SAP Concur ExpenseIt: Pricing and What It Reads

Plan rates come from the SAP Concur US pricing page. Product behavior comes from the SAP Concur ExpenseIt product page and SAP's own ExpenseIt training material.

ExpenseIt is the receipt scanning feature inside SAP Concur Expense. You photograph a receipt, and it creates a categorized, itemized expense entry rather than a spreadsheet row. The buying detail most people miss: ExpenseIt is not part of Concur's entry Base plan. It starts on Plus, which SAP Concur lists at starting at $11 per expense report against $7 for Base, so turning on receipt scanning raises the published rate by about 57 percent.

That single fact decides most ExpenseIt questions. Below is what it reads, what it refuses to read, the page limit almost nobody mentions, and how to work out whether the step up to Plus is worth it for your team.

Which SAP Concur plan includes ExpenseIt?

Plus and Premium. The SAP Concur pricing page lists Base with Concur Expense on its own, while Plus adds ExpenseIt, the User Support Desk and Reporting. Premium adds Concur Travel and Intelligence on top and is custom priced. Every tier lists unlimited users, so the plan step is about features, not headcount.

Concur adds that price can vary based on monthly commitment, and that the more you purchase with your contract, the less you pay per report. Both published figures are floors. A full breakdown of the tiers and the annual math sits on our Concur pricing page.

How much does ExpenseIt cost?

SAP Concur does not sell ExpenseIt with its own price tag on the public rate card. It is bundled into Plus, so the honest answer is that ExpenseIt costs you the gap between Base and Plus: $4 per expense report at list. What that means annually depends entirely on how many reports your people file, because Concur bills by the report rather than by the seat.

Those are list prices multiplied out, shown so you can check the arithmetic. Negotiated rates run lower. The pattern worth noticing is that the cost of receipt scanning scales with filing frequency, so the teams that most want automatic capture, the ones filing constantly, are the ones who pay most for it.

What data does ExpenseIt extract from a receipt?

SAP's training material lists the amount spent, the currency involved, the date of transaction, location, nature of expense, method of payment, vendor details, and hotel itemizations. The Concur product page describes the same job as identifying data like amount, date and location, then automatically categorizing and itemizing the expense using a mix of OCR, machine learning and generative AI.

Two of those predictions are smarter than plain text recognition. Expense type is predicted from your own previous history with that vendor, so the categorization improves as an individual files more. Payment type is worked out by matching the last four digits printed on the receipt against your corporate cards, which is how a charge gets flagged as card rather than cash.

What ExpenseIt does not capture

This is the part that catches restaurant-heavy teams, and it comes straight from SAP's own documentation rather than from a review site. ExpenseIt does not capture handwritten components like tips, sub-total, or tax amounts. It can read a handwritten total if the total is written clearly, but the handwritten tip line on a US restaurant slip is not extracted.

There is also a page limit that is rarely mentioned. A user can capture up to nine pages in a single receipt, but ExpenseIt will only analyze up to three of them: the first two pages and the last page. Anything in the middle of a long folio is not read. For a five night hotel bill or a multi-page conference invoice, that means the middle detail has to be handled by a person.

Is ExpenseIt accurate?

SAP Concur publishes no accuracy percentage for ExpenseIt, which is normal: no major receipt OCR vendor publishes one. Concur's public claims are directional rather than numeric, saying ExpenseIt is more accurate than ever with generative AI and that it can itemize even the most complex hotel bills in less than 30 seconds. Treat those as vendor claims, because there is no published benchmark behind them.

The documented limits above are a more useful accuracy signal than any headline number. A tool that skips handwritten tips and reads only three pages of nine has known, predictable gaps, and knowing where they are beats a percentage you cannot audit. We wrote up how the whole category reports this in our guide to how accurate receipt OCR really is.

Does ExpenseIt work offline?

Yes. Concur states that ExpenseIt works in airplane mode, with images added to a queue and processed once the device reconnects. That matters for travel, where the receipt lands in your hand on a plane or in a country where you have turned data roaming off. The capture happens immediately, the extraction happens later.

Why is ExpenseIt not working?

Most reported failures trace back to the documented limits rather than a fault. If a tip or a tax line came back blank, it was handwritten and is out of scope. If detail from the middle of a long receipt is missing, it fell outside the three analyzed pages. If a charge was categorized as cash instead of card, the last four digits were not legible enough to match. If nothing processed at all, the image may still be queued from an offline capture.

ExpenseIt vs a standalone receipt OCR tool

These are different products aimed at different endings. ExpenseIt exists to populate an expense entry inside Concur so it can be routed for approval and reimbursed. It is excellent at that and useless outside it, because the output lives in Concur. A standalone extraction tool ends with a file: an Excel or CSV export you can push into any ledger.

If your actual problem is a shoebox of vendor receipts that has to become coded data in a ledger, and nobody needs to be reimbursed, you are buying an approval workflow you will never open. That is the case our receipt OCR software is built for, and the Concur alternative page covers the switch in more detail.

Should you pay the step up to Plus?

Work it as a labor trade rather than a feature preference. Take your monthly report count, multiply by $4, and that is what ExpenseIt costs you at list. Then estimate the minutes your team currently spends keying expense lines by hand each month and price that at a loaded hourly rate. If the manual number is bigger, the upgrade pays. At 30 reports a month, ExpenseIt costs about $120 a month at list, so it pays for itself if it saves roughly two to three hours of finance time.

Two things sharpen that estimate. Volume discounts move the real per-report rate below list, so ask what the rate becomes at your committed volume before modeling anything. And the handwritten tip limit matters more than it sounds if your spend is restaurant heavy, because those lines still come back to a human either way.

The rollout itself is the cost people forget. Automatic capture only works if several hundred employees actually photograph receipts at the point of sale instead of stuffing them in a wallet, which is a habit change rather than a software setting. Most finance teams underestimate how much of the return depends on getting that adoption right, and it is worth handling the same way you would any other policy change you need to train and certify a whole company on, with a short course and a record of who completed it, rather than a single launch email.

The short version

ExpenseIt is a capable receipt reader with a specific job: filling in Concur expense entries. It reads amount, currency, date, location, expense type, payment type, vendor and hotel itemizations, it works offline, and it gets better at categorization as it learns a user's vendor history. It does not read handwritten tips, sub-totals or tax, and it analyzes only three pages of a nine page capture. It is not available on Concur's $7 Base plan, so budget the $11 Plus rate if receipt scanning is why you are buying. If you need receipt data as a file rather than as an expense claim, the whole per-report model is the wrong shape and a per-document extraction tool will cost a fraction of it.