Allen Holub
@allenholub.bsky.social
Author, international speaker, consultant, software architect, kitchen-sink wrangler.
Many engineers are scared of "sales" and "marketing." Don't be. Talk to friends. Write about your problem and how to solve it.
The smaller the problem, the sooner the revenue, so get crackin'!
7/7
The smaller the problem, the sooner the revenue, so get crackin'!
7/7
November 10, 2025 at 7:00 PM
Many engineers are scared of "sales" and "marketing." Don't be. Talk to friends. Write about your problem and how to solve it.
The smaller the problem, the sooner the revenue, so get crackin'!
7/7
The smaller the problem, the sooner the revenue, so get crackin'!
7/7
Use social media. Go to meetups. Find people who have the same problem that you do. Give your solution away (once you have customers who can give you feedback, and you improve the product based on that feedback, you can think about selling it).
6/7
6/7
November 10, 2025 at 7:00 PM
Use social media. Go to meetups. Find people who have the same problem that you do. Give your solution away (once you have customers who can give you feedback, and you improve the product based on that feedback, you can think about selling it).
6/7
6/7
Make it web-based because that runs everywhere (you do need it to look good and work well on a phone, though). By all means, use AI to help write the code if that speeds you up, but we're talking small, here. An AI assist won't make that much of a difference.
Then get it out there.
5/7
Then get it out there.
5/7
November 10, 2025 at 7:00 PM
Make it web-based because that runs everywhere (you do need it to look good and work well on a phone, though). By all means, use AI to help write the code if that speeds you up, but we're talking small, here. An AI assist won't make that much of a difference.
Then get it out there.
5/7
Then get it out there.
5/7
Then build it in the simplest way possible—no need for elaborate frameworks or a 100K-customer architecture (but you do want an architecture that can grow incrementally over time—take a look at DDD). You probably don't even need a database, at least not at first.
4/7
4/7
November 10, 2025 at 7:00 PM
Then build it in the simplest way possible—no need for elaborate frameworks or a 100K-customer architecture (but you do want an architecture that can grow incrementally over time—take a look at DDD). You probably don't even need a database, at least not at first.
4/7
4/7
You don't even need a business plan. (Remember, no VCs.) Things like "risk" don't matter—you don't have anything better to do with your time if you're not working.
3/7
3/7
November 10, 2025 at 7:00 PM
You don't even need a business plan. (Remember, no VCs.) Things like "risk" don't matter—you don't have anything better to do with your time if you're not working.
3/7
3/7
You do not need AI. (Remember, you're not selling to VCs, so you don't need a buzzword-based product.) You do not need millions in revenue—replacing your salary is enough.
Solve a tiny problem that impacts your actual life. Don't do a "great idea!" You don't want a "blockbuster" product.
2/7
Solve a tiny problem that impacts your actual life. Don't do a "great idea!" You don't want a "blockbuster" product.
2/7
November 10, 2025 at 7:00 PM
You do not need AI. (Remember, you're not selling to VCs, so you don't need a buzzword-based product.) You do not need millions in revenue—replacing your salary is enough.
Solve a tiny problem that impacts your actual life. Don't do a "great idea!" You don't want a "blockbuster" product.
2/7
Solve a tiny problem that impacts your actual life. Don't do a "great idea!" You don't want a "blockbuster" product.
2/7
There is a word for this way of thinking: "science"—a process and way of working that most people do not understand, particularly in the business world.
5/5
5/5
November 9, 2025 at 7:31 PM
There is a word for this way of thinking: "science"—a process and way of working that most people do not understand, particularly in the business world.
5/5
5/5
That said, if we DO drill into the "why," we're likely to learn things that would increase throughput even more, so it's an exercise worth doing.
4/5
4/5
November 9, 2025 at 7:31 PM
That said, if we DO drill into the "why," we're likely to learn things that would increase throughput even more, so it's an exercise worth doing.
4/5
4/5
Admittedly, you can often get away with just looking at the edges. For example, if the claim is that X practice improves throughput when compared to Y, you can prove that simply by measuring throughput without having to drill deeper into the why.
3/5
3/5
November 9, 2025 at 7:31 PM
Admittedly, you can often get away with just looking at the edges. For example, if the claim is that X practice improves throughput when compared to Y, you can prove that simply by measuring throughput without having to drill deeper into the why.
3/5
3/5
That's hard work, so I normally wouldn't do it, but the drilling process increases my understanding of why things work the way they do. That is, the act of proof increases my own knowledge.
Proof is not about convincing, it's about learning.
2/5
Proof is not about convincing, it's about learning.
2/5
November 9, 2025 at 7:31 PM
That's hard work, so I normally wouldn't do it, but the drilling process increases my understanding of why things work the way they do. That is, the act of proof increases my own knowledge.
Proof is not about convincing, it's about learning.
2/5
Proof is not about convincing, it's about learning.
2/5
It's not generic "people." It's Republicans. Trump has the money and refuses to release it. Republicans have tried to kill SNAP for decades, going back to Regan and his bogus "welfare cheats." It's just racism and cruelty all the way down. Trump's got $20B for Argentina, though.
November 9, 2025 at 1:52 AM
It's not generic "people." It's Republicans. Trump has the money and refuses to release it. Republicans have tried to kill SNAP for decades, going back to Regan and his bogus "welfare cheats." It's just racism and cruelty all the way down. Trump's got $20B for Argentina, though.
My travel-for-any-amount-of-time solution is merino wool for everything: shirts, pants, socks, underwear. The stuff doesn't need to be washed unless it actually has dirt on it, and even then, you can wash it in the hotel sink and it's dry the next morning. A carry-on works for all trip lengths.
November 9, 2025 at 1:46 AM
My travel-for-any-amount-of-time solution is merino wool for everything: shirts, pants, socks, underwear. The stuff doesn't need to be washed unless it actually has dirt on it, and even then, you can wash it in the hotel sink and it's dry the next morning. A carry-on works for all trip lengths.
Self-organization is a fundamental precept that is a key indicator of whether the organization is "Agile" or not. The presence of team-level managers is a huge red flag to me.
5/5
5/5
November 6, 2025 at 4:25 PM
Self-organization is a fundamental precept that is a key indicator of whether the organization is "Agile" or not. The presence of team-level managers is a huge red flag to me.
5/5
5/5
The _team_ decides how to work, what to work on, and when to do the work. To say a team is incapable of that is to infantilize the team. This is not kindergarten.
4/5
4/5
November 6, 2025 at 4:25 PM
The _team_ decides how to work, what to work on, and when to do the work. To say a team is incapable of that is to infantilize the team. This is not kindergarten.
4/5
4/5
To me, any place that treats SM as a job title has it wrong on a fundamental level.
SM-as-job-title thinking quickly devolves into the SM-is-a-manager anti-pattern. All "Agile" (including Scrum) teams are self-managing/self-organizing.
3/5
SM-as-job-title thinking quickly devolves into the SM-is-a-manager anti-pattern. All "Agile" (including Scrum) teams are self-managing/self-organizing.
3/5
November 6, 2025 at 4:25 PM
To me, any place that treats SM as a job title has it wrong on a fundamental level.
SM-as-job-title thinking quickly devolves into the SM-is-a-manager anti-pattern. All "Agile" (including Scrum) teams are self-managing/self-organizing.
3/5
SM-as-job-title thinking quickly devolves into the SM-is-a-manager anti-pattern. All "Agile" (including Scrum) teams are self-managing/self-organizing.
3/5
Those "accountabilities" can be concentrated in the hands of a single person, but I've actually never seen a high-functioning Scrum team where the SM was a full-time job. Typically, one of the engineers dedicated some of their time to SM stuff.
2/5
2/5
November 6, 2025 at 4:25 PM
Those "accountabilities" can be concentrated in the hands of a single person, but I've actually never seen a high-functioning Scrum team where the SM was a full-time job. Typically, one of the engineers dedicated some of their time to SM stuff.
2/5
2/5
This reminds me of how much software development is a social activity where rationality doesn't matter. If everybody in the group thinks in one way, anybody who proposes a different way will be ostracized, even if they were formerly part of the group. Nobody wants to lose their friends.
November 6, 2025 at 4:10 PM
This reminds me of how much software development is a social activity where rationality doesn't matter. If everybody in the group thinks in one way, anybody who proposes a different way will be ostracized, even if they were formerly part of the group. Nobody wants to lose their friends.
If you can talk about it, what in particular sent you running?
November 5, 2025 at 7:31 PM
If you can talk about it, what in particular sent you running?
More to the point, you can work much faster when you don't have to wrestle bugs, so high quailty dramatically lowers development time. Worth thinking about. </rant>
9/9
9/9
November 5, 2025 at 6:15 PM
More to the point, you can work much faster when you don't have to wrestle bugs, so high quailty dramatically lowers development time. Worth thinking about. </rant>
9/9
9/9
Nobody in a 996 environment "wastes time" with documentation or writes easy-to-read code.
I'm sorry, but quality matters—particularly if you're a startup, but even when you're not. If your competition is higher quality than your, people will move to the competition.
8/9
I'm sorry, but quality matters—particularly if you're a startup, but even when you're not. If your competition is higher quality than your, people will move to the competition.
8/9
November 5, 2025 at 6:14 PM
Nobody in a 996 environment "wastes time" with documentation or writes easy-to-read code.
I'm sorry, but quality matters—particularly if you're a startup, but even when you're not. If your competition is higher quality than your, people will move to the competition.
8/9
I'm sorry, but quality matters—particularly if you're a startup, but even when you're not. If your competition is higher quality than your, people will move to the competition.
8/9