Corporate Card Program KPIs: What to Track and What Good Looks Like

The corporate card program KPIs worth tracking are spend under card, rebate yield, out-of-policy rate, receipt and reconciliation lag, active cards, decline rate, and vendor card acceptance. Few have public benchmarks, so most programs set their own baseline.

Most card program reports show two numbers, total spend and rebate dollars, and both go up when the company grows. Neither tells you whether the program is doing its job. A program can double its rebate while half the cards sit in drawers and receipts arrive six weeks late. The ratios below are the ones that separate a healthy program from a large one.

Key Takeaways

• Rebate dollars rise with spend, so track rebate yield (rebate divided by card spend) and spend under card (eligible card spend divided by card-eligible spend) instead.

• Out-of-policy rate and reconciliation lag show whether the controls and the ERP posting actually work, which matters more to the close than any rebate.

• Dormant cards are risk with no benefit. Close cards with no activity for six months unless someone justifies keeping them.

• Decline rate and vendor card acceptance show where the program's friction sits, at the point of sale or with suppliers who won't take cards.

• Credible public benchmarks are rare. A twelve-month internal baseline, reviewed quarterly, beats a borrowed industry number.

What KPIs should you track for a corporate card program?

Track seven KPIs that cover adoption, value, control, and friction. Each one has a plain formula and a specific question it answers, and none needs special software, only clean card data and a defined denominator.

Several of these overlap with accounts payable KPIs once virtual cards carry supplier payments. Keep the two dashboards separate anyway, since AP metrics track invoices and card metrics track cardholders and merchants.

Which card program KPIs matter most to a CFO?

Spend under card, rebate yield, and reconciliation lag matter most to a CFO, because they answer whether the program is growing for the right reason, paying for itself, and keeping the books current.

My view is that rebate dollars are the most over-reported number in card programs. They make a good slide, and they say almost nothing on their own, since a 20% rise in spend lifts them by the same share with no change in how the program runs. When a CFO asks for best practices in managing a commercial card program, the honest short answer is to stop presenting rebate dollars without the yield and coverage ratios next to them.

The other four KPIs belong in the program administrator's monthly review, where they explain why the top three moved.

How do you measure spend under card and rebate yield?

Measure spend under card by dividing eligible card spend by total card-eligible spend for the same period, and rebate yield by dividing rebates earned by card spend over the same period. The hard part of both is the denominator, so define it before anyone reports the number.

Spend under card is narrower than spend under management, which covers every purchase governed by procurement. This KPI asks a simpler question about the purchases that should be on a card and how many actually are.

What counts as card-eligible spend?

Card-eligible spend is the spend your policy says belongs on a card, usually low-dollar, high-frequency purchases from suppliers that accept cards. Build it from AP history and card transactions together, deduplicated, and apply the same eligibility policy to both. Count only the card spend that meets those same rules, over the same period, in the numerator. If either side uses different rules or dates, spend under card can top 100%.

Delaware's state auditor gives a usable template. In its Fiscal Year 2019 Purchase Card Performance Metrics inspection, the Office of the Auditor of Accounts measured each agency's share of transactions under $2,500 paid by PCard, its share of $2,500 to $10,000 transactions on PCard, PCard spend as a share of total spend, and a card-to-employee ratio. That's a clean definition of eligibility by transaction size, and it works for a company as well as a state.

According to the same Delaware inspection, moving just 5% of the state's checkbook transactions under $2,500 to PCard would avoid an estimated $2,674,760 in processing cost, against $50,218 in added rebates. That ratio is worth showing anyone who thinks the card program exists for the rebate.

What does a good rebate yield look like?

A good rebate yield depends on your volume, your mix, and your contract, so the useful comparison is your yield against your own contract tiers. Federal figures show one mixed-program example rather than a target. According to GSA SmartPay's FY2025 program statistics, federal agencies spent $39.4 billion on SmartPay cards and earned $471 million in refunds, about 1.2%, across purchase, travel, and fleet cards combined.

Yield has a ceiling set by interchange, the fee the supplier's bank pays the card issuer, which funds most of the rebate. According to Visa's interchange reimbursement fee schedule effective April 18, 2026, the standard commercial card-not-present rate is 2.70% plus $0.10. Large-ticket transactions carry far lower rates, which is why a program that moves big supplier invoices onto cards can see yield fall while rebate dollars rise. Read the mechanics of interchange and card rebates before treating a falling yield as bad news.

Contract tiers matter as much as mix. Most programs pay higher rates as annual volume crosses thresholds, so consolidating cards with one issuer can raise yield on the same spend. Corpay pays its customers more than $800M in rebates each year across card programs, and virtual card rebates on supplier payments are usually where the yield grows fastest.

How do you track policy compliance and reconciliation lag?

Track policy compliance with an out-of-policy rate and reconciliation with a median lag in days, both pulled monthly. Together they show whether the controls work and whether card activity reaches the ledger in time for the close.

How should you calculate the out-of-policy rate?

Calculate the out-of-policy rate as transactions flagged for a policy exception divided by total transactions, then split it by exception type. A missing receipt, a blocked merchant category, and a split purchase are different problems with different fixes.

Blocked attempts deserve their own line. If card controls and spend policies are set well, many violations never post at all, because the authorization is declined at the terminal. A low posted out-of-policy rate with a high count of blocked attempts means the controls are doing the work, and that the policy probably needs a clearer explanation for cardholders. Repeated attempts in the same category are also an early sign of maverick spend looking for another route.

Commercial cards success story

See how commercial cards transformed expense management and reporting for a finance team — turning a manual burden into measurable savings and a more strategic AP function.

Read the success story

What's a reasonable receipt and reconciliation lag?

A reasonable receipt lag is days, and a reasonable reconciliation lag is short enough that every transaction is coded and approved before the books close. If card activity regularly lands after the close, the lag is too long, whatever the median says.

Set the receipt target from your close calendar and your policy, so every charge is documented and coded before the books close. For reimbursements, IRS Publication 463 treats accounting within 60 days as a reasonable time under an accountable plan, but that's a safe harbor for reimbursement, not a receipt deadline for card programs, and a target built around it would miss most closes.

Measure the median, then look at the slowest tenth of cardholders. In my experience the average is almost always fine. The trouble is a few heavy travelers who batch every receipt until the end of the quarter. Corporate card reconciliation gets faster mostly through capture at the point of purchase and automatic coding, not through more reminders. Direct integration with the ERP matters here too, because coded transactions that post without a manual import never sit in a reconciliation queue.

How do you measure card coverage, dormant cards, and declines?

Measure coverage as cards per employee or per eligible role, dormant cards as cards with no activity in a set window, and declines as declined authorizations over total authorizations. These three show whether the program has the right cards in the right hands.

How many cards per employee is right?

The right number of cards is the number of people whose jobs require buying, and no more. A higher ratio helps only when those cards carry spend.

Delaware's inspection shows how wide the range runs even inside one government. Among state agencies, the card-to-employee ratio went from 6.54% in the Judicial branch to 81.64% at the Department of Natural Resources and Environmental Control. Both can be right. Field staff buying parts need cards, and court staff mostly don't.

The approval side needs its own ratio. In its audit of Air Force purchase cards, GAO reported that installations were directed to keep each approving official responsible for no more than seven cardholder accounts. That limit is old and specific to the Air Force, but the idea transfers. Track cardholders per approver, and if one manager reviews forty statements a month, the review isn't really happening.

When does a card count as dormant?

A card counts as dormant when it has no transactions for a defined window, and six months is a common threshold. Close dormant cards unless the cardholder's manager justifies keeping them in writing.

The same GAO audit shows what a cleanup can find. Under a policy of closing accounts unused for six months, the Air Force canceled about 27,000 purchase card accounts, about 36% of the total, in the six months that ended March 31, 2003. Those numbers are two decades old, but most programs that have never run a dormancy sweep would likely find a smaller version of the same thing. Every idle card is a fraud target that earns nothing.

What does the decline rate tell you?

The decline rate tells you where cardholders hit friction, and it means nothing until you split it by reason code. Declines for a blocked merchant category are controls working. Declines for insufficient limit or suspected fraud on a routine purchase are friction that pushes people back to personal cards and reimbursement.

There's no credible public benchmark for corporate card decline rates worth quoting. Set your own baseline from the last six months, and investigate any cardholder or merchant category that runs well above it.

How do you measure vendor card acceptance?

Measure vendor card acceptance as the share of suppliers contacted for card payment who now accept it, and weight it by spend. Twenty small vendors on card can matter less than one large supplier that still wants a check.

This KPI belongs to the virtual card side of the program, where AP pays supplier invoices by card. It moves only with outreach, so track it next to a count of vendors contacted. A vendor enrollment campaign that calls the same list twice a year will produce a flat line no matter how good the offer is.

Supplier cost explains most refusals. Suppliers pay interchange on every card payment, and on a large invoice that's a real number. Passing Level 2 and Level 3 card data with the payment can qualify a supplier for lower rates, and according to the same Visa schedule, the enhanced-data commercial rate is 1.75% plus $0.10. Bring that rate into the conversation with suppliers who refuse cards on cost.

Fraud exposure is the other argument. Corpay pays suppliers with single-use virtual cards, each locked to one invoice amount, so a stolen card number is worth nothing to anyone else.

How do you set a baseline when there's no benchmark?

Set a baseline from your own last twelve months of card data, segmented by card type and department, and review it quarterly. An internal trend is more honest than a borrowed number from a survey with a different program mix.

Most published card program "benchmarks" come from vendors or from surveys whose methods you can't inspect, and many blend travel, purchasing, and fleet programs into one figure that matches nobody. If you can't see how a number was built, don't put it in front of your CFO as a target.

A workable baseline process looks like this:

1. Pull twelve months of transactions, authorizations, and cardholder data from the issuer.

2. Define each denominator once, in writing, especially card-eligible spend.

3. Calculate each KPI by card type, because travel cards and virtual cards behave differently.

4. Set targets as movement from the baseline, such as five more points of spend under card, not as an absolute number.

5. Recalculate the baseline each year so growth doesn't flatter the trend.

How often should you review card program KPIs?

Review the program administrator's KPIs monthly and the CFO-level KPIs quarterly. Monthly review catches dormant cards, decline spikes, and late receipts while they're still small, and quarterly review is the right pace for spend under card and yield, which move slowly.

Peer comparison stays an open question for most programs. The data mostly isn't public, and the closest thing many teams get is their issuer's view of similar clients. Ask how it was built before trusting it.

Track and grow your card program with Corpay corporate cards

The KPIs that matter depend on card data that arrives clean, coded, and on time, plus a supplier base willing to take cards. Most programs struggle on those two points rather than on the reporting.

Corpay corporate cards give finance teams spend controls, receipt capture, and card data that posts to the ERP with its coding attached, through 100+ ERP integrations including NetSuite, Sage Intacct, Business Central, and Acumatica. For supplier payments, Corpay virtual cards use single-use card numbers, and Corpay's team handles supplier enrollment so acceptance grows without your AP staff making the calls. Card spend earns rebates on every program, and AP teams save about 40% of the time they spend on payment processing. Most programs are live in weeks.

Frequently Asked Questions

What is a good KPI for a corporate card program?

Spend under card is the best single KPI, because it measures how much of the spend the program was designed for actually runs through it. Pair it with rebate yield and reconciliation lag to see whether that spend pays off and reaches the ledger on time.

How do you calculate rebate yield on a card program?

Divide rebates earned by card spend for the same period. Compare the result with your contract's rebate tiers rather than with another company's number, since yield depends on volume, transaction size, and how much spend runs through large-ticket or enhanced-data rates.

How do you find dormant corporate cards?

Run a report of cards with no transactions in the last six months. Ask each cardholder's manager to justify keeping the card in writing, and close the rest. Repeat the sweep at least twice a year, and remove cards promptly when employees leave or change roles.

What is a normal corporate card decline rate?

There's no reliable public benchmark for corporate card decline rates. Set a baseline from your last six months of authorizations, split it by reason code, and investigate cardholders or merchant categories that run well above it. Declines for blocked categories are controls working.

David Luther

Product Marketing Program Manager

David Luther, MBA is a product marketing program manager with years of experience in commercial banking, finance, and technology sectors, with research and writing appearing in financial publications.

Commercial Cards

Expense management