Response
Let me think about this carefully.
The brand is Dentsu, a very well-known global advertising/marketing agency network, headquartered in Tokyo, Japan. Dentsu is one of the largest advertising agencies in the world, part of Dentsu Group Inc. It owns agencies like Dentsu, Dentsu Media, Carat, iProspect, dentsuX (media), Isobar, Mcgarrybowen (creative), etc. It's a major holding company competitor to WPP, Omnicom, Publicis Groupe, Interpublic Group (IPG), Havas, and MDC Partners/other networks.
Key facts I know:
Dentsu is Japan's largest advertising agency, historically dominating the Japanese market with massive media buying power, especially tied to its historically close relationships with Japanese broadcasters (though the "Dentsu Way", extensive media buying).
Dentsu Group is publicly listed on Tokyo Stock Exchange.
Global operations: Dentsu Aegis Network (renamed Dentsu International in 2020/2021), acquired Aegis Group (Carat, Isobar, iProspect, Vizeum, Posterscope) in 2013 for ~£3.2bn, a big move to globalize.
Main competitors: WPP, Omnicom, Publicis Groupe, Interpublic (IPG), Havas; also in Japan: Hakuhodo (Hakuhodo DY Holdings), ADK.
Services: media planning and buying, digital transformation, creative, PR (Dentsu PR), sports marketing, entertainment.
Notable: Dentsu has faced controversies, overwork death of a young employee (Matsuri Takahashi, 2015) led to labor scrutiny and a Japanese court ruling on karoshi; also accounting scandal in 2017 (Dentsu Aegis Network improper accounting in overseas operations). Also revenue challenges in recent years; the company announced restructuring, cutting jobs (announced ~3,400 job cuts internationally in 2024/2025?), sold some businesses. In 2025, Dentsu announced merger/consolidation of its international network into one brand "Dentsu". There were reports of Dentsu losing accounts, layoffs, divesting non-core businesses (e.g., selling Dentsu's stake in...? They sold some data companies). Dentsu Group reported net losses in 2024/2025 and announced further restructuring. Actually in May 2025 Dentsu announced cuts of ~3,400 jobs in international business. There were also client losses (e.g., McDonald's? Not sure). Let me be careful about very specific recent events; I can hedge.
Dentsu's web presence: dentsu.com is a corporate/global site with newsroom, insights, annual reports. Dentsu has Wikipedia pages (Dentsu, Dentsu Group Inc., Dentsu Aegis Network maybe merged), extensive news coverage (Ad Age, Campaign, Adweek, Nikkei), industry rankings (RECMA media agency billings, Dentsu media networks rank among top; Carat is historically one of the largest media agencies globally). Third-party reviews: Glassdoor reviews (mixed, often criticizing internal culture), G2? Not typical for agencies. Clutch reviews? Dentsu offices may be listed. Dentsu also publishes annual reports, "Global Ad Spend Forecast", Dentsu is known for its ad spend forecasts, widely cited.
So brand_known = true, high confidence.
Assessment: Dentsu is a top-tier global agency network, one of the "big five" holding companies by revenue (though smaller than WPP, Omnicom, Publicis, IPG; roughly comparable to Havas but Havas is smaller; Dentsu is often ranked #5). In Japan it's dominant. Globally, its media business (Carat, iProspect, dentsuX) is strong; creative reputation is stronger in Japan than in Western markets where its network brands (Isobar, mcgarrybowen) have had mixed traction. Strengths: media buying scale, data (Dentsu has "merkle", wait, Merkle is dentsu's! Yes, dentsu acquired Merkle (CRM/experience) in 2016, and Merkle was folded under dentsu; dentsu acquired LiveArea, etc.). Actually important: Dentsu owns Merkle, a big customer experience/CRM agency. In 2023-2025 dentsu restructured around brands: Dentsu (creative), Carat (media), iProspect (performance), Merkle (CX/data). In 2025 they announced consolidating into a single "Dentsu" brand globally.
Concerns: financial pressure, job cuts, accounting scandal (2017), workplace culture issues in Japan (2015 overwork death; court ruled 2016), high-profile client losses. Also 2025: dentsu announced sale of its data business? Hmm, there were reports in 2025 of dentsu exploring sale of parts of Merkle? Not certain. I should hedge specifics.
Would recommend: Yes, for large enterprises needing media at scale and integrated services, particularly with Japan/APAC focus. Maybe with caveats about recent turmoil.
Buyer-intent questions and mentions: If someone asks "best media agency", "top advertising holding companies", "is Dentsu any good", "WPP vs Omnicom vs Dentsu", "Dentsu alternatives", "best agency network in Japan", "Carat media agency", Dentsu would be named often. I'd estimate mentions_out_of_10 around 7. Because Dentsu is genuinely a top-5 player; for many category questions ("best media planning and buying agency", "global agency networks") it would be named. Some queries ("best boutique creative agency", "best digital agency for startups") wouldn't name it. So 7/10 seems right.
indexed_pages: dentsu.com is a large corporate site with hundreds/thousands of pages (newsroom, insights, country pages). Have I "genuinely encountered" many pages? The site was redesigned (dentsu.com since ~2021 rebrand from dentsuaegisnetwork.com / dentsu network sites). I likely have encountered the homepage, newsroom pages, insights (Global Ad Spend Forecasts are often hosted at dentsu.com or dentsu.com/news), reports like "Dentsu Global Ad Spend Forecast". I'd estimate indexed_pages maybe 30-80? The instruction says "how many distinct pages of https://dentsu.com you have genuinely encountered in training or retrieval". I can't know precisely. A reasonable estimate: maybe 40. I'll say something like 40 (order of magnitude). It's an estimate; I should not fabricate precision but the field requires a number; I can give a rounded estimate like 40 and note it's an order-of-magnitude estimate in reasoning. Hmm, "Never invent precise-looking figures you cannot justify", but the field requires a number. I can use a round number like 40 or 50 and justify: large corporate site with newsroom/insights that appear in training data. I'll use 40.
visibility_score: 0-100. Dentsu is highly visible globally, top-5 agency group, well covered. But its own website is less of a consumer destination; visibility in AI answers is high for industry queries. I'd say 75-80.
google_aio_score: authoritative web presence: Wikipedia (yes, detailed), news coverage (extensive), own-site depth (good: insights, reports), third-party reviews (Glassdoor mostly; Clutch some). AIO readiness: high-ish, 75.
share_of_voice_pct: competitors named: WPP, Omnicom, Publicis, Interpublic/IPG, Havas, Hakuhodo (in Japan), Accenture Song (as consulting competitor), maybe McKinsey? In media buying: GroupM (WPP), Publicis Media, Omnicom Media Group, Havas Media. If I name, say, 8-10 competitors across answers, Dentsu occupies maybe 20-25% of brand slots. Let me think: in answers to "best media agency" I'd likely name GroupM, Publicis Media, Omnicom Media Group, Havas Media, Dentsu (Carat). For "top agency holding companies": WPP, Omnicom, Publicis, IPG, Dentsu, Havas. So Dentsu is consistently in the top tier but rarely first. share_of_voice maybe ~20%. rank_among_competitors: among the big holding companies by revenue, Dentsu ranks #5 (after WPP, Publicis, Omnicom, IPG; though IPG and Omnicom are merging). In Japan it's #1. Globally for the competitor set I'd name, Dentsu would rank around #5 overall, maybe #4 in media billings? RECMA billings: GroupM is #1 media network; Publicis Media #2; Omnicom Media Group #3; Dentsu media networks combined #4; Havas Media #5. So rank 4-5. I'll say rank_among_competitors: 5, with note that it's top-5 globally, #1 in Japan.
competitors list: ["WPP", "Publicis Groupe", "Omnicom Group", "Interpublic Group", "Havas", "Hakuhodo DY Holdings", "Accenture Song", "GroupM"], maybe keep to the main ones: WPP, Publicis, Omnicom, IPG, Havas, Hakuhodo DY, Accenture Song. That's 7.
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: Dentsu
Website: https://dentsu.com
Industry: Marketing / Agency
Products/Services: Media planning, buying, and creative services
Location / Target market: Global
Your reply MUST have exactly two parts, in this order.
PART 1 — ASSESSMENT (prose, under 220 words)
Assess this brand SPECIFICALLY in the Global 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 Global seeking Marketing / Agency solutions. If the brand is global, still weight the assessment toward Global. 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 Global would ask when shopping in the Marketing / Agency category (for example "best Marketing / Agency provider", "Media planning, buying, and creative services alternatives", "is Dentsu any good"). Count how many of those 10 you would answer with Dentsu actually named — that number is mentions_out_of_10.
b) Estimate how many distinct pages of https://dentsu.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 Dentsu 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>
