I’ll be at Dreamforce next month. Moscone, three days, the usual.
I’m not going just for the keynotes. I am excited about them
I’m going for one kind of conversation. The kind you can only have in person, with someone who owns support or operations and has had a quarter they still can’t fully explain. I have a short list of questions for those people. Same questions, everyone I meet.
Here they are.
How did you find out?
Not when the alert fired. Alerts fire all day. Your dashboards were green the morning it started. The deploy went in clean. Every check passed.
So how did anyone know something was wrong?
Usually the answer is a customer. Or a pattern someone noticed by hand. Rarely the monitoring you pay for. The monitoring did its job. It watched the things it was told to watch and it reported them accurately. It just had nothing to say about the one that mattered.
I want to know how the news actually reached you.
Was it the second time?
Not the second time you fixed it. The second time it happened.
Most teams can’t answer that with much confidence. An incident in March and an incident in July can be the same underlying thing, and nobody says so, because saying so takes one person holding both in their head at once and seeing it.
So I ask it as a question about recognition. If it came back, would your team clock it as a repeat while it was happening? Or would it look brand new, get worked brand new, and close brand new?
What did the change look like on the day it shipped?
The change that caused your last bad week. Go back to the day it went in.
It passed review. It passed the checks. It looked like every other change that week. Nothing on it, that day, said “this is the one.” If something had, you’d have caught it.
That’s the part I find most interesting. Not the failure. The ordinariness of the thing that caused it, right up until it stopped being ordinary.
Maybe it was a deploy. Maybe a config change in your Salesforce org. Maybe a contractor touched an integration and the only trace is a ServiceNow ticket nobody’s opened since. Walk me through it. What shipped, what it touched, whether anyone would have flagged it if they’d known to look. Usually the honest answer is no. Nobody could have. Nothing about it, on the day, carried the information you’d have needed.
Three questions. Notice what none of them ask.
None of them ask what your monitoring caught. None of them ask how fast you resolved it.
They ask how you knew, whether you’d know again, and what the cause looked like before it was the cause. Different questions. The tools you already own answer the first kind well. This kind, they mostly can’t.
That’s the conversation I’m looking for at Dreamforce. If you own support or operations at a company running Salesforce, and you’ve had the quarter I’m describing, I’d like to compare notes. Reply here or send me a message and we’ll find time on the floor.
If your team recognizes the pattern and you’d rather not wait for the conference to dig into it, we run a complimentary Investigation Cost Audit. Forty-five minutes. A structured look across five dimensions of how your team gets from a problem surfacing to understanding what caused it. You leave with a scorecard. Numbers on what last quarter actually ran you: in time, in repeat work, in the hours your most experienced people carried alone.