Anatomy of a Brand Conversation

HubSpot's audiences, jobs & pains — an ontology from 3,677 public signals
Who is actually talking about HubSpot online, what they are trying to get done, and where it hurts — reconstructed from 3,677 public data points: Google search demand, Reddit, independent forums, HubSpot's own community, and six social platforms. United States · July 2026.
2,033Google keywords
743Reddit mentions
300Forum mentions
280Social posts — LinkedIn, Bluesky, Pinterest, TikTok, Instagram, Threads

How to read this study

Every keyword and post was classified along four dimensions. Voice type — is this a genuine ask or need (intent), an experience shared (story), promotional content, or noise (the word "hubspot" appears, the conversation isn't about it)? Audience — who is speaking: existing customers, prospects evaluating a purchase, developers building integrations, agencies & partners, learners, job seekers, or investors. Job-to-be-done — what they are trying to accomplish: LEARN (understand), DECIDE (choose between options), ACQUIRE (buy or set up), USE (operate day-to-day), RESOLVE (fix a problem), BENEFIT (realize value). Progress type — whether the motivation is practical (functional), social/emotional (status, frustration, belonging), or neutral information-seeking. Charts are interactive: click segments, bars and map nodes to see the underlying keywords and posts.

1 · Search demand — Google 2,033 keywords → 1,279 on-topic → 757 unique

What people type into Google around HubSpot. On-topic = the query names the brand (1,279 of 2,033). Word-order and spelling variants are merged into 757 unique queries ("hubspot login / login hubspot / hubspot log in" count once).

Top unique keywords — volume & growth

KeywordVolumeGrowth %Variants

Fastest growing (volume ≥ 100)

KeywordVolumeGrowth %

The audience ontology of search

HubSpot at the center → branches to each audience segment → each audience ramifies into its jobs-to-be-done. Node numbers = keyword counts, node size = share. Click any node to see the actual queries.

2 · Communities 743 Reddit + 300 Forums

1,043 public posts mentioning HubSpot: 743 from Reddit and 300 from online forums. Every post is classified by voice type — the key move of the study, because most of what looks like "brand conversation" is not audience voice.
87 of 300 forum posts come from community.hubspot.com — the brand's own support forum. Included, but flagged: owned-community voice is not independent audience voice.

Reddit — 743 posts by voice type

Independent forums — 213 posts

HubSpot's own community — 87 posts

Reddit: only 169 posts (23%) are genuine intent and 76 (10%) genuine stories — while 194 (26%) are promotion (templated SEO content farms and astroturfed "we switched to HubSpot" testimonials) and 304 (41%) are noise. Independent forums look even more polluted: 57% is promotional content from blog-style pages posing as forums, though the real communities (Webflow, n8n, Make) are almost pure intent. On HubSpot's own community nothing is noise, but 26% is the brand's own promotional content (partner announcements, job ads) rather than customer voice.

Temporal distribution — Reddit posts per year by voice type

All 743 Reddit posts by publish year. Forum sources expose no publish date — their intent posts are counted in the Unknown box.

All 743 Reddit posts per year, by voice type

Unknown date

The audience ontology of communities

All 396 genuine-voice posts (intent + story) from Reddit and online communities (incl. HubSpot's own), grouped by audience segment. Each audience splits into intent (full audience color, ramifying into jobs-to-be-done) and story (same color, lighter — no JTBD by design). Node numbers = post counts. Click any node for the posts inside.

3 · Social platforms 280 posts

LinkedIn 113 · Bluesky 95 · Pinterest 23 · TikTok 18 · Instagram 16 · Threads 15 — classified with the same voice-type, audience and job-to-be-done lenses. (Reddit is analysed in the Communities section.)

Voice type — 280 posts

Voice type by platform

Jobs-to-be-done × progress type (86 genuine posts)

Social platforms are dominated by promotion (57%) — agencies, consultants and content marketing talking at the market. Genuine audience intent is rare: 7 posts (2.5%). Stories (28%) carry most of the remaining signal.

The audience ontology of social platforms

The 86 genuine-voice social posts (7 intent + 79 story) grouped by audience segment, same reading as the Communities map: intent in the audience color ramifying into jobs-to-be-done, story lighter, no JTBD. Click nodes for the posts with links. Note the shape: on socials, story massively dominates — genuine intent is rare and concentrated in Evaluators (DECIDE) and Customers.

Semantic clusters (86 genuine posts)

4 · Pain map where it hurts · negative sentiment · all channels

Sentiment analysis applied to every kept (intent + story) post across Reddit, online communities and socials. The 45 negative-sentiment posts (35 Reddit · 4 forums · 6 social) were thematically clustered and double-reviewed. Result: 31 genuine product/martech pains in 5 themes; 14 posts explicitly excluded as "not a product pain" (news reposts, phishing-scam warnings, stock chatter, generic advisory content). Click a theme card for the posts with links.

Where it hurts — 31 genuine pains, sized and colored by who voices them

Bar length = number of posts · segment colors = audience · click any bar for the posts with links. Pricing, contracts & vendor trust erosion is the dominant pain — twice the size of actual product breakage. The greyed row holds negative-tone posts excluded as not product pain.