Am I just overhead tracking overhead?
The Question: Am I actually providing value, or am I just corporate overhead tracking corporate overhead?
What's Going On: We've been building (and rebuilding) our cost model for a few years now, and we still can't get past "I don't trust the data." Every time we hit a data gap, we work around it with a better allocation rule, but by now the model leans on so many assumptions that nobody will make a real decision with it.
What I'm Afraid Of: Leadership just asked every IT team to justify its headcount for next year, and I'm afraid the honest answer is that if my job disappeared tomorrow, the company wouldn't suffer. They'd just have fewer charts.
I'll be honest with you, your fear isn't unreasonable. For larger enterprises, the endless cycle of trying to gain traction is truly costly when you add up the headcount and software. If nobody is making decisions with the model today, leadership is going to ask the same question you're asking. The good news is it's fixable, and I don't think more data cleanup or allocations is what fixes it.
What you're describing is a cycle a lot of programs fall into. Between chasing down data problems and creating allocation logic to minimize them, the more assumptions you add, the less useful the information gets. Even when the assumptions make sense, getting folks to understand them, and when they can use certain reports, makes adoption harder. Ultimately, your team is the only team that trusts the data and can defend the analysis that comes out of it.
I'd stop trying to engineer your way to trust. Keep your assumptions simple enough that anyone can follow them, even if they aren't as precise as they could be right now. Simplicity wins over complex precision. People will argue that statement, but I've seen too many eyes glaze over when explaining super complex allocation rules. When people say it's too simplistic, explain the data gaps and invite the data owners into the conversation. I assume you're already in this cycle with data owners, but try to get them to be more of a shield for you, instead of you being a backlog item for them. How can you get them to defend the data improvement journey for you?
Next, I would narrow the scope. Pick one decision leadership cares deeply about and needs to make within the next quarter, like a contract renewal or a growing cloud or AI bill. Focus on that portion of the model and sharpen it just enough to materially inform the decision. Bring Finance and the person who owns that decision into the work early. People trust numbers they helped build a lot more than numbers handed to them. The data only has to be good enough for that one decision, which is a much lower bar than good enough for everything.
Keep a running log of every decision the model informs, what was decided and what changed because of it. When leadership asks your team to justify its headcount, that log is your answer. It's a lot easier to defend a list of decisions you helped get made than a stack of reports.
If you can get even one real decision made with the model this quarter, you're not overhead tracking overhead anymore. Start there.
Hope that helps!
Have a question of your own? Submit here.
Want next week's answer in your inbox? Subscribe here.