Knowledge base and self-service
Asked as “knowledge base software”, and as “customer self-service help center tool”, on behalf of a mid-market B2B software company. 49 first choices recorded across the direct, paraphrase, budget and scale prompts, six models each.
01The standing
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all six models, for a mid-market B2B software company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
Show the five products at 0%, ordered by negative rate
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its vendor page.
Recommended versus criticized
Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.
Criticized challengerCriticized default
Negative label rate →
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative30%
Key
01Notion18%
Notion18%02Document36014%
Document36014%03Freshdesk12%
Freshdesk12%04Confluence6%
Confluence6%05Zendesk6%
Zendesk6%06BookStack6%
BookStack6%07Guru4%
Guru4%08Helpjuice4%
Helpjuice4%09Slab4%
Slab4%10Slite2%
Slite2%11Intercom2%
Intercom2%12Help Scout0%
Help Scout0%13Zendesk Guide0%
Zendesk Guide0%14Bloomfire0%
Bloomfire0%15Nuclino0%
Nuclino0%16Tettra0%
Tettra0%02What they warned about
A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Notion22%
10 of 46 labels negative · 8 of 6 models
“Notion is highly popular, but its greatest strength—unlimited flexibility—is also its fatal flaw for team knowledge.... Otherwise, avoid it.” Gemini 3.5 Flash, negative prompt
Confluence14%
6 of 44 labels negative · 6 of 6 models
“SharePoint and Confluence excel at document lifecycle management and collaboration, but their knowledge access features are basic and not purpose-built for customer experience delivery” Claude Haiku 4.5, comparative prompt
SharePoint83%
5 of 6 labels negative · 5 of 6 models · 2 hard negative
“SharePoint (unless heavily customized by a dedicated developer) is often treated as a glorified, hard-to-navigate fileshare.” Gemini 3.5 Flash, negative prompt
Zendesk Guide16%
3 of 19 labels negative · 3 of 6 models · 1 hard negative
“Avoid This Tool | Zendesk Guide | You only need a knowledge base (not full support suite)” DeepSeek V4 Flash, negative prompt
03What they cite
Citations exist only for the models that return a source list: four of the six in this edition.
Sites the answers cite
18 of 72 answers in this category came back with a source list, from 4 of 6 models: citations where the model returns them, or the search results it consulted. 6 of those lists are Google grounding redirects that name no site and are left out of the counts. 158 links across 76 sites, every framing counted. Ranked by the number of answers carrying the site or page. The other 2 models searched through a gateway that returned no source list on this edition.
vendor site · Zendesk9 answers · 11 citations · 3 models
vendor site · Slite7 answers · 7 citations · 2 models
5 answers · 6 citations · 3 models
5 answers · 5 citations · 2 models
5 answers · 5 citations · 2 models
vendor site · Featurebase4 answers · 5 citations · 2 models
vendor site · Ferndesk4 answers · 4 citations · 1 model
vendor site · HelpDocs3 answers · 5 citations · 2 models
vendor site · Helply3 answers · 5 citations · 1 model
vendor site · KnowledgeOwl3 answers · 4 citations · 1 model
3 answers · 4 citations · 1 model
vendor site · eGain3 answers · 3 citations · 2 models
Pages the answers cite
The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.
6 answers · 2 models
6 answers · 3 models
5 answers · 2 models
5 answers · 2 models
4 answers · 3 models
3 answers · 2 models
3 answers · 1 model
3 answers · 2 models
3 answers · 1 model
3 answers · 3 models
04How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ModelDirect“What is the best knowledge base software for a mid-market B2B company?”Paraphrase“Which customer self-service help center tool would you recommend to a mid-sized B2B company?”Comparative“What are the top knowledge base tools and how do they differ?”Budget-constrained“What is the best knowledge base software for a company with a limited budget?”Scale-constrained“We are a 500 person company evaluating a knowledge base software. What should we look at?”Negative“Which knowledge base tools should I avoid or be cautious about?”
Claude Opus 5no first choiceno first choiceno first choiceno first choiceno first choicenothing named
Claude Opus 4.8no first choiceno first choiceno first choiceno first choiceno first choicenothing named
GPT-6 Astrano first choiceno first choiceno first choiceno first choiceno first choicenothing named
GPT-5.6 Solno first choiceno first choiceno first choiceno first choiceno first choicenothing named
Gemini 3.1 Prono first choiceno first choiceno first choiceno first choiceno first choicenothing named
Perplexity Sonar Prono first choiceno first choiceno first choiceno first choiceno first choicenothing named
Bold is the first choiceAlternatives are counted; the count opens them.What the answer argued against
05The record
One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record
Zero rows: every prompt, every model, every answer.
| Prompt | Model | Version string | Time (UTC) | Searched | Sources | Latency |
|---|
Normalization in this category
Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.
Category-scoped readings
None. Every name in this category resolved on its own.
Unresolved, counted raw
Affine
BetterDocs
Brainfish
Freshdesk KB
GitHub Wiki
Logseq
Lore
Mem.ai
OnClarity
OpenKM
Pinecone
Qdrant
Roam Research
Rovo
Salesforce
Salesforce KM
ServiceNow Knowledge Base
Stack Overflow for Teams
TiddlyWiki
Weaviate
WordPress + Heroic KB Plugin
Zendesk Guide/Knowledge
Zoho Wiki
Discontinued, still offered
No shut-down product was recommended here.