Have you ever used low-code automation tools?

Hi! I would appreciate if you could share your opining on low-code automation tools. Have they really helped you?

Moderator’s Edit: For context, Tania works for aqua ALM (a testing tool vendor).

2 Likes

I’m going to make the massive presumption that you’re talking about low-code test automation tools.

They have not helped. It was a while ago, but what I found was the more effort I saved the further it put me from my desires. The tool puts heavy limitations on what is possible and shapes the way that you work - you end up working for the tool, rather than the tool working for you. I spent more time fixing the steps than I would if I’d just coded in an established framework. I know what I want to check when I’m testing and I want my tools to help achieve that, not to proscribe for me how and what to test, or put limitations on both my actions and my thinking because it excludes those possibilities. Working in automation is already limiting to testing, and then abstracting that out with recorders and the like feels like a limitation on top of a limitation until I feel like I’m too corseted to do anything valuable with the time I’m supposed to be saving.

If I were going to try this road again then what I would settle for, and may even exist, would be a way to generate useful data in an automatic way that could save some time when it gets to the real coding. Maybe a way to generate primitive page objects, parse field names, that kind of thing. Low-code tools powering high-code tools, all powered by tester decisions.

There may have been huge leaps in the low-code test automation tool world that I’m oblivious to, though, who knows.

5 Likes

Hi @taniazhydkovaa

Yes, I have and they have been successful for SaaS applications as elaborated here

They are not tools to solve everything, but they do have a purpose depending on the situation.
I have a company white paper, and blogposts on the topic as I have been working with these tools the last 5 years.

Currently there seems to be a new tool pitching every so often. My analysis would be that the market for these tools are already overloaded with options and promises.

2 Likes

I have found that low code solutions can work very well in certain circumstances. For example, helping to speed up repetitive task that aren’t particularly complex.

For automated testing I have stayed away from them for multiple reasons including, wanting the developers to contribute and feel like tests are part of their workflow. Using a test framework that is a library in their codebase, makes much more sense.

Recently though I have experienced an exception to this - An ERP system that rhymes with “tap”. Our existing automation tools could help but it was hard work and just didn’t fit well with the skills for those working on this system.
We trialled several ‘low-code’ and ‘no-code’ solutions and they are not all equal. Most fell into 1 of two categories: Too expensive or Too opinionated about how automation should work. Eventually we found a tool that was reasonably priced and allows us to follow all of the importance principles and practices we apply with our code projects. It’s early days but so far it has been very successful.

So to conclude I would say that where you have software engineers crafting code - find a code solution for testing that they are happy with. But if your challenge sits out side of that context, low code might well be worth a look, but be prepared for a lot of up front tools evaluations and comparisons.

3 Likes

I used them way back.

I viewed them as temporary scripts, 30 mins to create running daily covering basic user flows, often throwing them away in about two weeks and another 30 mins to create new one for next couple of weeks.

For me, good value when I treated them as quick to create, temporary and throw away after short life span perspective.

I hear they are more robust now and easier to have a slightly longer life.

Broader coverage of fragmentation risk is also a potential plus but at the same time coding standards have improved so ideally fragmentation risk has also dropped even though we still have sky rocketing variations.

There is a separate angle for me though, I love discovery focused testing and this was mainly scripting so a bit less fun, I also did not need to learn coding which was part of the interesting bit of code based tools so also lose out on that aspect.

The phrase “we do not do these things because they are easy” springs to mind when I see the adverts for the tools, what do I learn as a tester if a lot of my time ends up on a codeless tool?

3 Likes

Nice one, @adam.pike. Thanks for sharing.

This reminds me of the excellent thread on how testers evaluate a new tool. It’s full of such useful advice.

You can try it :slight_smile:

2 Likes

Hi Adam, would you be happy to share which tool it is that you are now using and happy with?

I have tried many tools like this.
The ones that I really appreciated are Testim and Functionize.

You can also have a look at TestRigor.

It also depends on your needs.

1 Like

I have used Zapier and IFTTT, both are awesome :slight_smile:

2 Likes

Sorry to be so late to the party here but yes, low code tools are HEAVEN in some cases. I’m a consultant and I often have to go to a client and setup an automation framework. But I also have to share the knowledge towards the internal people. Which are often not technical nor do they know how to program, so we’ll sit together and discuss all the options, if they wish to learn it or not.

Most of the time a low-code tool is the best option for the company due to time/money/investments and most of all people. Many do not have the time nor do they want to invest time into learning how to program (they’ll always say ‘I WOULD LOVE TO LEARN IT’). In reality it’s different, so a low-code tool fits their needs better and they actually enjoy it since it has a lower learning curve and it’s always better to work with something you enjoy then you hate.

So yea, I’ve worked with both options and low-code tools are nice at some point.

Developers who moved away from developing and gone to testing will probably always prefer a full coding framework like RestSharp or RestAssured. But that doesn’t mean it’s the best fit for your company/project. Make sure to assess it carefully and hey, maybe even the ‘Business’ jumps in and uses a tool like SoapUI or Postman.

Hey @kristof ! No worries, it was really helpful, thanks!

1 Like

Hey - I have used TOSCA and Eggplant. Both proved to be efficient in creating scripts quickly and for running less number of scripts. But with increasing number of scripts and complexity it was difficult to manage. Recent postman flows is a good tool but its been marketed as a development tool rather than testing.

2 Likes

Kinofrost summarized it well. I too dislike low code tools so much that I’d suggest job seekers to avoid companies which use such tools or leave the company if they adopt them.

If you’d like, here is one of my posts on why I dislike such tools - link. You might want to look at the 2nd link in the post I shared.

2 Likes

@sivaganesh.sivakumar Wow, thank you, 2 tools :fire:

2 Likes

Thank you @anon68517856 ! Yes, it was definitely insightful to find out the opposite point of view

1 Like

I was working for a company where we had QA Engineers doing automation and QA Analysts doing manual testing. QAAs once decided that our automation is going too slow so they said they will try out some low-code/no-code tool.

They didn’t ask us (QAE) for opinion and they didn’t make some research / pilot, they just went head-in with then-popular Katalon Studio. It was missing an important feature that we needed (can’t remember now) so they had to ditch it, and then switched to Testim. They were happy at first (few weeks) but then the questions started to roll in… how to do this, how to do that (mostly about JavaScript). We had a complicated web app that needed a lot of custom scripting which Testim also offered, but that required more and more effort and knowledge of code.

The project infamously got shelved after some time. Granted, this was 4-5 years ago, I think current tools have advanced a lot, especially the flurry around AI. Found this video just yesterday:
10 AI-Powered Software Testing Tools You Have to Know

2 Likes

I used the Eggplant and SikuliX before (kinda automate image capture based script).

In my personal opinion, I think low-code automation should be tools act as supplement when you want to something repetitive quick and can be throw away. It shouldn’t be the main or only one tool to use.

  • Your reliable reference source going to be documentation and request to support and quite rare to find community to ask. (If it a paid tools)
  • When some testcase start to get long, complex and flaky. I felt it a bit annoyed when I found there more factors that make test flaky (e.g. same image run in different VM and image can’t be found, Image appear too quick, Handle same multiple image and tell the script which one to click etc.)
  • When your tests scale up (If it a paid tools) It kinda felt like you tied/force to use their service to scale up and have few option to customise.
3 Likes

One of the benefits of low-code automation tools is that they can help streamline the development process, reduce the time and effort required to build software applications, and make it easier for non-technical users to participate in the development process. With low-code automation tools, businesses can build applications more quickly and cost-effectively, allowing them to be more agile and responsive to changing market demands.

In addition, low-code or codeless automation tools can help address the shortage of skilled developers by allowing non-technical users to build and maintain applications. This can be particularly beneficial for smaller organizations that may not have the resources to hire a dedicated development team.

That being said, low-code automation tools do have some limitations. While they can be used to build simple applications quickly, they may not be suitable for more complex applications that require custom coding and more advanced technical skills. Additionally, the visual, drag-and-drop interface may not provide the same level of control and flexibility as traditional coding methods.

Overall, low-code automation tools can be a useful tool for businesses looking to build applications quickly and cost-effectively, particularly for simpler applications. However, it’s important to carefully evaluate the capabilities and limitations of these tools to ensure that they are the right fit for your specific needs.

1 Like

I have used Ghost Inspector and found that it helped me learn a lot about test automation: Use Low Code as a springboard for learning – TestAndAnalysis

2 Likes