Skip to content

2023-07-03 team meeting

Brent’s update about Mathesar users

Brent did an update on what he’s learned from the Mathesar users he’s been talking to.

Brent’s notes (private): https://hackmd.io/4DId5uYKR1icDXZs8Bq9oQ

Notes

  • Brent has been responsible for responding to installation requests through the website (~15 so far)
  • This means he’s met some potential users
  • There are some things we can learn
  • Notes are here: https://hackmd.io/4DId5uYKR1icDXZs8Bq9oQ, Brent won’t go through all the notes in the call, please read if interested.
  • Users are referred to in general terms (since notes are public)
    • Business in Italy, wants to use Mathesar for ERP/CRM
      • Loves data modeling capabilities, he is already familiar with entities relevant to his business
      • Wanted to do it locally
      • Trying it out before committing to it
      • Stuff he wants:
        • Basic UI stuff like color coding cells, etc.
        • Calculations / formulas
      • Tried and didn’t like NocoDB
      • Understands portability and having a standard format for the data set
      • Didn’t need installation help – needed help with some UI confusion
    • Real estate office under large conglomerate
      • Use case: more like CRM than ERP
      • Currently using Airtable
      • Wants to integrate with mapping
      • Wants to access DB directly, knows SQL
        • Was frustrated with Airtable’s inability to do this
      • Stuck at installation on existing hosting provider
        • Shared hosting
        • Couldn’t figure out SSH access
        • Wants to use cPanel to install software
    • User at large French company
      • SaaS provider
      • Wants to use Mathesar to enable customer support using Postgres
        • Can’t modify schemas
        • Since Mathesar’s install adds schemas, it’s a no-go from the tech team
    • User want to use it for project management
      • Was confused about installing it on a web server
      • Seemed to want a SaaS solution, but didn’t express it in so many words
      • Currently using NocoDB
      • Using Windows
    • A couple of different users want to use Mathesar instead of a Postgres client or code
      • Currently using: pgweb, pgadmin, command line / Python
  • Summary:
    • We’re attracting interest from business users, not tech users
    • We’re not best equipped to help these types of users because of our installation
    • These users are not using Mathesar for critical operations, and we wouldn’t suggest they do, because they’re not technical enough to manage it.
      • We can improve Mathesar for these use cases (e.g. offering hosted Mathesar, or hosted DB)
  • Discussion:
    • It’s interesting that some people want us for our capabilities, not just because we’re a “free Airtable”, it proves some of our hypotheses
    • It’s good to know there are business users interested in data tools
    • Seems like some users are coming from a PoV where running their business on spreadsheets is no longer sustainable
      • Don’t think about things like backups
      • We may need to educate them or make things easier for them
    • People who use shared hosting might be satisfied with a SaaS offering (rather than integrate with things like cPanel)
    • We might need to have an opinionated flow on how to set up Mathesar for specific use cases (have a hosted database, cost $X per month)
    • Ideal customer seems to be customers who need non-technical users to have existing databases (like the SaaS provider)
      • Will pay a bunch of money for Mathesar
    • We should keep track of installation requests and users in a database
      • We don’t really have standardized data for this

What should we be working on in the next two weeks?

  • Wrapping up current work.
  • Cleaning up open PRs.
  • Organizing our GitHub issues and prioritizing ones for the next release.
  • Planning out projects for the next cycle.
  • Discussing goals, priorities, and vision for Mathesar – we should all be on the same page by the time the next cycle ends.

What should our next projects be?

  • Getting user feedback from existing users
  • Should be focused on increasing user adoption
  • At least:
    • Getting a better picture of Mathesar’s users (maybe personas)
    • Having a clear idea what getting to beta and v1.0 looks like
    • Simplifying installation
    • Making data in Mathesar sharable
    • Addressing bugs and features requested by users
    • Designing the first experiments towards sustainability
  • After this cycle, we should aim to have enough information to select the next round of projects.

Projects

  • Kriti & Ghislaine: Strategy / user research
  • Rajat: Localization
  • Brent and Anish: RSQLA1
  • Mukesh: Simplifying installs
  • Pavish: Shareable links
  • Sean & Dom: Improvements for 0.1.3