@garbas I have some time to spend, so I sent a mail to webmaster@nixos.org
Anyway, I thought I’d post the answers to the latter questions here, too.
What is your idea what marketing team will work on, this year? Next year? Next 5 years?
We should start thinking about a rough marketing strategy mostly to be clear on what our goals are. If this is not clearly communicated, I’d be wary of subjective nit-picking in the community around what should and should not be done.
Sure, we want to make Nix more mainstream, but what exactly does that look like, if we are successful? More NixOS users? More Nix code on GitHub? More success stories about Nix adoption in organizations?
Identify a target audience that fits with the marketing strategy, e.g.
- DevOps engineers or site reliability engineers
- System administrators
- Software developers (maybe segregated by tech stack)
As we have a lot of experience in the field ourselves, we can make some assumptions around what the relevant pains and gains of our target audience are to begin with.
However, ideally this should be followed up with user research. For example, we can review research papers as far as available, create online polls or conduct interviews with people from the target audience to validate or falsify these assumptions (outside NixOS community).
In any case, when we identify the issues that Nix can solve, we know what to advertise. This can mean explaining in simple terms that Nix can solve a problem and demonstrate how Nix does this. This should be done on the website, of course, but also by strategically placing conference talks outside the Nix community. There are certainly a lot more avenues for public outreach to explore.
In addition we will likely learn about issues that Nix cannot solve well enough yet. If additional features, documentation, tooling, packages, etc. would help Nix solve these issues, we can then communicate this to the Nix community to consider prioritising their development.
While addressing issues, changing our advertising, we can continuously conduct user research to monitor how Nix is perceived by our target audience.
This would be an on-going process, but I think it is realistic to get quite far already during this year 2020.
Next year we could start to widen the target audience and address a different set of needs. Just to give an example, we may focus on making Nix a great choice for local development environments and building containers now, and continue to widen the scope to managing complex infrastructure entirely with Nix-based tooling.
With the website hopefully in much better shape we can focus our efforts next year more on driving a part of the development efforts on Nix, NixOS and nixpkgs into a direction that is aligned with the marketing strategy.
Five years is far away. But I hope that Nix will be mainstream by then. The marketing team continuously monitors the needs of Nix users, derives suggestions for features and tools for the development from insights gained and actively reaches out to more target audiences to explain them how Nix will solve many of their infrastructure issues.
What could be few easy fixes to the website that would solve problems that you heard others complain about Nix/NixOS?
Without actual research, this will naturally be a little subjective. Generally, I believe https://nixos.org should cater much more to newcomers. Here are a few things that I believe are easy to fix.
Quickstart
Right now, you have to dig a little to find out how to get started with Nix.
There should be a very smooth path for people to get into their first simple nix-shell as soon as possible. This can be done with a few shell commands. Then they could already be writing their first shell.nix to keep that shell around. This is all very easy to do on most Linux distros, but the website does not highlight this. It should be the obvious thing to try first, when you are a new visitor to the website.
My hypothesis here is that, if we can show a potential new user in a few minutes how Nix can help them manage environments, this goes a long way in creating interest and make them stick around for more.
Clean up frontpage
There is too much irrelevant info on the frontpage
- release history beyond the current release of Nix/NixOS
- blog posts are interesting much later, but should not be your first impression. I’d bet experienced users get this from discourse or reddit instead of the homepage.
Confusing navigation
The Home/Nix/NixOS/Nixpkgs split is confusing: focus the navigation around what the user wants to see or do. I would merge all the separate navigation systems into a single menu focussed around the use case. For example, nixpkgs and the package search address the same concern and should be grouped together.
This requires some iteration with user feedback, but the situation can certainly be improved.
Package search
The package search could show more relevant results, i.e. when I type in python or vim or emacs or redis, the respective packages should really show up first. This goes for nix search, too, by the way.
