A rational CapEx purchase and tracking process for IT

How often does someone in your company (often the CIO, or the CTO, or the head of infrastructure) end up running through the halls, waving a purchase order that “has” to be signed off that very day, or else key systems will allegedly go dark? Maybe you’re in the fortunate situation of being in a company where this frenzy doesn’t happen, but in my experience, that’s unusual.

I’ve written before on the importance of technology carefully shepherding its fiduciary responsibilities. Nothing contributes to the IT stereotype/stigma as much as a loud demand for a major purchase, at the last minute, justified solely by dire predictions of doom, and topped (often) with acronym-laden technobabble. Amazingly, it’s not that hard to avoid this situation, if you exercise a little forethought and planning.  The benefits of doing so are indirect as well as direct: you can change perceptions of IT into being viewed as a partner of business concerns, rather than as a troublesome, risk-fraught, and confusing cost center.

It all goes back to Management 101: plan the work, then work the plan. Surprises are a bad thing. Not only do you need a solid plan, but then you want to diligently track actuals against that plan. None of this is exactly a radical idea, yet I’ve served as an executive now in at least three different companies where none of it was happening before I arrived, with respect to capital expenditures.  To the extent there was a capital expenditure plan for the year at all (as opposed to just one big CapEx number!), it had been thrown out the window by February. Sure, this can and does happen in fast-paced Internet companies in particular, but the rankling thing was that no one really was tracking the changes against plan, or could envision how funds were shaping up for the year. Even if a plan has undergone radical changes, there still needs to be a current plan. Walking in, any executive (not to mention any auditor!) should be able to see one or two core documents that detail that current plan, as well as the progress against it.  If that’s not there, then the technology area (and by extension the whole company) is just operating by the seat of its collective pants, and that’s not acceptable.

Here are the minimum elements of responsible CapEx stewardship, in my view:

• Construct, as part of the annual budgeting process, the CapEx plan for the year, project by project, with anticipated months for each expenditure.  When the budget is finally approved, this CapEx plan becomes the Plan of Record.
• In addition to the expense grid, compose an initial outlined description/justification (at least a paragraph, definitely not just a vague phrase) for each project, assigning that project some kind of arbitrary ID for tracking.  This description needs to be signed off by the finance department for Sarbanes/Oxley purposes, and to ensure that the appropriate corporate guidelines for capitalization are being adhered to.  Note that often, a given project will end up having its cost split over several expenditures (e.g., multiple hardware purchases made in different months, so assigning each project an arbitrary but immutable ID is useful to help track back to the original plan.
• As a project gets actually kicked off during the year, there needs to be a deeper analysis and planning phase, during which the expected costs and timing get solidified and, of course, written down.  Here, it’s critical to document the alternatives to the proposed CapEx expenditure, including the risks and cost of staying with the status quo rather than going with the proposed approach.
• A regular process needs to be instituted and expected by all: a monthly formal review with the CFO (or the CFO’s delegate), using the following tracking document: an itemized list of expenditures planned for last month, expenditures actually incurred, expenditures deferred, expenditures cancelled, expenditures upcoming for this month.
• The purchase order mentioned at the beginning of this post? That PO should come to the signing executive with predictable standard attachments: the analysis document for the project, the CapEx plan of record, and the current monthly CFO review showing the expected expense. What, this is an emergency and we don’t have those? OK, but emergencies had better be rare and explainable as exceptions.  In any case, the PO should appear at least a week before it “has” to be signed. No surprises.

Side notes:

• The analysis document for each project? That’s an excellent place to make sure that all the costs are taken into account, including non-capitalizable elements such as maintenance.  Again, “no surprises.”
• Any CFO or CEO who is being presented the justification for a technology project that lists and analyzes only one proposed solution has a right to be intensely skeptical: astoundingly, that single-minded approach seems to be typical at many companies.
• Here’s a key and often misunderstood basic tenet: just because a given expense is “in the budget” doesn’t mean that the amount is right or that the purchase should be made at all.  A budget is simply the bare bones of a plan, made with limited information at the time; for nearly all projects, a detailed investigation with multiple vendors is necessary to get true pricing and timing. The annual budget, your initial “plan of record”, should be regarded as a mere best-faith estimate.

If I were walking into a company, as a new CIO, CTO, or CFO, I’d immediately want to see the current CapEx plan, and the solid record of monthly reviews occurring as plans gel and change.  What, that isn’t all there? That tells me a lot.

And if you’re feeling that doing all this is a lot of work?  Well, if you want to be able to spend the money, you should be more than willing to put up with a certain amount of due process in order to get it.