Testing in no-code?
Testing in no-code seems to mean broadly three things:
1- Can these tools support the desired workflow? Can I do what I want to do?
2- How will the workflow handle errors?
3- How do we provide visibility to folks in our tools that they may be triggering workflows?
#1 - Can I do what I want to do?
When it comes to whether something is possible without code, we have to check whether our glue – Zapier, Integromat etc. – supports the different actions & triggers we need. Zapier probably does the best job of this with their
app directory.
There really is no other way than being intimately familiar with the tools you’re using to know whether a workflow is possible using no-code.
I’d love for an entrepreneur to create a
draw.io type diagram creator that depending on the tools you want to connect and the expected workflow, tells you whether it’s possible. That said, like many of you I’m sure, my workflows are rarely so complex from the get go that I can’t just give it a try.
#2 - How will the workflow handle errors?
Out of the three, I think mainstream no-code tools perform best here. Zapier will automatically notify you when Zap breaks and offers granular settings for when to be notified.
Parabola isn’t great and breaks when any API it’s working with throws an error.
They’re aware that it’s an issue and I’ll definitely keep annoying them until they have better error handling!
Just like #1, handling errors more gracefully is an opportunity for entrepreneurs! Take Chris Spags from
Jetboost which builds add-on features (Boosts) to
Webflow. Their next Boost will replicate something you
could do with Zapier (with a lot of difficulty) but at a much cheaper price point (why pay for Zapier if all you need is one flow?). And his Boost will handle Webflow specific errors gracefully! The effort for Zapier to handle Webflow errors gracefully in every case isn’t important enough for them to prioritize which means an opportunity for folks like Chris.