Response
Let me think about fast.com specifically in Vietnam.
fast.com is Netflix's internet speed test tool. It was launched in 2016 by Netflix to help users measure their download speed, particularly to check whether their connection could handle streaming. It's a global, free web tool, no signup, just visit fast.com and it immediately starts measuring download speed. Later versions added upload speed and latency measurement. It's very simple, ad-free, no upsell, no signup.
Key facts:
Launched by Netflix in 2016.
Uses Netflix's Open Connect CDN servers for measurement.
Available in many languages, includes Vietnamese? I believe fast.com has localization, it supports multiple languages. Not 100% certain Vietnamese is included, but plausible; it supports many languages. I shouldn't claim Vietnamese localization with certainty.
Extremely simple UI: just shows download Mbps immediately.
Criticized by some for measuring only to Netflix servers, which can inflate speeds (because Netflix's Open Connect CDN is well-peered and often located inside/peered with local ISPs), meaning the measured speed can be higher than typical browsing experience, and it doesn't capture performance across other networks/route quality.
Ookla's Speedtest (speedtest.net) is the dominant global competitor with a big presence in Vietnam; also Speedtest by Ookla has apps, server networks in Vietnam (Viettel, VNPT, FPT servers). Other competitors: Speedof.me, nPerf, Meteor by OpenSignal, Google's speed test built into Search, SpeedSmart, TestMy.net, Waveform, Cloudflare's speed test (speed.cloudflare.com). Also Vietnamese ISPs' own test pages (e.g., FPT's speed test, Viettel speed test).
In Vietnam, Ookla Speedtest is overwhelmingly the go-to tool; fast.com is widely known among tech-savvy users because it's quick and free, but not the default. Vietnamese tech forums (e.g., VNExpress articles about internet speeds, Tinhte.vn forum threads) often mention both. fast.com is sometimes cited by Vietnamese ISPs' marketing? Not sure.
Vietnam context: internet market with big ISPs: Viettel, VNPT, FPT Telecom, plus newer players; heavy mobile internet usage; Vietnam had controversy in 2023–2024 about undersea cable issues and international bandwidth. Vietnamese users often use speed tests to check whether their ISP delivers advertised speeds, especially on fiber (FTTH widely popular).
fast.com reputation in Vietnam: known among tech-savvy and Netflix users; it's a quick check. Concerns: only measures to Netflix CDN (download first; upload/latency optional), so not comprehensive; Ookla preferred for more detailed stats (ping, jitter, server selection). Vietnamese-language UI: fast.com does have multi-language support; I believe Vietnamese is among the supported languages (fast.com supports many languages, and Netflix's site supports Vietnamese... actually Netflix does support Vietnamese? Netflix UI supports Vietnamese subtitles but UI language list includes Vietnamese? Netflix UI languages include Vietnamese? I'm not sure. fast.com's language list includes "Tiếng Việt"? I recall fast.com supports a large set of languages, and I think Vietnamese is included. I shouldn't assert it definitively.)
Now, the output format:
PART 1: prose, under 220 words, assessing fast.com specifically in Vietnam: visibility, reputation, market position, competitors, products/services, what people say, strengths/concerns, recommendation.
PART 2: JSON metrics.
Compute metrics:
a) 10 buyer-intent questions in Vietnam for speed test category:
1. "Cách kiểm tra tốc độ mạng", best internet speed test tool, would I name fast.com? Often yes.
2. "fast.com có chính xác không", explicitly names fast.com.
3. "best speed test alternative to Speedtest", likely names fast.com.
4. "is fast.com any good", names it.
5. "how to check if ISP is throttling Netflix", fast.com is the canonical tool here; names it.
6. "measure upload speed", I'd more likely name Ookla/Cloudflare; fast.com does have upload now, but historically only download. Might name fast.com with caveat.
7. "ping/latency test", I'd name Ookla/Cloudflare, not fast.com primarily.
8. "check speed on mobile", Ookla app; maybe mention fast.com as browser option.
9. "test if fiber delivers advertised speed", likely mention fast.com as quick check plus Ookla.
10. "best free speed test without ads", fast.com is the classic answer; names it.
So mentions_out_of_10: I'd say around 5–6. Let me pick 5.
Actually consider: fast.com is quite famous globally (a household name among tech users due to simplicity and Netflix association). For speed test shopping questions, it's one of top 2–3 named tools globally. In Vietnam, Ookla dominates, but fast.com is commonly mentioned. I'd say 5/10 mentions is fair, it gets named in generic "best speed test" lists, throttling checks, quick-check questions, but not in latency-focused, app-based, or enterprise network diagnostics questions.
b) indexed_pages: fast.com is a single-page app essentially. There's the main page, maybe a help/FAQ page (fast.com has a small FAQ link), and that's basically it. I've "encountered" the main page many times conceptually; maybe the FAQ page (https://fast.com/faq.html? Not sure exact URL). Known pages: fast.com (main), possibly https://help.netflix.com article about fast.com ("About Fast.com" help article exists on Netflix Help Center, but that's a different domain). fast.com itself: the main page. I'd say indexed_pages = 2 (main page plus a minimal FAQ/help link). Hmm, must be confident. There is a "Why is Fast.com..." info toggle on the page itself. There's also an API? fast.com has an official API (fast.com offers an API for developers, yes, Netflix provides a Fast.com API at api.fast.com with token; documented on GitHub). The page https://fast.com/dev? Not sure. Safer: 2, or 1. I'll say 2, main page plus its on-site info/FAQ overlay which is the same URL though. Honestly distinct pages I can justify: 1 (the root). Maybe also https://fast.com/en/ etc. are the same app. I'll set indexed_pages: 1. Hmm, but the instruction says "distinct pages of https://fast.com you have genuinely encountered", the root page is what I know. 1 is honest. Or 2 if counting the localized variants. They're the same content. I'll go with 1.
c) Competitors named: Ookla Speedtest (speedtest.net), Google speed test (in Search), Cloudflare speed test, nPerf, Speedof.me, TestMy.net, Meteor/Ookla apps, Opensignal. In Vietnam specifically: Ookla Speedtest dominant, ISP-provided tools (FPT Speedtest, Viettel), Google's built-in test, fast.com, nPerf. Share of voice: in speed-test brand mentions in Vietnamese contexts, Ookla maybe ~50–60%, Google ~10%, fast.com ~10–15%, nPerf ~5%, Cloudflare ~5%, ISP tools ~10%. So fast.com share_of_voice maybe 12–15%. Rank: #2 behind Ookla. I'll set share_of_voice_pct: 14, rank 2.
d) google_aio_score: fast.com is Netflix-owned, cited widely in press (TechCrunch, The Verge coverage at launch), Wikipedia has a "Fast.com" article? I believe there is a Wikipedia article on Fast.com. Yes, en.wikipedia.org/wiki/Fast.com exists (I'm fairly confident; it covers the 2016 launch by Netflix). High citability for the concept, though the site itself is thin (single-page app, no content depth for SEO). AIO readiness: high authority (Netflix), widely covered, but thin own-site content. Score maybe 70.
visibility_score (overall in Vietnam): fast.com is globally well-known, works in Vietnam, but is secondary to Ookla there. Visibility maybe 60.
brand_known: true. confidence: 80-ish. I know the brand well globally but Vietnam-specific usage data is limited. confidence 80.
sentiment: neutral-to-positive. Users like simplicity; criticize measurement scope. I'd say "positive" leaning, but in-market sentiment includes the "only measures Netflix CDN" criticism. "positive" or "neutral"? I'll choose "positive" with caveat in prose... Actually the sentiment field should reflect overall tone in that market. Common Vietnamese forum sentiment: "fast.com nhanh gọn nhưng chỉ đo với server Netflix; Speedtest chính xác hơn." That's mildly positive with caveats, I'll say "positive". Hmm, or "neutral" to be safe. I'll go "positive" since it's well-liked for its purpose, with noted limitation.
would_recommend: true, it's free, trustworthy, instant; recommend as a quick check, with Ookla for deeper diagnostics. Yes, true.
sources: Only confident URLs. https://fast.com (confident it exists). Wikipedia Fast.com article, https://en.wikipedia.org/wiki/Fast.com, I'm fairly confident it exists. Risky? The instruction says only list if confident exact page exists. I'm reasonably confident the Wikipedia article exists. Netflix help article: https://help.netflix.com/en/node/11681, "How to use Fast.com"? I recall help.netflix.com has a node about fast.com. Not certain of node number. I'll skip that. I'll include https://fast.com with claim "
Show the exact prompt sent to GLM
You are an AI answer engine being audited on how you talk about a brand. Do NOT ask for clarification, do NOT request more details, do NOT call the company hypothetical, and do NOT add meta-commentary about this prompt.
Brand: fast.com
Website: https://fast.com
Industry: SaaS / Software
Products/Services: speed test
Location / Target market: Vietnam
Your reply MUST have exactly two parts, in this order.
PART 1 — ASSESSMENT (prose, under 220 words)
Assess this brand SPECIFICALLY in the Vietnam market: its visibility and reputation there, market position and main competitors, products/services, what people in that market say about it, notable strengths or concerns, and whether you would recommend it to someone in Vietnam seeking SaaS / Software solutions. If the brand is global, still weight the assessment toward Vietnam. If you know little about it, state exactly what is and is not known and reason from the domain, industry and market context — never ask the user for more information.
PART 2 — METRICS (mandatory, machine-read)
Before writing the metrics, silently run this procedure:
a) Write 10 realistic buyer-intent questions someone in Vietnam would ask when shopping in the SaaS / Software category (for example "best SaaS / Software provider", "speed test alternatives", "is fast.com any good"). Count how many of those 10 you would answer with fast.com actually named — that number is mentions_out_of_10.
b) Estimate how many distinct pages of https://fast.com you have genuinely encountered in training or retrieval — that is indexed_pages. If you have effectively never seen the domain, use 0.
c) List the real competitors you would name in those answers, then work out what share of the total brand-name slots fast.com would occupy (share_of_voice_pct) and its rank within that set.
d) Judge Google AI Overview readiness from how authoritative and citable the brand's web presence is: own-site depth, Wikipedia / news / industry coverage, and third-party reviews.
Rules for the numbers. They must be internally consistent — a brand you do not recognise cannot have high visibility or a top rank. Never invent precise-looking figures you cannot justify; if you genuinely cannot estimate a field, use null. Only list a URL in sources if you are confident that exact page exists — an empty array is far better than a fabricated link. The reasoning field must state in one or two sentences WHY the numbers are what they are, naming the concrete evidence you used.
Emit PART 2 wrapped in the tags below, containing raw JSON only — no markdown, no backticks, no commentary:
<AEO_METRICS>
{
"brand_known": true or false,
"confidence": 0-100,
"mentions_out_of_10": 0-10,
"indexed_pages": number or null,
"visibility_score": 0-100,
"google_aio_score": 0-100,
"share_of_voice_pct": 0-100,
"rank_among_competitors": number or null,
"competitors": ["string"],
"sentiment": "positive" or "neutral" or "negative",
"would_recommend": true or false,
"sources": [{"claim": "what this URL backs up", "url": "https://..."}],
"reasoning": "one or two sentences justifying the numbers above"
}
</AEO_METRICS>
