For life sciences teams

A true-up is not a correction. It is the one moment the organization finds out which of its assumptions was wrong, and most organizations spend it arguing about the number instead.

This guide shows how to decompose the variance by driver before booking it, route each component to its functional owner with the assumption quoted, report gross movements before netting, and keep a record of which assumptions failed. Skip it and you repeat the same variance every quarter, paying the full cost of being wrong and collecting none of the information.

Financial NavigationPhases 2 and 3The 5 Layers of GTN™ Layer 05

What this guide is for

Actuals have arrived and they do not match. This guide is about what to do with that information, and specifically about how to stop the true-up from being the end of the conversation when it is the most useful beginning available.

The failure it addresses

A true-up gets treated as an accounting event. The difference is quantified, the entry is booked, the rate is adjusted, and the period closes. Everyone involved has done their job correctly and the organization has learned nothing.

Four steps that turn a true-up into information A true-up should be decomposed by driver before it is booked, so each component can be routed to the function that owns the assumption behind it, with that assumption quoted. Gross movements are reported before netting, because netting can hide two errors that happened to offset each other. Finally the assumptions that failed are recorded, so the same variance is not repeated next quarter. A true-up handled as a pure accounting event costs the money and produces no learning. STEP 1Decompose bydriverBefore booking it, split thevariance into its causes.STEP 2Route to theownerEach component goes to thefunction, with the assumption quoted.STEP 3Report grossbefore netNetting hides two errors thathappened to cancel out.STEP 4Record whatfailedKeep the list of assumptionsthat did not hold.A true-up treated as an accounting event costs you the money and teaches you nothing.Skip this and you pay the full price of being wrong, every quarter, and collect none of the information.
FIGURE 1. The variance is the price. The decomposition is the only part you get to keep.

The information was in the variance, not in the entry. A variance is the only reliable signal that an assumption held somewhere upstream is no longer true, and it arrives already attributed to Finance because Finance is where it becomes visible. That attribution is the problem. The assumption that failed was almost never Finance’s.

There is also a quieter failure. A true-up that nets to something small looks like a good period. It can just as easily be two large movements in opposite directions, one in coverage and one in affordability, cancelling each other out. Netting is how an organization congratulates itself on an accrual that was wrong twice.

The use cases you can run

Use case 01

Decompose before you book

Split the variance by driver before anything is recorded: volume, mix, rate, lag, and eligibility. Booking first fixes the number and ends the inquiry, because once the entry exists the question feels answered.

What you needthe claim detail behind the variance and the assumption inventory the accrual was built on.
What you geta variance you can attribute to a behavior rather than to a period.
Use case 02

Route each component to its owner

Every decomposed piece goes back to the function that owns the assumption underneath it, with the assumption quoted. Not as blame. As information that function has no other way of receiving, because they do not see the financial consequence of their own operating change.

What you needthe decomposition, and a standing forum where Commercial, Market Access, Trade, Patient Services and Finance all sit.
What you getthe loop closed. This is the single highest-value habit in the whole discipline and it is almost never in place.
Use case 03

Ban the net

Report gross movements, both directions, before any netting. Then net.

What you neednothing but the discipline to do it.
What you getthe two large offsetting errors you were about to miss.
Use case 04

Write the assumption obituary

For every assumption that failed, record what it was, when it stopped being true, how long it took to notice, and what would have made you notice sooner. Keep the file. It becomes the most honest forecasting asset the organization owns.

What you needsomewhere durable to write it, and the willingness to keep a record of being wrong.
What you geta detection-lag number that gets shorter over time, which is a better measure of a finance organization than accuracy is.

What good looks like

Variances are decomposed before they are booked. Each component has a named owner outside Finance. Gross movements are visible before netting. And the organization can say how long it takes to notice that an assumption has stopped being true, because it measures that.

What it costs to skip it

The same variance, next quarter, with the same explanation. An organization that trues up without governing the true-up is paying the full cost of being wrong and collecting none of the information.

Where this connects

The accrual this trues up is Accrual Design. The reason a variance that surfaces in reporting was created three layers upstream is the 5 Layers of GTN™. The room where the routing in use case 02 actually happens is the Alignment Labs™. The patient-side view of the same ten exits is at Where did your treatment stop? What is actually testable at the first three checkpoints after launch is set out in Your First Ninety Days.

Former employers and products referenced are matters of professional history. The views expressed are my own and are not made on behalf of any current or former employer, and no commercial outcome for any named product is attributed to me personally. All trademarks belong to their respective owners.

Disclosure: who pays for this work
LinkedIn Newsletter Podcast YouTube Instagram