Small SaaS help desks: chat and email support with help articles, 13 firms
Summary
Customer support conversations from 13 small US SaaS companies, each answered by a team of 2 to 8 people, with the help center articles agents linked in their replies. The mix of chat and email across booking, invoicing, e-commerce and analytics products gives a breadth one vendor's desk cannot. End-user names, emails, phone numbers, order and account numbers are replaced with consistent placeholders.
Collected to order. We collect it from US companies for your order. Scope, volume and timing are confirmed in writing before you sign.
Pricing on request
Request scope and pricingAt a glance
- Planned scope. Volume, sources and timing are confirmed in writing before you sign.
- Records
- 1.8Msupport conversations and help articles
- Size
- 7.4 GB
- Tokens, estimated
- 1.1B
- Years
- 2016 to 2026
- Firms
- 13
- Owner profile
- 13 US SaaS companies with 8 to 40 staff each, selling to small businesses and agencies
- Industry
- Software
- Data type
- Support and service
- Headcount
- 8 to 40 staff per firm
- Region
- United States
- Languages
- English
- Formats
- JSONL
- Updated
Sources
| System | Kind | Records |
|---|---|---|
| Intercom | Support tickets | 980K conversations |
| Help Scout | Support tickets | 512K conversations |
| Freshdesk | Support tickets | 250K tickets |
| Help center exports | Shared files | 9,800 articles |
Good for
- training support bots to answer from a help center and know when to hand off
- evaluating grounded answers against the article the human agent linked
- learning tone and escalation across many small products instead of one
- building intent and routing classifiers for small-business software support
Fields
| Field | Type | Description |
|---|---|---|
| conversation_id | string | De-identified conversation id |
| firm_code | string | Pool member code |
| product_area | string | Product area tagged by the agent |
| channel | enum | chat or email |
| started_at | timestamp | First customer message |
| messages | array<object> | Turns in order with role (customer, agent, bot) and text |
| linked_articles | array<string> | Help article ids sent in replies |
| resolved_by | enum | bot, agent or unresolved |
| csat | enum | good, bad or none |
Example record
{
"conversation_id": "CV-F05-0318840",
"firm_code": "F05",
"product_area": "online booking",
"channel": "chat",
"started_at": "2025-05-02T18:21:47Z",
"messages": [
{
"role": "customer",
"text": "Clients are still getting booked on Friday after I blocked it off yesterday. What am I missing?"
},
{
"role": "agent",
"text": "The time off was added to [PERSON_44]'s calendar, not yours, and bookings check the staff member being booked. I can copy it to your calendar and cancel the two bookings that got through. Want me to go ahead?"
}
],
"linked_articles": [
"KB-F05-0142"
],
"resolved_by": "agent",
"csat": "good"
}Licenses available
Exclusive
One buyer. Nobody else can license it during the exclusive term, usually 2 years.
Non-exclusive
Licensed to more than one buyer.
First look
One buyer gets 2 to 4 weeks to make an offer before anyone else sees the dataset.
What ships with it
- Datasheet: what it is, what it is good for, verified size, sources, years, format and limits.
- Provenance file: the owner shown by code, staff notice and consent status, exclusions, mapped to the EU AI Act Article 53 training-data summary and California AB 2013.
- Manifest with a hash for every file.
- QC report and de-identification report.
- Your own copy, with unique canary records.
How it is de-identified
Default exclusions removed, files flattened to text, identifiers found in three layers and replaced with consistent placeholders, and the result measured against a 99% bar for each identifier type in each source.