Why Your Soft Close Isn't Actually Saving You Any Time

I spent three years running a soft close that I was convinced saved my team time. We'd knock out a preliminary set of numbers by business day two, get them in front of the VP of Finance, and feel good about ourselves. Then the hard close would roll around a week later and we'd re-examine every accrual, re-pull the same bank activity, and discover that half the estimates we booked needed real numbers anyway. We weren't closing faster. We were closing twice.

That's the trap. A soft close is supposed to give you a directionally accurate P&L quickly so leadership can make decisions before the full close is done. It is not supposed to be a dress rehearsal that you throw away and redo from scratch. If your soft close and your hard close don't share any work, you've built a reporting treadmill, not an acceleration.

Where the Duplication Actually Happens

The waste is rarely in the big, obvious entries. It's in the dozens of small judgment calls that get made twice because nobody wrote down the first decision in a reusable way.

The recurring accrual you re-estimate every month

Take a professional services accrual. During the soft close, someone estimates $42,000 for outside legal based on last month's run rate. During the hard close, the invoices arrive totaling $38,500, and someone reverses the estimate and books the actual. Fine. But next month, the person doing the soft close estimates from scratch again, not knowing that legal spend has a predictable pattern: it spikes in quarters with board meetings and dips otherwise.

The information to make a better estimate existed. It just lived in someone's head, or in a reconciliation tab that got overwritten. The soft close estimate and the hard close actual were never connected, so the variance never taught anyone anything.

The reconciliation you do, then undo, then redo

I've watched teams reconcile a prepaid schedule during the soft close, flag it as "preliminary," then reconcile it again at hard close because they didn't trust the preliminary version. If your soft close reconciliations aren't trusted enough to carry forward, they're not reconciliations. They're guesses with a spreadsheet attached.

The Fix: Decide What Is Provisional and What Is Final

The single most useful thing I ever did was categorize every close task into three buckets before the period even started:

• Final at soft close — items that will not change between soft and hard close. Depreciation on fixed assets already in service. Fully invoiced vendor costs received before soft close cutoff. Payroll that's already run. These get done once and locked.
• Estimated at soft close, trued up at hard close — accruals waiting on invoices, revenue pending final usage data, bonus accruals. For these, the soft close number is a real estimate backed by a documented method, not a placeholder.
• Hard close only — items with no useful early estimate, like a one-time legal settlement being negotiated. Don't pretend to close these early.

In a typical month, I found roughly 60% of our entries were truly final at soft close, 30% were estimate-and-true-up, and 10% were hard-close-only. Once we stopped re-touching that 60%, our hard close dropped from six business days to four.

Make the estimate method the asset, not the number

For the estimate-and-true-up bucket, the method matters more than any single month's figure. If your legal accrual method is "trailing three-month average adjusted for known board activity," then next month's soft close estimate is a 10-minute calculation, not a 45-minute investigation. The person doing it inherits the logic instead of reinventing it.

Why Excel Quietly Sabotages This

Everything I described is theoretically possible in Excel. In practice, it falls apart for reasons anyone who's maintained a close workbook will recognize.

When your soft close lives in a tab and your hard close lives in another tab or a separate file, there is no enforced link between the estimate and the actual. Someone copies the soft close file, renames it Close_Final_v3_USE_THIS.xlsx, and the variance analysis between estimate and actual evaporates. You can't learn from a history you overwrote.

There's also no memory of why something was marked final. When a staff accountant leaves, the reasoning behind which items were locked at soft close walks out the door. The next person, unsure, reconciles everything twice to be safe. That's exactly the duplication you were trying to kill.

A structured close platform like ComplyBar changes the economics because the task and its status are the same object. When you mark the fixed-asset depreciation entry as final at soft close, it stays final through hard close unless someone deliberately reopens it, with a logged reason. The estimate you booked for the legal accrual is attached to the actual that replaces it, so the variance is captured automatically. Over three months you accumulate a real record: this accrual was estimated at $42K, $39K, $45K and actualized at $38.5K, $40K, $44K. That history makes next month's estimate sharper and defensible to your auditors.

The handoff problem solves itself

When close tasks carry their status, owner, and supporting method in one place, the soft-to-hard transition stops being a scramble. Your team sees exactly which of the 120 close tasks are locked and which 36 still need truing up. Nobody re-reconciles a prepaid schedule out of anxiety, because the platform shows it was completed and signed off at soft close with the preparer's name on it.

A Concrete Target to Aim For

Here's the test I use to know whether a soft close is actually saving time: measure how many soft close entries survive untouched into the hard close.

If fewer than half your entries carry forward, your soft close is a rehearsal and you should either tighten your cutoffs or stop pretending you're closing early. If 60% or more carry forward, you've built something real. The hard close becomes a focused exercise in truing up a known set of estimates rather than a full re-execution.

Track one more number: the average absolute variance between your soft close estimates and hard close actuals. When we started, ours was around 18% on accrued expenses. After two quarters of documenting methods and reviewing variances, it dropped under 7%. At that point the soft close P&L was trustworthy enough that leadership stopped waiting for the hard close to make decisions, which was the entire point.

A soft close that saves time isn't about moving faster through the same work. It's about doing each piece of judgment once, recording why you made it, and letting that decision stand. The moment you're redoing settled work, you've lost the time you thought you were saving.