Weāve accumulated a lot of feedback about Automations in Fibery and thinking about full re-write of this area. It is not an easy project. Here are some things we want to hear from you that may help us to make a decision:
What use cases you imagined with Fibery, but canāt do them because of poor automations?
Do you compensate that with n8n, API, Make, Zapier or other external tools?
What a single thing more important for you to have in Fibery instead of better automations? Please do not provide many, just one
Hello,
The use case I can think of is āimproved workflowsā as discussed in this thread:
I think a lot of companies have complex workflows which mix automations, processes and manual procedures (tasks). Itās difficult to document the processes well and to mix both worlds (automation with human in the loop).
The most difficult thing is to follow each ongoing process execution. (e.g. a specific customer onboarding, a project lifecycle, a marketing campaign).
The alternatives are:
Domain locked workflow solutions. A lot of Saas products propose workflow solutions
And well.. N8N which is great. But it is not really āhuman in the loopā friendly
In the end, itās important for a company to track and improve the efficiency of its processes. Itās also important to quickly onboard on those.
EDIT:
Maybe materializing Workflows, allowing to manage a set of rules and their correlations. Properly handling the ātime / transition / contractā aspect of automation.
For an example. A task status can switch to āDoneā or āCanceledā only when it is not āDraftā. And depending of the ācompletionā or ācancellationā event. It would trigger other rules, like Creating a āReportā or a āTestā. To ācancelā a task, maybe it should be mandatory to fill a āformā with a proper reason.
And you could see all this logic on a proper Workflow page.
Before responding, curious what you mean by āfull re-writeā, and what (from the feedback collected) justifies re-writting instead of adding missing features. What features would a re-write unlock?
I personally think automations in Fibery are already much more powerful than most competing platforms. (The āStep 1 Linked Entityā is amazing for complex automations).
Are there things that could be improved? Yes. Looping, conditions, using step results are examples. But the workarounds needed (at least for me) have not yet justified spending the time and energy into a full re-write of automations. Again, Iām curious to understand what features canāt be added without a full rewrite.
Iām still incredibly new to Fibery, but as far as I can tell, the ability to templatize a project plan seems limited. If I could allow my users to easily build their own templates that capture the multi-level Milestones, Taks, and Subtasks needed to complete the project, it would be game-changing.
Our current approach is pretty time consuming and fragile due to how we have to map/look up field values using the native rule builder. I havenāt attempted to build this process with Zapier (which I use for some other automations), but I imagine it would still be really time intensive to set up and manage our templates.
As a simple enhancement, itād be a great start if I could just duplicate an action in the rule builder.
You can cascade this logic to as many levels as youād like. The āWhen linked to entityā trigger is so so powerful, especialy when combined with lookups. Do pay attention to the correction in the comments of the video to reduce the number of automation runs.
Webhook Action in Automations would be my top request for automations. I can handle scripting, but I often just want the more quickly understandable layout of a GUI around webhooks (both during setup and reviewing later).
Working with both collection fields and non collection fields. (perhaps this is more of a formulas item, but this is usually the āsimpleā thing that I find myself wishing I could quickly do in automations)
No
More robust communications within Fibery. For example, publishing something like the Threads feature (I think I put all my thoughts on this in other posts).
What use cases you imagined with Fibery, but canāt do them because of poor automations?
Address entities not linked to trigger entity ā This is a BIG restriction. Example: I want to record a transaction for a specific Contact. From that Contact entity, I want to create a Journal entry. That Journal entry should always have at least 2 individual Line Item entities. Because those line items are not directly linked to the Contact entity, I canāt create them with a button on the Contact. Instead, I have to do one of 2 things: create an unnecessary relation between line items and Contact (this really confuses and complicates the data model) or try to 2-step the automation: create the Journal entry and then have a rule that creates the Line Items (which is much more fragile than a button where I can specific what exactly needs to be done in the context of the button).
Loops ā e.g., I need to create an entity that pairs a single person with a single project for each new project. I know I can do it with scripts, but a āfor eachā automation is actually a pretty common use case. Scripts are a lot harder and so automations just end up not happening.
Variables ā this is one of the biggest things I miss from Notionās automations. Being able to introduce a variable and then operate on it is very helpful in calculations, dynamic automations, etc. Scripts compensates for this, but there are lot of things I just put off automating or suffer with manual entry because scripts are a quantum leap in complexity and I donāt have the time to sit and write them.
Conditional / case PER STEP ā e.g., suppose I have an automation for a Project. If itās a certain type of project, I want it to avoid creating one task or populate a given field differently. Right now, I have to do one of: a) script it (ugh - I just donāt do it then), b) multi-stage it with cascading rules (fragile because they are linked by conditions instead of directly ā and the logic might fail unexpectedly), or c) set up a different automation for EACH different case (complicated and becomes a behemoth to maintain).
For me the need is possibility to combine formula and manual input. The idea is that formula outputs value and I can modify it. When using formula only this is simple blackbox. Ability to alternate value is something I miss.
The reason behind it is that for me Fibery has the potential to be the best email client with relations and the whole system.
When I will be able to alter formula I can send emails.
