
What’s good
0.15 s server response, HTTPS with a 301 redirect, and a valid certificate until January 2027.
Findings #7, #8, #9, #16.
Layer 3 · Usability (UX)
The products are hidden behind a gesture visitors have to guess
An empty first impression
The desktop homepage opens on a grey circle and “press and move your cursor”. Visitors who don’t press never see a product.
Flavours without names
Product pages use a numbered carousel and round icons; the flavour name appears only after a click.
Phones are locked down
Zoom is disabled (user-scalable=0) and 13 of 13 buttons are smaller than the 44 px touch standard.
Recommendation
Put the product in the hero and keep the gesture as a reward, not a gate. Name every flavour.
Findings #4, #11, #13, #19–23. Annotated screens on the next four slides.
UX · desktop homepage, first screen
The first screen shows a mascot, not ice cream
1
2
3
1
Product hidden. The circle holds the taiyaki mascot; the ice cream appears only after pressing and holding.
2
13 px instruction at the very bottom of the screen — easy to miss.
3
Menu hidden behind a hamburger icon, even on a wide desktop screen.
Before this screen, a full red loading screen shows for a few seconds (#19). Screenshot 22 Oct 2026, 1280×800. Yellow boxes = problem areas.
UX · desktop homepage, products
Four brands queue up in one carousel
1
1
2
3
4
1
One brand per screen. Frostbite, JCone and Waku Waku only appear after clicking the arrows.
2
14 px description with a typo (“berkualiltas”).
3
The only button is “Lihat Detail” (115 × 40 px). There is no buy button.
4
The menu icon sits on top of the product photo.
Findings #21, #14, #20. Screenshot 22 Oct 2026, 1280×800, scrolled to Our Products.
UX · product page (Frostbite)
Flavours are chosen through icons without names
1
2
3
4
1
12 flavours as small round icons with no names; you click them one by one.
2
Switching brand through small logos in a sidebar, without labels.
3
Another gesture: “press and move the cursor”, in 12–13 px text.
4
A weak buy link: “Dapatkan di” in 11 px, same weight as “Lihat Detail”, one store only.
Findings #22, #3. Screenshot 22 Oct 2026, 1280×800, /products/frostbite.
UX · phone (375 px)
On a phone, visitors are told to move a cursor they don’t have
1
2
3
4
5
1
Still hidden: the product sits behind the mascot circle.
2
“Move your cursor” shown on a touch screen.
3
12 navigation dots of 20 px with no labels.
4
Buttons 35 px and 26 px tall, 10 px text. The touch standard is 44 px.
5
Footer links 23 px tall, 12 px text, crowded together.
And zoom is locked (user-scalable=0), so the small text cannot be enlarged.
Findings #4, #23, #13, #11. Viewport 375×812.
UX fix plan
Products visible without a gesture, buying in two taps
| Area | Today | Fix | Done when |
|---|---|---|---|
| Navigation | Hamburger menu, also on desktop | Text menu in the header from 1024 px | Menu visible without a click |
| Phone sizing | Buttons 20–35 px, zoom locked | Touch targets ≥ 44 px, zoom on, text ≥ 16 px | 0 targets under 44 px |
| Hero | Loading screen, then mascot | No splash; hero shows products and two buttons | A product in the first screen |
| Brands | Carousel, 1 brand per screen | Grid of 4 brand cards | 4 brands without arrows |
| Flavours | 12 round icons without names | Named flavour grid with photos | Every flavour name readable |
| Buy | “Dapatkan di” 11 px, 1 store | Main “Buy” button + choice of ≥ 2 stores | “Buy” click tracked in GA4 |
The press-and-hold gesture can stay as a playful extra — as long as the product is already visible without it.
Layer 4 · Content
News stopped in 2021, product info stops at a promotional paragraph
Out of date
The latest item on /activities is dated 21 October 2021; the latest article is from May 2019.
Product info is missing
“Lihat Detail” only shows marketing copy — no ingredients, nutrition, allergens, BPOM number or halal logo.
Typos and mixed language
“berkualiltas”, “sebelum nya”, “Chasew”, “pemenang nya”; “Our History” next to “Lihat Detail”.
Leftovers
GTM template HTML (“Creation Date/22/2021”) is visible in the page source. The copyright year (2026) is current.
Findings #6, #12, #14, #18.
Layer 5 · Look & feel
A strong, recognisable look — which is also what makes it heavy
Keep
Consistent brand red, Japanese motifs (お祝い / Celebration) that support “From Japan to Indonesia”, the taiyaki mascot and good recipe photography.
Reduce
Large PNG ornaments (chocolate, almonds, waves) add weight. Thin small text on photos is hard to read, especially on phones.
Direction for a redesign
Keep the identity, products and playful interaction — but let the product lead and the decoration follow.
This layer is assessed last on purpose: the visual identity is an asset worth keeping.
Cross-cutting checks
Visitors are tracked, but there is nobody to contact
Privacy law (UU PDP)
Meta Pixel and GA4 run without a consent banner. No privacy policy (/privacy-policy is 404), no address, no data controller named.
Accessibility
22 of 22 product images without alt text, empty lang attribute, zoom locked on phones.
Measurement
GA4 and Pixel are installed, but no event for clicks to a store — the site’s contribution to sales can’t be measured.
Operations
glicowings.co.id has no MX record, so it cannot receive e-mail; content not updated since 2021. SSL renews automatically.
Data integrity: not applicable — the site hosts no user reviews or ratings.
All findings (1/2)
Findings 1–12
| # | Area | Finding | Severity | Recommendation |
|---|---|---|---|---|
| 1 | Legal | Meta Pixel and GA4 run before any consent; no privacy policy | Major | Privacy page, consent banner, Consent Mode v2 |
| 2 | Function | Footer “Our Products” opens a blank page (/products/wakuwaku) | Major | Fix the link; 301 the wrong URL |
| 3 | Function | “Dapatkan di” goes to one Blibli store; many items out of stock | Major | More channels; hide links when out of stock |
| 4 | UX | Homepage hero is empty until you press and hold | Major | Show products at once; keep the gesture as a bonus |
| 5 | Content | No contact, address or customer service | Major | Contact page with company, e-mail, WhatsApp |
| 6 | Content | No ingredients, nutrition, allergens, BPOM or halal info | Major | Nutrition & certification tab per flavour |
| 7 | Technical | Homepage downloads 3.0 MB; one banner is 1.44 MB | Major | WebP under 200 KB, lazy loading |
| 8 | Technical | No HSTS/CSP; session cookie not Secure; jQuery 3.3.1 | Major | Security headers; Secure cookie; update jQuery |
| 9 | Technical | No meta description, Open Graph or sitemap; names rendered by JS | Major | Meta per page, sitemap, names in the HTML |
| 10 | Access | 22 of 22 product images have no alt text; empty lang | Major | Alt text with the flavour name; lang="id" |
| 11 | Access | Zoom is locked on phones | Major | Remove maximum-scale / user-scalable |
| 12 | Content | Latest activity Oct 2021, latest article May 2019 | Minor | Publish 2025–2026 news or pull from Instagram |
All findings (2/2)
Findings 13–23
| # | Area | Finding | Severity | Recommendation |
|---|---|---|---|---|
| 13 | UX | 13 of 13 tap targets on phones are under 44 px | Minor | Tap areas of at least 44 × 44 px |
| 14 | Content | Typos and mixed English–Indonesian | Minor | Proofread; one language for navigation |
| 15 | Operations | Domain has no MX record, cannot receive e-mail | Minor | Set up MX and a customer-service inbox |
| 16 | Technical | Server version exposed; 404 page has no title | Minor | Hide the version; titled 404 with menu |
| 17 | Strategy | No conversion goal or event | Suggestion | GA4 event for clicks on “Dapatkan di” |
| 18 | Content | Leftover GTM template comment; two Facebook URLs | Suggestion | Clean the HTML; one Facebook URL |
| 19 | UX | Full-screen red loading screen for seconds | Minor | Remove the splash |
| 20 | UX | Hamburger menu even on desktop; icon covers the photo | Minor | Text menu in the header from 1024 px |
| 21 | UX | 4 brands in a carousel, only 1 visible | Major | Grid of 4 brand cards |
| 22 | UX | 12 flavours shown as unnamed round icons | Major | Named flavour grid with photos |
| 23 | UX | “Move the cursor” shown on phones | Minor | Disappears with the new hero |
#19–23 were added after the annotated UX review. Status of every finding is tracked in the revamp proposal deck.
Roadmap
Three phases: remove the risks, open the path, then keep it alive
Phase 1 · week 1 — Remove risks
Privacy policy and consent banner (#1), footer link (#2), security headers and cookie (#8), unlock zoom (#11), desktop menu (#20), MX and contact (#5, #15).
Phase 2 · weeks 2–4 — Open the path
Hero without a gesture (#4, #19, #23), brand and flavour grids (#21, #22), store choice (#3), nutrition & halal (#6), image weight (#7), meta & sitemap (#9), alt & lang (#10).
Phase 3 · ongoing — Keep it alive
Regular news (#12), proofreading (#14), touch targets (#13), conversion events (#17), then a re-audit against this baseline.
Scope and limits
What this audit did and did not test
| Tested |
|---|
| 9 main pages and the footer links |
| Headers, certificate, DNS, image and script sizes |
| Console, network and every homepage CTA on desktop |
| Homepage and recipes on a 375 px phone |
| The Blibli buy destination |
| Not tested |
|---|
| Admin area or anything behind a login |
| Recipe search and activity detail pages |
| Field Core Web Vitals and Lighthouse scores |
| Full screen-reader and keyboard navigation |
| Which GA4 events are actually sent |
First audit, 22 October 2026. This deck is the baseline for the next re-audit.