For anyone that has brought in automation to solve a problem - Has it worked consistently? How do you manage it?
Automation is table stakes. What's the context?
What areas are you looking to automate? Sales, Marketing, SDR, Deal composer? Thousands of automation options available.
Great points, Jerome and Pat — context is everything. We've seen automation succeed when it's tied to a clear business outcome, not just 'let's automate everything.' Vern, if you're looking at sales/deal automation, we've done this recently with Salesforce CPQ + DocuSign for a trade association — automated quote generation, approvals, and e-signatures. Cut turnaround time significantly. If it's more infrastructure/DevOps — we handle that too (Terraform, Azure, CI/CD). Happy to share what worked (and what didn't) depending on your focus.
I think the answer depends a lot on what kind of automation you're referring to. Are you thinking about internal workflows, engineering, sales, customer support, or something else?
The automations that hold up for me are the ones with a clear owner and an easy way to see what failed. I’d start narrow, review exceptions weekly, and track whether the output is still helping the person who uses it—not just whether the workflow ran. Once nobody is checking edge cases, small errors tend to turn into a trust problem pretty quickly. What kind of automation are you trying to keep consistent?
I'm not looking to automate. I'm looking to get a feel/learn from peoples/teams experiences with adopting automation. Be it a SaaS product or AI. The trend I'm seeing is that automation tools that are brought in have not delivered as expected. Either you have to constantly manage the workflow process or it delivers what is required for a month and then stops.
Vern H. You definitely need to manage and observe the automation workflows in the beginning. But as it matures, you can even automate the alerts on the workflow. I understand your pain that it delivers for a short period before stopping. But for me personally, that has happened when I was relying too much on third party services. I believe that workflow automations in general work reliably when all the components driving it are your own. But even that is use case specific.
Vern H. The way I see it is that the first quesiton anyone should ask before jumping to the automation is "what's the size of the problem?". It appears that people want to automate sth that won't save time, increase revenue, reduce cost, whatever. Once the problem space is clear, I'd recommend to have very custom automations. Because you use specific set of tools, need specific system prompt, have specific business context, etc. and most automations product out there are not the right fit for that So in short: yeah automations work a ton. But not everything should be automated in a generic way
Consistency usually breaks at the edge cases (bad API responses, hallucinated outputs, or missing variables). For key workflows, we've found that adding a strict validation/QA step before the final execution or dispatch makes all the difference in keeping it hands-off.
