Notes on developing a marketing team manifesto

In the last few weeks, @garbas, @arsleust, and @fricklerhandwerk met multiple times to write down a strategy and roadmap for the Nix marketing team, based on prior discussion with people involved in the past months (@bjth @djacu @idabzo @jeremiahs @kranzes @pxc @ron @tomberek and others).

This is an account of the thought process that resulted in a mission and roadmap for the Nix marketing team.

Why Growth

Growth is instrumental to something. What is that?
Why do you want growth (personal reasons)?

  • @arsleust:
    • more up-to-date packages in Nixpkgs
    • more contributors
    • better quality of software
    • more users of Nix → more users of your software if it leverages Nix
  • @garbas:
    • more companies using Nix
      • companies can supply the resources to make all of the above happen
    • create a sustainable ecosystem
    • if I had to do programming, rather that way
      • don’t feel smart enough to use other tools
    • more effective development leads to better quality, things will just work
    • scalable growth is better than convincing people individually
      • more Nix users are more people to collaborate on software in a sane way
  • @fricklerhandwerk:
    • all of the above, because:
    • I don’t want to care anymore about computer problems that should have never existed

Apparently we have different kinds of Nix users we are talking about:

  • user
    • we want them to be able to use (our) software without friction
      • it should be easy to try or use software without even thinking about it
  • developers
    • the people we want to collaborate with on our creative work
    • we want to share our work with them
  • contributors
    • people who will solve problems we will never manage to solve
      • people who make small improvements, such as “bumping versions”
      • those who are invested in the project and will do amazing things
      • maintainers who will do work literally no one else will do
    • they will help making the ecosystem work at all, and let everyone else get on with life
  • maintainers
    • people who can help contributors getting their work merged and curate the direction of the project

We care about all of them, because there is a progression from user to contributor.
But also, we cannot have users without contributors, because the system has to work to begin with.
Right now we care mostly about users and to a lesser degree about developers, because we assume that funneling people into becoming Nix users will necessarily capture developers’ attention, and eventually those may become contributors.
This provides a natural boundary to the responsibilites of the documentation team that mostly focuses on developers and contributors.

Marketing cares also about “pre-users”, because the question is how people become users to begin with:

  • people who know about Nix and have never tried it
  • we’ve all been ther: what is that, why should I care?

Now we roughly know why we are here. We want the ecosystem to work, and we know we need more users, developers, and contributors.

Objective: What can we do, what should we do to achieve that general goal?

We have to start with where we are. @garbas provides the following estimates:

  • 5k contributors to Nixpkgs (15k GitHub stars)
    • assumption: 1:5 ratio of contributors to users
    • guess that 4x that many people are using packages from Nixpkgs
    • conclusion: 20k user-developers of Nixpkgs
  • assumption: 1:5 ratio of early adopters
    • conclusion: 100k people know about Nix(pkgs)
  • we also care about Nix contributors
    • much lower number, in the hundreds

Objective: Measure actual user base, and figure out the fraction of developers (and ideally the number of potential contributors), and what they need:

  • need to know what is the right thing to work on

  • measure our effectiveness

  • track where in the technology adoption life cycle we are

    • @garbas’ intuition: currently crossing the “chasm”
  • @garbas: conclusion: widen the funnel to grow more contributors and maintainers

    • @fricklerhandwerk: objection: may grow existing users just as well
    • if we succeed, the ratio will actually grow because we get into more late adopters who we can expect to be less likely to become highly involved
      • we probably want to onboard them as contributors/maintainers as well, but that may need a different approach
        • @fricklerhandwerk: have to keep an eye on the maintainers:contributors ratio and workflows not to overwhelm maintenance

We currently have certain things in terms of metrics, but they are dispersed:

  • we probably want to have it all in one place, it would be very important
    • e.g. funnel the numbers into Grafana
  • needs continuous monitoring
  • to some extent blocked on the infra team

Anything more important than metrics?

  • figure out the messaging and adapt the website?
    • based on target audience, can’t serve everyone
    • @fricklerhandwerk: isn’t this blocked on the rewrite?
      • @garbas: no, messaging gets more bang for the buck
  • outreach on social media, newsletter
    • “maintenance” tasks, current main activity
    • probably first priority

How will we know how to make decisions in a changing situation (high growth) if we don’t have metrics?

  • @garbas: we will know from social media

Discussing the next community survey:

  • there is a bias: most “engaged” people are “loudest”
  • main outcome from the surveys are trends
    • those will only be useful in 5 years, but we need to do it now
  • still, last year’s survey was useful per se
  • we need to listen to users, but also go in the direction “we” want users to be

The pitch: “NixOS is the best Linux developer OS out there”

Mapping out objectives for the next three months:

  • a year is too much

  • what capacity the team will have in the next quarter? what can actually be done?

    • media outreach
      • posts for social media
      • revive the newsletter
        • monthly
        • 2h/month
        • all the content is more or less prepared
          • aggregation of what was posted on social medias
      • continue “what’s new with nix?”
        • one video per ~3 months
          • goal: release one video
        • effort estimate:
          • script: 2h for the team (~3 people)
          • recordings:
            • intro/outro: 2h (~3 people)
            • presentations (~10 topics, 1min/topic): 2h (2 people)
          • post-processing: 2h
          • total ~18-20h; ~9-10 h for two people
            • ~1h/w for two people
        • currently involved:
      • (optional) write a content policy
        • what to include, what not to
        • which langauge to use
        • define audience personae
        • estimate: 2h for full team (5 people, two meetings)
          • total: 10h
      • total estimate: 30h (outreach) + 20h (videos, mostly done) + 10h (content policy) = 60h
    • get the survey out
      • goal: publish in June
      • blockers:
        • get back the limesurvey EC2 instance
          • fallback: google forms
      • effort estimate:
        • 2d create questionnaire
        • 8h review and discussion
        • 1h write and publish announcement
        • (1-2h outreach, may be merged into outreach activities)
        • 2x8h statistics evaluation
        • 2-3d qualitative analysis
        • total: ~55h (3-5 people)
      • currently involved:
    • stretch goal: prepare discussion on messaging
      • idea: write down decisions to guide future contributions
  • current team capacity: 5 people, ~2h/w total (120h)

    • @garbas has ~4h/w for marketing activities
      • 1h/w marketing meetings
    • total effort estimate ~115h

After compiling this overview, @garbas and @fricklerhandwerk followed up with finding concise wording that capture these thoughts in order to put on the team page on nixos.org: add marketing team mission and strategy by fricklerhandwerk · Pull Request #1069 · NixOS/nixos-homepage · GitHub.

7 Likes

A colleague at work has done a fair run down on nix blockers to adoption. I think that analysis is a really valuable artifact for your work and might help in establishing a baseline of blockers, even if they’re more or less well known, the compact and rather complete nature of the analysis has value.

I’d wish that we can arrange for that information to reach you. As it’s a thorough (I mean really thorough) critique, it may not get permission to publish here (it may be inciting).

I’ll try to reach out both ways.

4 Likes

Something that’s worth noting, I think, is that the benefits of growth basically all come from growing the number of contributors, rather than the number of users. So when pursuing growth, that’s the metric we should be focusing on.

8 Likes

I argue the real bottleneck are maintainers, and the work they do is usually so boring or tedious or just so much that it has to be paid for to get done at all. But that’s, in my view, accounted for in the discussion. @garbas argues we first need users for them to become contributors and eventually maintainers.

2 Likes

Yes, definitely. But my point is that not all populations of potential users are as likely to become contributors, so if we’re honest that the growth that is actually useful to us is growth in contributors, then we should be focusing on the populations of potential users that are most likely to end up being contributors.

4 Likes

One different way of looking at it (and I think it’s somewhat covered in the discussion):

Users transition into commitment, commitment transitions into budgetary defendability, budgetary defendability transitions into the funding that @fricklerhandwerk alludes to.

That’s probably one aspect of what “scalable growth” means, there are (positive) network effects that we don’t have yet access to (e.g. still limited funding, etc).

1 Like

Thank you for putting this together!
A general ask from folks is that there are many opportunities to talk, teach, educate and expose Nix, as we are a global community do reach out if you recognize anything we should be involved in as we would love to help!
We’ve also been working on getting us tables and presence at some of the many open source related events out there - examples: SCALE/SOOCON/Devoxx(just now with Rok!)
The second point to further emphasize - it’s quite critical for us to understand the graduation between the stages for individuals in our community so that we can make sure to also empower folks to grow into the later ones (contribution/support/helping others). How does one go from discovering Nix → Start using Nix : defining separate types of users which can be based on usage/type of usage/persona → Helping others with Nix → Contributing to Nix.

3 Likes

I agree with both sides here, but generally think we should be thoughtful in our approach. It would be quite ideal if we had some assumptions on what type of persona would be most likely to graduate from user to contributor/maintainer, then potentially focus on those core areas.
The other point is in my general message below, in putting some thoughts together on how we can help folks graduate from active user to the later stages (again contributor/maintainer)
@fricklerhandwerk do you feel the survey we are working on has the right questions to help guide us on this?

2 Likes

This would be super helpful!

1 Like

Re survey: I recently discovered a nicely instrumentable typeform alternative that could be used as un-obstrusive situational market research in the different contices (plural of context?) in which users, contributors and in general members of the community experience their moments of truth:

A bit of a lead question example: :grinning_face_with_smiling_eyes:

  • After a stack trace
  • Would you like to help us improve? - Click this shortlink:
    • There: “How useful did you find the stacktrace (1-5)?”
    • Followed by a Net Promoter Score (NPS), i.e. How likely are you going to recommend Nix to a friend or colleague? (1-5)

But don’t get fooled on “lead question”, it’s still a raw moment of truth capture, even if we probably already can infer the tendency of the answers.

1 Like

Fully agreed. I suppose that should be covered by the communication strategy to be developed next quarter.

That’s another thing, as for defending budgets we have to cater more to organisational users. We talked about that at some length in the developer workshop (I will publish a report soon™), and it needs an entirely different approach than just attracting whatever type of individual users.

Yes, the documentation team will hopefully be able both support with getting those insights as well as making that graduation happen.

I added some suggestions based from my usability studies from last year.

3 Likes

Just a minor clarification:

That sounds true, but is probably omitting the most important point: let’s remember, that it’s a buying center.

I was referring to the user category in that buying center that is indistinguishable from the other individual user.

1 Like