Sector · SaaS

The support queue grows with your users. The team does not.

Onboarding, API keys, integrations, billing. The same questions arrive every week, and the answer is usually in documentation you have already written. MotherX answers from your own docs, help articles and release notes, in the language the user writes in.

The problem

Nobody reads the documentation. They write to you instead.

The documentation is usually good. The problem is that the user is mid-task with an error on screen and two minutes to spare, and will not search a docs portal. She opens the support form. Then the team spends the afternoon pasting links to pages that already existed.

  • Answers taken from your own docs and help articles
  • Answers in the language the user writes in
  • Hands over to support when the question needs a person
User
why am I getting a 401 from the API?
MotherX
A 401 means the key was not accepted. The most common cause is sending it in the wrong header. The key belongs in Authorization as a Bearer token. Source: API reference, Authentication.
User
and if we just rotated the key?
MotherX
Then the old one stops working immediately, so it has to be removed from the environment variables in every environment. Source: Key management, point 4.
Recurring questions

What fills the support queue.

Eight categories account for most of the volume in typical SaaS support, and every one of them is documented somewhere already.

🚀

Getting started

First setup, inviting users and what has to be in place before anything works.

🔑

API and authentication

Keys, tokens, scopes and why the call returns 401 or 403.

🔌

Integrations

What connects to what, and which fields sync in which direction.

🐛

Error messages

What the code means, and what the user should do next.

🧾

Billing and licences

What is counted, when it is counted, and what happens on a mid-period upgrade.

👥

Users and permissions

Roles, who can see what, and how access is revoked.

📤

Data export

Which formats exist, and how to take the data with you.

🚪

Cancellation

Deadlines, what happens to the data, and how long it stays available.

How it works

From docs to first answer the same day.

01

Connect the documentation

Point at the docs portal or sitemap. The AI reads the pages and builds a searchable model of them.

02

Add what is not on the web

Internal guides, PDFs and help articles only support has had access to.

03

Put it in the product

A small script in the app or on the docs pages, so the answer is where the user is stuck.

04

Read what they ask

Every question is logged, so you see which parts of the documentation fall short.

Insight

See which parts of the documentation fall short.

A topic that triggers many questions is usually a docs page missing a step. The dashboard ranks them, so you know which article to rewrite first instead of guessing from whichever tickets someone happens to remember.

  • Questions ranked by topic and volume
  • Questions that got no good answer
  • Evidence for which articles to rewrite
Most asked · last 30 days
setting up SSO146
401 on API calls118
changing plan mid-period no answer74
exporting data to CSV52
FAQ

Questions from SaaS teams.

Can it replace first line support?

It takes the repetitive part, meaning the questions that are already documented. Anything that needs an account check, judgement or a decision is handed over. The point is to clear lookups out of the queue, not to remove people from support.

Can it read our documentation?

Yes. Sitemaps and page links are described as sources, along with PDF, Word, Excel and PowerPoint. If the documentation sits behind a login, access has to be sorted out first.

Does it answer in English?

Yes. The product pages list Norwegian and more than 90 other languages, and it normally answers in the language the user writes in, regardless of the language of the source.

Does it keep up when we ship new versions?

Sources can be set up with scheduled crawling, so updated docs are picked up on their own. How often depends on the plan, so it should be confirmed specifically before you build it into a routine.

What happens to what users write in the conversation?

Conversations can be logged in the dashboard and deleted. If you turn on lead capture, the chat can ask for name and email, and you configure the purpose, consent and storage yourselves. Customer data is not used to train public foundation models.

30 days free

Let the documentation answer for itself.

Connect the docs and help articles, and see what the queue is actually made of. No card required. Setup included.

Try for free