I'm working with some awesome engineers who have day jobs but still find time to write some posts. How do we make sure we make best use of their time?
Do "How we did X to improve Y" posts work? Do you want backgrounders to our key area or should we assume knowledge and give a bucketload of detail? I'm assuming nobody wants promotional content, but if we say how awesome we think something is because we solved our biggest pain point, does that count?
- Don't form voting cabals on hacker news / reddit. There's a lot of sophistication in detecting it. Let the natural popularity of a post be an indicator of what the audience likes. I'd further be paranoid if these sites detect a voting cabal, it could hurt the posts from the domain (to be clear, I don't know this for certain).
- Make it casual, interesting, and authentic.
- Don't put too many barriers between the author's authentic techie voice and the reader
- Find a way to give your great writers more autonomy on the blog. For example, instead of an upfront review process, give them permission to directly publish, but not share for 24 hours to allow a quick review for legal/PR reasons
- Give authors clear, consistent guidelines with what they can feel free to talk about / what they shouldn't talk about in public
- Create support for your newer writers. They're the ones who need encouragement and a writing coach. But don't take away their authentic voices, help them discover them.
- Don't get too cute with protecting IP. The in-the-weeds tech details are almost never the thing that makes a great organization or product.
- Understand the goals of your tech blog. Is it for recruiting? To get consulting clients?
- Focus on the long-game, strong community engagement, relationships, and domain authority, not short-term conversions (https://moz.com/learn/seo/domain-authority)
Ken Shirriff started blogging about very average Arduino projects that he did but he taught himself how microchips work and he shares it with the rest of us.
There is a mental model for blogging that goes: "we'll submit this page to Hacker News, get 15,000 views; 12,000 of those will bounce right away; 30 will sign up for an e-mail newsletter; 3 will talk to a salesperson; 0.3 will start a trial; 0.03 will become profitable customers"
If you think that way you won't get the most out of blogging!
I'm sure every day somebody visits Ken's blog for the first time and spends the whole day reading it.
All of the content you describe is good. You need a lot of it. It has to be high quality. Some articles will be better than others and sometimes you'll struggle to get articles out the door. It will take a long time to accumulate enough articles to make a destination.
Be careful with "calls to action". On one hand you want people to take action. I've been told to add them to blog posts and really wanted to because I was just as hungry as my co-workers. Sometimes I feel like they drained all the authenticity out of the article though and were a subtraction instead of an addition.
Make it all about you and your reader. Don't let on that you are blogging because somebody else blogs, told you to blog, etc. Make it clear that you always respect yourself and the reader (e.g. watch self-deprecating remarks like "this blog is occasionally interesting".) There are many signs that make bad blogs seem inauthentic and you don't want to be that person.
1. Patterns in confusing explanations - https://jvns.ca/blog/confusing-explanations/
2. Blog about what you've struggled with - http://jvns.ca/blog/2021/05/24/blog-about-what-you-ve-strugg...
3. "advice" for aspiring tech bloggers - http://jvns.ca/blog/2016/05/22/how-do-you-write-blog-posts/
Mention versions and dependencies where possible/practical.
Show me the code!
Point to relevant source material.
Lots of code examples and lots of simple explanations of complex ideas.
The tone is casual, like a conversation, rather than a lecture.
I try and replicate that a bit (but not nearly as well) on my own blog (sambernheim.com). The best way to get better is to keep practicing.
There are few things I hate more than posts where the author provides some half-assed instructions, maybe even omits very important things, and then when you try to re-create, it just fails to build halfway through.
- this is how we did what we did
- this is why you need to think about this thing, and here's a way (or ways) to think about it
- I got really bored, and did this thing - have at, Intarwebs
- here's something that I tried that didn't work, and I wish I knew why, or here is why (at least in this use case)