4 ms·

I have built a couple very successful SaaS businesses from scratch. 8 figure businesses. I have never in my life written a unit test or any kind of testing code

I have built a couple very successful SaaS businesses from scratch. 8 figure businesses. I have never in my life written a unit test or any kind of testing code in 20 years of professional software development.

jollofricepeas 5y ago
Over the last 15 years, I built a unicorn and sold it for more than a billion and half. We wrote unit tests extensively.
- Invictus0 5y ago
  Clearly one of you must be wrong or lying
  - ohmanjjj 5y ago
    Maybe if I had written unit tests my companies would also sell for over a billion
    - chris_wot 5y ago
      Why not sell the unit tests separately from the business?
    - holler 5y ago
      It sounds like you did quite well! I'd be curious to learn more about your experience, would you be open to a chat on sqwok.im? I'm somewhere in the middle with moderate unit tests and e2e for critical paths.
  - CyanBird 5y ago
    Or both could be right
    - jollofricepeas 5y ago
      Both are likely ;) The idea that you need to throw unit testing out to be successful is kind of silly. When you’re growing a very complex system (huge micro services pattern and dealing with sensitive data) you’ll be happy you have tests to save your ass. …speaking from experience.
  - necovek 5y ago
    Clearly the value of business being sold is not dependent on the application of standardized programming approaches to quality. I mean, who would even think it is? There are also alternative approaches to achieving quality, and there was that report/"study" circulating around that claimed Team Software Process actually beats most agile/extreme programming methodologies in at least one metric (or two: defect rate and time to completion, IIRC). Of course, none of the actual "agile" development teams do clear-cut TDD or Extreme Programming either, so you are never comparing apples-to-apples. As a corollary, I'd also say that neither does the UI or UX matter that much to success, and there are simply a bunch of intangibles that can make or break a product. But one "tangible" thing is that you've got to be serving some customers' needs, and need to continue to do that. And rare are businesses which have a need of "my button needs to be rounded and 10px away from the next button" or "my code needs to be unit tested": those are actually "my customer can get their work done through my product" and "it should rarely break". None of this means that you should not invest in pretty/functional UI/UX or well-tested code. Just don't expect that to mean much in the grand scheme of things. They will affect your ability to hire top-level people once you are an established business, though. (I personally am too attached to "high-quality" [as in well tested] code and would struggle to kick off a start-up for that reason only, even though ideas keep popping up :)
    - jollofricepeas 5y ago
      Exactly this. Product/market fit has more to do with functionality and the minimum quality design aesthetic required by the customer. In our case, we disrupted a rather boring, traditional field. Our UI/UX was never great neither did we hire “frontend” people until post-acquisition but performance and data security were huge to our clients thus testing.

jollofricepeas 5y ago

Over the last 15 years, I built a unicorn and sold it for more than a billion and half. We wrote unit tests extensively.

Invictus0 5y ago
Clearly one of you must be wrong or lying
- ohmanjjj 5y ago
  Maybe if I had written unit tests my companies would also sell for over a billion
  - chris_wot 5y ago
    Why not sell the unit tests separately from the business?
  - holler 5y ago
    It sounds like you did quite well! I'd be curious to learn more about your experience, would you be open to a chat on sqwok.im? I'm somewhere in the middle with moderate unit tests and e2e for critical paths.
- CyanBird 5y ago
  Or both could be right
  - jollofricepeas 5y ago
    Both are likely ;) The idea that you need to throw unit testing out to be successful is kind of silly. When you’re growing a very complex system (huge micro services pattern and dealing with sensitive data) you’ll be happy you have tests to save your ass. …speaking from experience.
- necovek 5y ago
  Clearly the value of business being sold is not dependent on the application of standardized programming approaches to quality. I mean, who would even think it is? There are also alternative approaches to achieving quality, and there was that report/"study" circulating around that claimed Team Software Process actually beats most agile/extreme programming methodologies in at least one metric (or two: defect rate and time to completion, IIRC). Of course, none of the actual "agile" development teams do clear-cut TDD or Extreme Programming either, so you are never comparing apples-to-apples. As a corollary, I'd also say that neither does the UI or UX matter that much to success, and there are simply a bunch of intangibles that can make or break a product. But one "tangible" thing is that you've got to be serving some customers' needs, and need to continue to do that. And rare are businesses which have a need of "my button needs to be rounded and 10px away from the next button" or "my code needs to be unit tested": those are actually "my customer can get their work done through my product" and "it should rarely break". None of this means that you should not invest in pretty/functional UI/UX or well-tested code. Just don't expect that to mean much in the grand scheme of things. They will affect your ability to hire top-level people once you are an established business, though. (I personally am too attached to "high-quality" [as in well tested] code and would struggle to kick off a start-up for that reason only, even though ideas keep popping up :)
  - jollofricepeas 5y ago
    Exactly this. Product/market fit has more to do with functionality and the minimum quality design aesthetic required by the customer. In our case, we disrupted a rather boring, traditional field. Our UI/UX was never great neither did we hire “frontend” people until post-acquisition but performance and data security were huge to our clients thus testing.