I’ve been looking for the answer to this for a while and I’m sure the answer is there somewhere and I just haven’t found it, so perhaps the thing to take away from this is that you need to put the answer into a FAQ.
As someone interested in using Temporal at my company and putting a tremendous amount of effort and trust into doing so, I would like to understand how Temporal plans to make money in the future. Everything you are doing is greatly appreciated, but not sustainable in a financial sense. I understand that there are plans to create a Temporal service that will generate revenue, but that doesn’t seem sufficient if you are to get past the “valley of death.”
Then there is the scenario where some well-funded company sees the value in Temporal and buys you outright. That happens all the time and is often the motivation for startups in the first place. What would you have us believe about this?
Ultimately, why should I take the risk to make Temporal a critical component of my infrastructure? I love what you are doing. I trust your motivations, but that and $5 will buy you a cup of coffee.
I appreciate the concerns you have and completely understand why you have them. My answer is pretty direct:
We make money from our cloud service and nothing else. We do not offer paid support, enterprise edition or on-prem deployments.
Temporal Cloud is already generating revenue. We have 10+ paying customers, multiple companies are in production. Amazing companies such as Checkr, Bolt and Descript are already public users of our cloud. Our first cloud customer was a 20B+ publicly traded company.
We cannot guarantee the longevity of the company or anything like that. It’s a crazy world and there is no way to predict the future. As of now there is zero interest in selling our business and that’s quite unlikely to change. Maxim and Samar literally spent the last decade of their lives just trying to solve this problem. How many people do you know who have dedicated ten years of their life to a single thing and then stopped doing it all of a sudden?
It’s important to not underestimate the critical mass behind Temporal. Most people would be shocked if they knew about the number and types of companies already running Temporal for critical use cases. Even if you just look at public users like Stripe, Datadog, Netflix, Coinbase its clear there is serious investment in Temporal. These are all companies who have a strong vested interest in ensuring Temporal stays maintained even if our entire team were to quit today.
Hope that clears things up. Please let me know if you have anymore questions.
We don’t have plans to provide any commercial offering beyond Temporal Cloud.
Our experience shows that in most cases “not all usecases support cloud” is based on a misunderstanding of the Temporal Cloud security and execution model.
Some of the common objections:
We cannot run our code in the cloud
Temporal Cloud hosts only the Temporal Service. Application code runs on customer premises. Customers fully control the deployment and management of the application code. This includes both workflow and activity code.
We cannot trust our data to a third-party service
All payloads sent to Temporal service can be encrypted using customer own algorithm and keys. Temporal cloud never sees any customer data in clear text. Both UI and CLI contain features allowing local decryption of the payloads for troubleshooting. Many customers send only keys to the PII data stored in their internal DB even when encryption is used.
We have legal requirement that our customer data cannot leave their country even if encrypted
Temporal runs in many regions. You pick the region when provisioning a namespace in the cloud.
Our security team would not allow opening connections to our datacenters.
Temporal SDKs require only outgoing connections to Temporal service. So no inbound connections are needed to the customer network. AWS PrivateLink is used by many customers to avoid any traffic over open internet.
It is cheaper to run open source version
Temporal cloud is consumption-based. So you pay a monthly support fee + consumption. For the majority of workloads, the cloud is cheaper or comparable to self-hosted AWS cluster cost. This doesn’t include engineering time needed to maintain OSS cluster.