Most companies discover their margin leakage too late to act upon the insight. The quarter closes, gross margin comes in below plan, and finance starts asking questions. An analyst spends a week or two stitching invoice data to rebate accruals and carrier bills, and a waterfall chart appears in the quarterly review about a month after close. Everyone agrees it’s useful. By then the orders that caused the miss have shipped, and a good share of next quarter’s orders are already booked with the same issues.
In the Six Things piece, the second job on the list was to show where margin is leaking while there’s still time to act. Its sample line set the bar: finance and sales can see pocket price and pocket margin by customer and SKU on any day of the month, without pulling an analyst off something else for a week. Last week was about the record that makes that possible. This week is about why leakage tends to be found so late, and what changes when you can see it in the same week it happens.
Leakage shows up late because the data shows up late
The invoice is the only part of a transaction that arrives on time. Everything between invoice price and pocket margin runs on a different clock.
Rebates accrue monthly at best, often against an estimated rate, and settle quarterly or annually. The freight bill from the carrier shows up weeks after the truck leaves, and nobody ties it back to the order unless someone is asked to. Credits and claims trickle in over the following month. Standard cost may be revalued at period end, which quietly changes the margin on everything already sold.
So the full (final, actual) waterfall can only be built once all of those pieces have arrived, and in most companies that means after close. Building it is also a project rather than a routine. Somebody has to pull each source, line up the keys and spread the off-invoice amounts back across the transactions that earned them. That person usually has a day job, which is why the exercise happens once a quarter, when the variance forces it.
Put the timeline together and the gap is wider than most people expect. A discount that starts slipping in the second week of a quarter is typically first seen in a review deck three or four weeks after the quarter ends. That’s roughly fifteen weeks of the same terms on every order that matched them.
Assumes a leak starting in week 2 and a review deck about four weeks after close, as in the post. Hover or focus a row to read it.
A quarterly SUM or AVG hides where the leak is coming from
The quarter-end view also tends to be a total, and totals move slowly. A pocket margin that slipped 0.8 points looks like a modest variance on a slide. Underneath it, the leak is almost always concentrated.
Here is an illustrative example for a distributor doing about $60M a quarter. A 0.8-point slip in pocket margin is about $480K. When the waterfall is finally built at the transaction level, that $480K might break down like this:
- Two reps applying manual discounts well beyond their usual range on a handful of accounts: $190K
- Lines shipping below cost because a supplier increase reached the cost file before it reached the price file: $140K
- Freight waived for customers whose terms say they pay it: $95K
- One large job quote priced well below its peers and never revisited: $55K
Illustrative, not client data. The total is invisible in a quarterly average; each driver was visible within days of its first invoice.
None of these is visible in the total, and each one has a different owner and a different fix. Most of them were visible within days of the first affected invoice, to anyone looking at the right line.
We covered why the distribution matters more than the average in the price waterfall primer back in April. The point here is about timing. The same breakdown that explains last quarter could have prevented most of it.
What finding it early actually takes
Seeing leakage in the same week starts with the one record from last week’s piece, because every line has to be joined to the price, agreement and rebate program that applied to it. On top of that record, a few practical changes make the difference.
The view has to refresh as invoices post, daily or weekly, rather than when someone has time to rebuild it. That sounds obvious, but it rules out anything that depends on an analyst running a workbook.
Off-invoice amounts get estimated at the time of sale. If a customer’s rebate program pays 3% on this product group, the line can carry a 3% accrual the day it’s invoiced. If freight is billed by weight and lane, the line can carry an estimated freight cost before the carrier invoice arrives. When the actuals come in, they replace the estimates, and the difference becomes a useful measure of how good the estimates were.
Each line gets compared to its peers instead of to an overall average. A 22% discount means one thing on a small order of a commodity item and something quite different on a large order of a specialty product. Grouping lines by product group, customer segment and order size, and then looking at where each line sits within its group, is what separates a real outlier from normal variation.
Every flag also needs a dollar size and an owner. A list of 2,800 lines below target is not something anyone can work through on a Monday. A short list sorted by margin dollars, with the rep, the customer and the reason code attached, usually is. Turning those flags into specific recommendations is a separate job, and it gets its own piece later in the series.
Estimates are good enough to act on
The usual objection comes from finance, and it’s a fair one: accruals and freight estimates aren’t actuals, and nobody wants to chase a number that turns out to be wrong.
In practice, estimates built from the agreement terms and recent carrier rates are close enough to show direction and concentration, which is what an early warning needs. Knowing that one customer’s pocket margin dropped six points this month is worth acting on, whether the true number turns out to be 5.6 or 6.3. The quarter-end true-up still happens. It just becomes a check on numbers people have already seen, instead of the first time anyone sees them.
Someone has to be looking
The last gap is ownership. Leakage tends to sit between sales and finance. Sales owns the discount and finance owns the margin, and the waterfall belongs to neither. Without a named owner and a regular slot on the calendar, even a good early-warning view becomes a page nobody opens.
The companies that do this well usually keep it simple: one person owns the leakage review, it happens weekly, and it covers the top ten flags by dollars. Thirty minutes is often enough. The habit matters more than the tooling, although the tooling decides whether the habit is realistic.
Where to start tomorrow
Take the explanation for last quarter’s biggest margin variance, the one that ended up in the review deck. For each driver on it, find two dates: the date of the first invoice where the problem showed up, and the date someone first raised it in a meeting or email. The weeks between those two dates are your detection lag.
Then look at the driver with the longest lag and ask what you would have needed to see it within a week. Usually the answer is a join that nobody has automated, or a report that nobody owns. That’s a short and concrete list to bring to finance and sales, and it gives you a number to improve on next quarter.
Extra credit
For this last part, I need to credit Rick Robinson at KKR, who is a bona fide pricing expert that I’ve had the pleasure of working with a few times in our careers. Everything I wrote above starts from the invoice, because it’s often the cleanest record of what was sold and at what price. Orders come earlier. For a distributor shipping from stock the gap may be a day or two, but for made-to-order products, project work or blanket orders, an order can sit in the backlog for weeks or months before it ships and gets invoiced. Running the same price realization and leakage checks on orders means a discount outside its peer range, or a line priced below current cost, shows up while someone can still review it, often before anything has shipped and sometimes before the customer has been confirmed at that price.
Order data comes with caveats. Orders get changed, cancelled and partially shipped, and freight and rebates on an order are estimates by definition. So treat the order view as an early read and confirm it against invoices as they post. Where the backlog is long, though, orders are usually where the earliest useful signal is, and the detection lag from the exercise above can drop below zero: the leak gets caught before the first invoice goes out.
Next in the series: seeing the cost shock before it hits your margin.
At Revomo, this weekly view is part of Analytics & Insights, built on the Data & Context Studio. If you’d like to see your own price-to-pocket waterfall at the transaction level, our Margin Leakage Assessment is a free first look.