Do you hate writing tests?

I think it depends on which tests we're talking about. The biggest amount of effort is spent on the unit tests in any software project (if there's any!). And ideally unit tests need to be updated with every single line of code that's being added, again ideally even before you write the actual code if you're following TDD. So it's simply not possible to outsource unit tests as they're a part of the project itself and every software developer has to be able to write them along with the actual live code.

If we're talking about other kinds of tests, like integration, stress and load tests, it might be possible to outsource them but because the biggest amount of effort is spent in unit tests, they're kind of secondary and wouldn't be quite beneficial to outsource them in my opinion.

Im using Django for backend development and React for the forntend. Every time when I add a new endpoint in the server I also add unit test cases for it, so I know exectly return codes, data and error messages. It makes frontend development much faster.

Honestly, I see a tremendous value with writing unit tests. When we update an existing feature or add a new core feature, chances are that things are going to break. Unit tests help me identify those right away rather than in production.
That way I make sure that broken code is not deployed anywhere and customers won't call me at night even when they are paying me.

Pros of outsourcing writing unit test-

1. It will save (very little) time in the short run.

Cons-

1. I find it very difficult to trust outside my team with the source code. Managing expectations is going to be difficult.
2. There's going to be a recurring steep learning curve for the outsourced test team based on how complex the logic is.
3. Additional test environment to test and simulate your exact use case. Will users pay for the infrastructure or the developer time?
4. Quality control.
5. Deadlines and syncing up with developer sprint cadence can be tricky. Can be a blocker to push new features.

As an indie-hacker, tests rarely make sense. The product is changing so fast, that keeping the tests up to date and relevant would require more work than actually building the product. Plus, with a small team things are less likely to break as everyone knows the codebase really well.

Storybook + Chromatic allowed me to test my (vue) frontend and actually increase my development velocity. That's what testing should be like

IMO for a one person team, unit testing should be used for critical aspects of your app (e.g. payments). You don't want your customers losing money for no reason.

But other than very critical aspects, I don't really think its necessary and can get in the way of developing other aspects of your app that rely too heavily on the tests to pass.

With that said, unit testing is a good way to jog your memory as well (you can write unit tests as a way to walk through how you built specific aspects of your app and iterate on it).

It's not clear what exactly you offer.
Automated tests?
Unit tests?
Manual testing?
How does it work? How much does it cost?

The problem is that in order for someone else to write my tests, I need to communicate to them what the app's desired behavior is. And I hate communication even more..

Hi,
A few thoughts form me, apologies if some have been mentioned before.

1. Spelling and grammar. The page needs a proof read.
2. The design: Personally I don't understand the characters and their purpose. I feel you could use an illustration that summarises what you are doing and the process. Perhaps a flowchart of UI designs with red areas needing attention (just off the top of my head).
3. Do you have any case studies whereby you could mention some stats. Number of bugs found. Number of potential lost users saved. Flex some numbers, we need to see social proof ideally with some testimonials with headshots.
4. Be clearer what will happen when I click your CTA. I get a box to enter a link. What then? What will happen? Folks don't like leaping into the unknown. a 3-step illustrated process here would work well.
5. What does it cost?

Hope that helps for now.
Steve

One of my developer mentors once told me a feature is 40% feature specific code and 60% tests. I believe that premise to be true. Tests is a must have. I'm not a stranger to outsourcing it though.

I wouldn’t use it because my mentality is coders should write their own unit tests (so even having someone else on your team do it is not good let alone another company)

A slight pivot would be a testing company to find new bugs via manual testing or review your testing code to look for gaps.

My feedback on the site is there is nothing there to make me think you grok testing. I would expect to see someone like “I have been coding for x years and written tests in Jasmine, Selenium, Mocha and NUnit. I have refactored large test suites and made them twice as fast” or similar (obviously based on your experience doesn’t have to be exactly this) with some social proof (maybe colleagues?) . Also double check all your grammar!

Tests bring peace of mind. You will sleep much better a year from now, and you will thank yourself someday.

I can't say I write many tests during indie hacking. At work, however, it is heavily implemented. I think once you grow to a certain amount of users and the complexity of your application it is inevitable to start using tests.

I actually think that for MVP building they are not a MUST and in some cases contribute to the overpolishing and overengineering that some of us fall victim to. I say, if you have a very complex piece of code that is business critical test it from the start otherwise i think you will be fine.

Great Idea but I thought the tests were code tests like unit tests haha!

For our MVP, we didn’t write any tests. Once we start scaling, tests will be very important and we would love to use your service. Look for tech startups that are funded but early in their scale process.

This is a great idea!

Do not trust anything you read about tests online. Its like googling "should I exercise?" "Should I eat healthy?"
Online you will read the things people say they do, while in the real world you will see the things people actually do.

I hate them really..

If you're biasing the poll by only providing negative options, you're setting yourself up for confirmation bias.

I've been doing Test-Driven-Development for 15 years. It's a major (major) time saver if you think long-term. Tests allow you to refactor, which is invaluable as your code base evolves over time. They also allow onboarding new team members more easily, because they can feel free to refactor without silently breaking things.

Tests should not be an after throught, or outsourced. You can outsource acceptance testing, but as soon as a bug was found, you should write an automated test that reproduces it, then fix it.

There are are some good articles on the advantages of TDD:

• https://www.geeksforgeeks.org/advantages-and-disadvantages-of-test-driven-development-tdd/
• https://fortegrp.com/test-driven-development-benefits/

@slavaGanzin given the comments I read here, I thought that maybe it would be a good idea to explain that your tests (correct me if I'm wrong) will be more on an e2e level. Because you probably will not outsource unit tests, right? It would require access to the codebase, no? Otherwise, automated e2e tests seem a loveable idea. I would love to test it.

I like the path you are following with the UI. On the UX however, I think you need to improve feedback on clicked buttons. I clicked twice on the button that appears on helpmetest/step-2 and it sent me two e-mails. Gotta make it disabled and show a loading.

Despite this and things I saw other people saying here already, I think this idea might have a market fit! It's very common in my experience to see developers that don't like to test. But we all know that we need it, especially when our paying customers hit the door. (Oh man, I understand this frustration 🥲)

This is a good landing page @slavaGanzin, aesthetic for sure.

Some grammatical nits (lmao forgive me, couldn't help it):

• "Never Lose Money to a BUG" (not glitch)
• "Sometimes your app DOES not work as expected"
• "And you may NOT even know about it"

---

I don't hate tests, but for my side project, I only write unit tests for a small part of business-critical parts of the codebase.

Once I reach ~10k ARR I think I'll write a couple of E2E tests too (or maybe use your service 😉). Before such a point, as other IHs have mentioned, the time investment isn't really justifiable.

I never did that in my previous five failed projects but the current one I do it a lot. a) it's because my recent contract job I need to write a lot of tests and I get now why it's important b) I done with my MVP so every time I ship new features to the existing MVP I added a new test case.

Just ship the code without tests if you're building it from scratch. Once you hit the MVP mark then write some tests to ensure the intended functionality works.

There's an excellent lecture where Uncle Bob explains why creating a project just to add tests is doomed to fail:

https://youtu.be/BSaAMQVq01E?t=4542

In the same video, a little bit later, he explains the best thing about TDD is that it makes tests less boring.

https://youtu.be/BSaAMQVq01E?t=5759

Overall, all your videos are super interesting and I definitely recommend watching them.

I hate writing tests but not sure its so easy to outsource since it really requires both an understanding of the code and the business (to know what to cover by tests). Feels like writing tests goes hand in hand with building the actual features.

Hi!

From my experience, I've never seen test outsourcing,
Usually writing tests depend on the company stage but personally, I would write tests when I have actual users and in critical places in the application, and usually, it is not that hard and doesn't take too much time for me at least.

I wonder if I would like to outsource Tests it could take much more time from me rather than just writing my own test. 🤔

I think the idea has potential, if I am building a website with login, subscription, search etc ... These general functionalities are known and relatively easy to test, and above all testing is a must.
As for more sophisticated systems, I believe the best tester is the engineer who built it, or at least a testing team that is very close to the engineering team.
Having said that, the idea is plausible and needs a try.

hello @slavaGanzin, i am actually working on creating similar app, i am at a phase of collecting workforce, please let me know if you want to collaborate, thanks