BUG: Users not reaching 'tag goal' in automation when tagged. They just stay stuck in the automation.
I needed to remove all of my users from an automation, so I created a tag goal at the very bottom of the automation (Goal: Tag Applied "Newsletter - Remove From Sending Sequence"). Underneath that goal, I added a 'remove tag' action removing "Newsletter - Send", and applied the goal tag to every one of my contacts.
All of the users in that automation should have hit the tag goal, and had "Newsletter - Send" removed from their tags.
27,000 contacts received the tag, made it to the goal, and had the "Newsletter - Send" tag removed.
592 contacts received the tag, did NOT make it to the goal, and did NOT have the tag "Newsletter Send" removed. They are still simply still sitting in the automation, with no reason for the goal to have failed. They have the tag, they just weren't moved to the goal.
There are no errors in the Fluent logs for these users. They all received the tag, and are in an active state. They simply didn't move to the goal. There are no errors in my server logs at all... pertaining to server resource constraints, or PHP.
Being able to move users to a tag goal is one of the single most important functions of an automation... and it just blatantly doesn't work.
This is on top of a host of other inexplicable automation failures that I've experienced...
Please help me to understand why this is happening...
Support ticket 133643 also addresses this issue.
Hi Travis, have you had any luck resolving this? I'm considering moving from ActiveCampaign but I have some significant concerns about automations and issues like this make me nervous. What other "inexplicable automation failures" have you experienced? I'd love to be better informed about those things if you're open to sharing - even if you want to chat privately that's fine. DM me / email israel@israelsmith.com. Thanks!
Hey Israel,
After 10 years with Active Campaign and paying $1,500/month for enough time to put my kids through college, I would 100% switch to FluentCRM again.
I did manage to figure the issue out, and it was inherently a mix of user error and a misunderstanding of exactly how goals in FluentCRM work.
In AC, when you set a goal, regardless of where the user is in a campaign, when the user achieves that goal, they will zip right to the goal right away. It's like a teleportation device.
That's not how it works in FluentCRM.
#1. If the user is in a 'wait' period, like waiting 2 days for whatever the next action is, if they achieve the goal, they will have to wait until the wait period is done before they're moved to the goal. That goal action is not triggered until their existing action has finished. It's a delayed teleportation.
This is annoying because if you want a user to be able to sit at a certain spot in an automation until they click something... it doesn't work. You can't put an indefinite wait period because they will wait indefinitely before the goal is triggered.
As a workaround, I tried to just put an 'end automation for this user' action instead of a wait... so they'd just sort of sit at a dead end until the goal was achieved. "No further campaign actions will happen. But, that doesn't work. Users won't be moved to a goal within the automation if it has been ended for them. Again, pretty annoying. You can drop a completely new user IN to the campaign at a goal... but a user who is sitting in a campaign after an 'end' will not achieve that goal. And they're never actually 'out' of the automation. You can't remove users.
Now this does NOT apply to automation trigger goals... so if you START an automation with a tag goal, that can be achieved no matter what. They pop right to the top of the automation when the trigger goal is achieved... AFTER their wait period ends, assuming they're in a campaign.
So, for more complicated campaigns where goals truly need to be infallible, you might break up the automation into multiple chunks, where the trigger goal serves as a more effective teleportation device.
What I'm probably going to end up doing is just attacking it from the database level. So, I will completely remove the user from the automation with a db query triggered by a webhook... so, "remove user_id 22222 from campaign ID 35... and then I will trigger the tag goal which allows them to teleport to the goal, right away... regardless of prior campaign actions or waits.
The long and short of it is that every system has it's own flaws and limitations. Fluent is no different... it just takes longer to dial it in because you're going to need to make server adjustments to go along with it. PHP timeout values need to be dialed in, memory constraints, email rates, etc. But once it's up an running... it's a major relief to not be paying for active campaign any more.
The more you know about databases and chatGPT script prompts, the more flexibility you have. I can do things that I could never do with AC, and i'm not limited by their IP reputation.
You just have to be willing to get your hands dirty, or hire some who can do it for you.
The support is thorough, but you will not get 'emergency' response time. Any ticket you have, you can expect 1 good answer on it to come in overnight. If you respond in the middle of the night, there's a chance you can get more communication.
I hope this helps!
Travis Roesler wow man that’s an amazing answer. Thanks so much! Good to know about server scripts and dialling in the values etc.
What kind of hosting are you running on? And is FCRM running within the same Wordpress instance as your main website and shopping cart? Any notable performance hiccups? Thanks again!!
Israel Smith I’m on an m7a.xlarge EC2 instance.
I have my woo commerce shop on a subdomain, but I run checkouts and sales funnels on my main site with fluent.
I have had some performance issues, but these again pertain to understanding how fluent works. I had a heavy email sending automation which would basically sort users based on what ‘group’ tag they had, and then apply different wait times etc so my newsletter would be partitioned out over time.
But when you add 30-50k people to an automation that immediately does 5 db lookups per user and then tags the user 15 different ways, and then sends 5k of them an email… adding users to that automation will be slow, and in some cases timed out for me.
I learned to add 10 minute ‘waits’ to the beginning of these automations, so users would be firmly added before all of the nonsense ran… and I cleaned up a bunch of that nonsense.
Fluent needs to program for the least common denominator. Not everyone has a dedicated EC2 instance, so they limit the number of people that can be processed, so as to not crash your server. This can make complex processes take longer.
I have managed to bypass many of these checks, so I can harness the power of my server, but no matter what, you do learn how to be ‘cleaner’ with your tagging and funnel steps.
It’s another case of, “this was broken… but not actually broken… I just needed to learn how to work within the system.”
Travis Roesler brilliant reply as been seeing the same issue and was wondering what what going on. Totally agree re the switch to Fluent after using keap/infusionsoft for nearly 15 years (along with small fortune)
Just a note re hosting we had all sorts of issues with hosting security product Immunify. So if you host has this and can't or wont disable then move!! Have now moved our fluentcrm to Ultahost VPS and is working brilliantly.. not competing with any of our other sites when doing large contact automations etc.
We have just the fluentCRM installed and use WPfusion from other sites to integrate.
bren Is immunity a 3rd party web application firewall? The lack of flexibility for stuff like that can be rough, especially when you’re dealing with rate limiting, among other random blockages.
I do use cloudflare to filter out the basic attacks, but I self host Atomic Protector and it’s great. Any time I have an issue with an application being blocked, they patch the software immediately. It’s incredible. But I can see 403’s in my own logs… that’s crucial for being able to figure out what’s happening.
I haven’t updated my woo commerce site in 7 years… no hacks. But yeah… security is a completely different dance with a self hosted CRM.
Like I said, the journey hasn’t been ‘immediate’ for a lack of a better word, but having made it… I’m super happy with where I’m at.
Travis Roesler Immunify360 is a product that some web hosting companies have on their servers.. full details here https://imunify360.com/
very annoying as it was Killing the php process on certain FluentCRM tasks. Mostly POST requests to site. There may have be some iffy or malformed posts but this killing of the PHP process was causing massive "pauses" and time outs. Switched to host who doesn't have this.. 100% uptime and no processes killed... side effect we had on shared hosting is that when the php got killed on fluentcrm it affected other 3 sites on same account... real pain to debug.
Brutal. I feel your pain. I remember when I was using Sucuri WAF, and for whatever reason, it didn't like the length of the URL parameters on my UPSELL links. So, people were clicking links to buy stuff and just getting slammed by a firewall.
It's always something... right? This is why when you finally achieve stability, just don't touch anything! ;-)