Ask an ops leader to walk you through their stack. They’ll do it without notes.
Salesforce for the cases. Jira for the work. GitHub for the code. ServiceNow for the changes. A monitoring platform or two. Every one of them is named, licensed, owned, and renewed.
Then ask a different question. When a case spike needs to be matched to a change, where does that happen?
The answer is never on the list.
It’s a spreadsheet, a shared doc, or a channel one person reads end to end so the two ends can hear each other.
Walk into any operationally serious team and you’ll find one. They built it because they saw the problem clearly: the systems each hold a piece, and nothing holds the whole. So somebody made a place where the pieces meet.
That artifact is doing the most important job in the operation. The licensed systems record. The spreadsheet connects.
Now compare what they’re each given to work with.
The licensed systems have backups. Permission models. Audit history. A vendor on the other end of a support contract. An owner, a budget line, a renewal date somebody argues about every year.
The spreadsheet has a person.
It gets updated when that person has time. It’s current until the week it isn’t, and it looks exactly the same either way. It has no test environment, no second maintainer, no version history worth the name. It appears in no architecture review, because officially it isn’t architecture.
Unofficially, it’s what holds the architecture together.
I want to be precise here, because this is easy to hear as criticism. It isn’t.
Building the workaround is THE sophisticated move. It means the team diagnosed a structural problem correctly and closed it with what they had. The teams without a spreadsheet aren’t ahead of this one; most of them just haven’t seen the problem yet.
The risk isn’t that the workaround is dumb. The risk is where it sits. The single point where your systems meet is also the single artifact with no support model, no successor, and no line in anyone’s budget.
Take the person who maintains it away for two weeks. The systems keep recording. The connection stops being made. The operation finds out what the spreadsheet was holding up.
Every stack has one system that would never pass the review the rest of the stack went through. It’s usually the one doing the connecting.
That seems worth saying out loud before Dreamforce, where the conversation will be about everything except this.
Teams running the connection between their systems on an artifact with one maintainer and no support contract are exactly who we’re thinking about.
If that’s your team, we offer a complimentary Investigation Cost Audit. Forty-five minutes. Structured diagnostic across five dimensions. You leave with a scorecard that quantifies what the investigation is actually costing you: in time, in recurrence, in the load your spreadsheet was never rated to carry.