Published 7 Jul 2026

I spent a long stretch of my career in financial reporting, closing the books and putting my name to numbers that were, if I am honest, already history by the time anyone opened the deck. That is the quiet problem with a spend report. It is a rear-view mirror. It tells you what you paid, cleanly and accurately, a few weeks after you could have done anything about it.
The leaks were never invisible. They were sitting in the ERP, in the purchase orders, in the invoice file, well before they surfaced in a report. The report just was not built to catch them early. It rolls everything up after the fact and shows you the total. It cannot tell you why the total moved, or what to do next week to stop it moving the same way again.
So here are nine places source-to-pay tends to leak, why your reports keep missing them, and what changes when the signal reaches you while you can still act. I will use one company to keep it concrete: roughly 180M in annual spend, 340 suppliers, about 12,000 invoices a year.
Every finance team has that catch-all GL line. It is where spend goes to disappear. In our example company, about 14M is being bought off-contract, at list, from suppliers we already hold negotiated rates with. Nobody decided to do that. It happened one rushed purchase at a time, and got coded into a category broad enough to hide it.
Rolled up by account, that 14M reads as ordinary category spend. You would never pick it out. The report is not lying to you, it simply cannot see a contract, so it cannot tell you which purchases ignored one.
Procurement Spend Analysis classifies spend down to the contract it should have used, and hands you the list: this purchase, this supplier, the rate you paid against the rate you already had in a signed agreement.
Pull your spend by value and the picture is reassuring. A dozen suppliers, most of the money, all under management. Now pull it by transaction count. Suddenly there are two hundred-odd suppliers at the bottom generating most of your POs and most of your invoices for a sliver of the spend. That is the tail, and it is where procurement and accounts payable quietly lose their week.
A value-ranked report buries the tail by design. The cost of it was never in the spend column. It was in the volume, and nobody was counting the volume.
Tail Spend Detection ranks the tail by the work it creates rather than the money it moves, and shows you what a catalog or a consolidation would lift off your team.
Anyone who has run an accounts payable function knows this one in their gut. Same invoice keyed twice. A credit that never landed. A currency or rounding slip on a big number. Out of 12,000 invoices a year only a handful go wrong, and every one of them is money walking out the door.
Here is the trouble. On the report a duplicate looks exactly like a legitimate payment, right up until someone reconciles it, and reconciliation happens after the run has gone out.
Accounts Payable Exception Prediction reads the invoice and payment stream before the run clears and flags the ones that do not smell right, with the reason attached, while you can still pull them.
You know the moves, because everyone who has ever approved a PO has seen them. Orders split to sit just under the sign-off threshold. A PO raised after the invoice turned up, to tidy the paper trail. A supplier who is not on the approved list but happens to know someone. None of it is fraud. All of it is control quietly slipping.
Your report shows these as processed orders, because that is what they are. It has no idea what your policy says, so it cannot tell you which ones bent it.
Purchase Order Anomaly Detection learns what a clean order looks like in your business and flags the ones that stray, ranked by how far.
This one still catches good teams out. A supplier wins the line on unit price. Then the freight is worse, the reject rate is higher, the reorders eat the saving, and by the time you add it all up they were the expensive option all along. The PO was right. The decision was wrong.
A spend report only ever saw the price. The rest of the cost lived somewhere else entirely, in logistics, in quality, in the inventory you were forced to hold, so it never made it into the comparison.
Total Cost of Ownership Modeling pulls those other costs back in and shows the real landed cost per supplier, so the cheapest quote gets judged on what it actually costs you.
Input prices rarely jump. They drift. A key material creeps up 9 percent across two quarters, and because no single order looks wrong, nobody raises a hand, and you carry on buying at spot on the day you happen to need it, which is usually the worst day to buy.
Your report compares this month's price to last month's. Useful, and about a quarter too late. It has no view of where the price is heading, or that a different buying rhythm would have beaten it.
Material Price Prediction forecasts the input prices that matter and shows you where buying ahead, or simply buying differently, protects the margin.
This is the one that used to keep me up. Half your suppliers are paid days early for no reason anyone can name, tying up cash you needed elsewhere. The other half are sitting on early-payment discounts worth more than that cash, and those are lapsing unclaimed. Both problems at once, pulling in opposite directions.
A payables report shows what is due and what has gone out. It will not tell you the right day to pay each invoice, weighing the discount against the cash position against how much the supplier actually matters to you.
Payment Timing Optimization and Vendor Payment Optimization work that out invoice by invoice, so the cash stays put where it should and the discounts worth taking get taken.
Look closely at a single category and you will often find it spread across seven suppliers where three would cover it. Every extra name is a rate you did not negotiate hard enough, plus onboarding, plus management, plus one more line in the payment run. Fragmentation costs you in ways the spend total never shows.
The report lists spend by supplier. It does not know that several of those suppliers are interchangeable, or that pulling them together would win you a better price and a lighter load.
Vendor Consolidation Analysis finds the overlap and models the trade honestly: the rate you would gain against the risk you take on by leaning on fewer names.
Somewhere in the 340 is a single-source supplier carrying more of your spend than is comfortable, and the early signs are usually there if you are looking. Deliveries slipping. Payment behavior changing. The odd financial-distress signal. If that one fails, a line stops, and you tend to find out the hard way.
Your spend report tells you how much you buy from them. It will not tell you they are turning into a risk, because the signals that matter are not in the spend data at all.
Supplier Risk Prediction pulls delivery data, financial signals, and payment behavior into one score, and surfaces the critical suppliers early enough to line up an alternative before you need one.
Notice what these have in common. In every case the signal was already in your data, and the report was a lap behind it. That is not a knock on your reporting team. It is simply what business intelligence is, a very good account of what already happened. It stops at What.
The two questions that actually protect the number come next. Why is this happening, and what do I do about it before the next run. That is the gap eyko is built for. It reads your source-to-pay data on a beat, works out the Why, and gives you the What Next in plain terms: hold this payment, redirect this purchase, qualify this supplier. Not a prettier dashboard. A shorter list of things to go and do while they still matter.
If you want to see the whole picture joined up, spend, suppliers, contracts, and payments read together, that is the Source-to-Pay page.
New to the category? Learn what decision intelligence is and why it changes how teams act on data.

COO & Co-Founder
7 Jul 2026
Jon Louvar is the COO and co-founder of eyko. He was previously VP of Product Marketing at insightsoftware and, before that, Manager of Financial Reporting at Silgan Containers, building BI and reporting platforms across finance, operations, and supply chain for enterprise organizations. At eyko he leads operations and delivery, translating customer insight into product execution.
The usual suspects: off-contract spend, a growing supplier tail, duplicate or erroneous invoices, purchase orders that break policy, buying on price instead of total cost, input prices bought at the wrong time, payment timing that wastes cash or misses discounts, too many suppliers for one category, and a critical supplier turning into a risk. Each one shows up in your source-to-pay data before it reaches a spend report.
A report is a record of what happened. It rolls spend up after the invoices are paid and the contracts are signed, so you see the result of a leak rather than the signal ahead of it, with no read on why the number moved or what to do next.
A dashboard shows you spend. eyko reads the same data and answers the two questions a dashboard cannot, why the number moved and what to do about it, returned as a ranked set of actions while there is still time to act.
It reads straight from the ERP and procurement systems you already run, including JD Edwards, Oracle EBS, SAP, and NetSuite, alongside your existing data platform and BI tools. Nothing to warehouse first.
Yes. It sizes each leak against its driver: what the off-contract spend is costing you by supplier, how much of the duplicate payments you can recover, which discounts are about to lapse. A number to act on, not a hunch.
Join the enterprises replacing weeks of manual analysis with a single prompt. See what eyko Playbooks can do with your data.
The latest industry news, interviews, technologies, and resources.

A revenue report shows last quarter's number. Here are ten order-to-cash warning signs that surface too late, the evidence behind each, and where the signal is hiding in your own systems.
Read post

A treasury dashboard shows the position after it has moved. Here are ten cash and treasury warning signs that surface too late, the evidence behind each, and where the signal is hiding in your own systems.
Read post

Gartner says CFOs want to cut costs and forecast better, yet only 36% trust AI to help. The gap is not AI. It is the layer that turns data into a decision, after the dashboard.
Read post