What happens when you run out of Automations?

[RnJ]

Because you can’t purchase more Automations on the Pro Plan in the event you run out (like integromat/zapier), what happens to your workflow? Is everything halted?
Upgrading to enterprise would not be an option due to extreme price difference.

For example - Say I have my automations set for my entire workflow working great, but I accidentally trigger an automation that uses a ton of unnecessary automations… oops… my fault and understand and wish I can buy more to pay for that mistake, but I can’t.
Now I am out of automations and almost have an entire month left with zero automations :frowning: and all my workflows are stopped.

I just want to find out the “what if” and hope never end up in that situation :slightly_smiling_face:

Thanks.

[ScottWorld]

This is a great question. I’d like to know what happens in this situation, too.

[Bill_French]

RnJ:

Now I am out of automations and almost have an entire month left with zero automations :frowning: and all my workflows are stopped.

So, this is precisely why there should be no limits on automations for paid users. Airtable should cherish users who avoid API and glu-factory integrations that pummel servers constantly seeking info about record changes and states. Automations are event-driven and because of this, they are very easy on servers - they fire when – AND ONLY WHEN – something actually changed. Compared with polling integrations, this is a minute fraction of the largely present integration model banging on Airtable today.

In fact, Airtable should be paying users to switch to internally-driven event-based automations.

If Airtable wants to encourage the community to continue to build heavy-handed polling architectures for data interchange they will continue to sustain automation caps.

[kuovonne]

I agree that this is a great question. I also agree that it would be great to be able to purchase more automation runs.

However, automations is still relatively new, and I suspect that these things take time to roll out.

@Bill.French, while I agree with you about automations within Airtable being less resource intensive versus the glu-factory integrations, I think that a lot of people are creating automations for things that they were not using glu-factories for, things that were not automated before at all.

Running out of automations is yet another factor that people need to take into account when making decisions about using an automation script versus using script app. There are no run limits on when using script app.

[Bill_French]

kuovonne:

a lot of people are creating automations for things that they were not using glu-factories for, things that were not automated before at all.

Good point. And it’s likely that the vast world of polling processes are not going to be recast with event-driven replacements for some time. But, this is on Airtable; they should have addressed this ages ago. So today …

Should they constrain new innovation by making it guesswork for customers if a new and innovative use of event automation will be (a) sustainable, and (b) affordable?

In my view, the past should not compromise Airtable’s ability to create competitive distance and the customer’s inability to assess or predict whether or not their innovations will fit a given budget.

Another dimension of this is in-platform and out-of-platform calls. Amazon is very clever about data movement across its network boundaries with far more expensive rates when data leaves their data center. They make it almost zero cost to perform compute operations inside the AWS domain. Other PaaS providers (like Airtable) should embrace this approach or force all of us to subsidize the bandwidth costs associated with the few customers who move data into and out of the Airtable realm.

kuovonne:

Running out of automations is yet another factor that people need to take into account when making decisions about using an automation script versus using script app.

Indeed, but this is a metric that few - even experienced developers - can easily predict. As such, it’s like building a leasing model where you are charged based on the number of times the engine’s cylinders actually fire. We don’t do that to consumers because they would never be able to relate to this metric. Instead, we use something relatable - like miles.

[SSA_Admin]

I set up automations to test various options and ran out (free plan) now I know what will be best for me I can’t implement it until the next billing period and I can’t find out my date to know when that is.
I like automations but will be conservative in using them

[Bill_French]

SSA_Admin:

I like automations but will be conservative in using them

Indeed; same here. The outcome from an automation that fails because of a quota breach is no different than a SaaS platform that is down. And there’s no way to anticipate such “outage” and gracefully fail and recover.

[Eriks_Reks1]

A few notes here because this just happened to us.

• We have 15 automations currently on. 2 are extremely active and are likely the culprits, but there’s no way for me to tell you Automation A takes up XX% of the run limit because visibility, as you know, is limited.
• We ran out at the very end of the day. My client emailed me at 11pm EST with the below message.

• Because I’m not an owner of the workspace I don’t have visibility into the little meter that’s only available on the workspace view, per this article.
• I’m also not familiar with how others are using Automations (another one was turned on earlier this week), and so it’s impossible for me to keep track of where we stand in terms of our limit overall.
• No alert came my way with a message like, “Hey, you’re 75% of the way there and there’s 10 days left in the month.”

I agree with @Bill.French’s comment about how Airtable should have a greater vested interest in supporting Airtable customers who are taking their platform to the max and spending hours and hours of time on their site. Automation runs limits should not be fixed to some arbitrary number, but perhaps proportional to the base size, the dashboards being used, APIs coming in and out, etc.

Upgrading to Enterprise for just a few days so we can keep our optimized operation going is a chore and quite a steep jump in terms of pricing.

I’ve reached out to a few folks to see what kind of support we can get given there are 8 days remaining.

[Bill_French]

Eriks_Reks1:

We ran out at the very end of the day. My client emailed me at 11pm EST with the below message.

This is precisely the surprise clients are not expecting.

BTW - Google manages quotas at 24 hour intervals. If you breach on any given day, by the next morning you are operational again and why I typically create self-healing processes. With a daily quota you can become aware of scale and activity in near-real-time even if you aren’t the account owner.

Airtable, pretty much, has this model completely wrong but I must also suggest that a lot of script/action creators have a few things wrong as well.

[Eriks_Reks1]

Well hopefully we can resolve it soon! Big Friday bummer mood.