Six spend channels. Hard controls on one of them.

Spend control software: spend visibility, budget control and expense control alerts for every department

Every roundup in this category lists the same platforms and calls all of their features controls. Some of those features stop money leaving. Most of them tell you afterwards. This page separates the two, channel by channel, and works out how many days of warning each kind of control actually gives you on a real department budget.

Compare the alert tools

No sales call. Plans are flat per month, not per user.

Try it on a real budget

Which day would your control have fired?

Set a department budget below, choose the thresholds you use today, and replay a month of spending through the engine. The point of the exercise is the gap between the day a percentage threshold trips and the day a run rate already knew the month was going to end over. This runs in your browser and nothing is uploaded anywhere.

Data source

Loading transactions…

Monthly budgets

Anomaly sensitivity

Low Medium High

Alert channels

Email SMS Slack

Pick a company profile, set budgets and sensitivity, then press Run monitoring. The engine replays 30 days of card spend right here.

Short answer

What is spend control software, and what can it actually control?

Spend control software is the set of tools that decide, enforce and report how a company is allowed to spend money: spending limits, merchant rules, approval gates, budget thresholds and alerts. The distinction that decides most purchases is that only one of the six channels a department spends through can be genuinely blocked. On cards the platform issued itself, rules are evaluated at the authorization message and violating charges are declined outright. On your existing bank cards, employee reimbursements, vendor invoices and signed commitments, every tool in the category is reporting after the fact, and the lag runs from a few days to a full month.

The finding

A control stops money. Everything else is a notification.

We took the six ways a normal department actually spends money and asked one question about each: can any tool in this category refuse the spend before it happens, or does it only find out afterwards? The answers are not evenly distributed, and the channel with the strongest controls is usually not the channel with the largest numbers.

Swipe the table sideways to compare all columns.

Spend channel	Can it be blocked?	The mechanism	Realistic time to notice
Card issued by the platform Ramp, Brex, BILL, Emburse, Navan	Yes, hard	Spend limits, per-transaction limits and merchant category rules are evaluated at the authorization message. Violating charges are declined and the cardholder is notified.	Immediate. The money never moves.
Virtual card for one vendor	Yes, hard	A single-vendor or single-use card with its own ceiling. Pausing the card stops the next renewal without contacting the vendor. Only works on cards that platform issued.	Immediate, within that platform only.
Your existing bank cards Chase, Amex, Citi, Capital One	No	Read-only feed import. The issuer decides authorizations, not the software. Pending transactions are generally not imported at all.	1 to 3 business days after the charge, once it posts.
Employee out of pocket	No	A policy rule flags the claim when it is submitted, which is after the employee has already paid. You can refuse the reimbursement, which is a different and much worse conversation.	Whenever the employee files. Often weeks.
Vendor invoices and AP	Partly	An approval gate can stop the payment, but not the purchase. By the time an invoice exists the obligation exists too. You are controlling cash timing, not the commitment.	On invoice receipt, which is set by the vendor.
Signed but uninvoiced commitments SOWs, renewals, open POs	No	There is no transaction yet, so there is nothing for a card or expense platform to see. Only a procurement system that captures the commitment at signature knows about it.	Often not until the first invoice lands.

Read the table by column and the shape of the category becomes obvious. Hard controls exist on exactly one channel and its variant: the cards the platform issued itself. That is not a criticism of those products, it is how card authorization works, and Ramp's own documentation confirms that limits and merchant rules are checked at the authorization message and violating transactions are declined on the spot. The problem is that a department's largest lines are usually agency work, contractors, events and software renewals, and those arrive through the bottom three rows.

The arithmetic

The tiered threshold everyone recommends is the slowest signal

Ask any vendor how to set budget controls and you will get some version of a three-tier threshold: warn at 70 percent, escalate at 90, hard stop at 100. It is sensible advice and it prevents alert fatigue. It is also, on both of the overruns that actually happen, the last thing in the room to notice. Here is the same $60,000 monthly department budget under two different failure modes, with every number worked out so you can check it.

Swipe the table sideways to compare all columns.

Control	Steady overrun $2,600 a day, ends at $78,000 (130%)	Commitment overrun $1,700 a day plus an $18,000 SOW, ends at $69,000 (115%)
Threshold at 70%	Day 17. $42,000 is crossed between day 16 ($41,600) and day 17 ($44,200).	Day 25. Paid spend reaches $42,500 on day 25. One day of warning.
Escalation at 90%	Day 21. $54,000 is crossed at $54,600.	Day 26. The invoice posts and the budget jumps straight past 90 to 103.7 percent.
Hard stop at 100%	Day 24. $60,000 is crossed at $62,400. The budget is gone.	Day 26, the same day as the escalation. Both tiers fire at once and the money is already committed.
Run rate on transactions	Day 5. $13,000 spent over 5 of 30 days projects to $78,000, or 130 percent. Twelve days earlier than the 70 percent tier.	Never. Paid spend of $1,700 a day projects to $51,000, which is 85 percent. The projection stays comfortably inside budget all month.
Commitment at signature	Nothing to catch. This overrun has no contract behind it.	Day 3. $18,000 committed plus $51,000 projected is $69,000, or 115 percent, the day the SOW is signed. Twenty-three days of warning.

Two things fall out of that table. The first is that a run rate beats a threshold by twelve days on the ordinary case, because it uses the slope rather than the level. Five days of data on a $2,600 daily burn is enough to know the month ends 30 percent over, and no percentage threshold can know that until the percentage arrives.

The second is more uncomfortable. In the commitment case the run rate is not slow, it is wrong. It reports 85 percent all month, because the thing driving the overrun is not a spending rate, it is a signature. Every control that watches transactions is blind to it until the invoice appears on day 26, and by then the only decision left is which other budget absorbs it. This is why the honest answer to "which control should we buy" depends on which of your two failure modes costs you more, and most finance teams have never separated them.

Both scenarios use a 30-day month and a $60,000 budget for legibility. Substitute your own numbers and the ordering does not change: level-based controls always fire later than slope-based ones, and neither sees a commitment. If you want to watch this happen on your own figures, the simulator higher up the page runs the same arithmetic interactively.

The controls themselves

What spending limits can you actually set?

On issued cards there are five limit types and they nest inside each other. A charge that passes one can still be declined by any of the other four, which is the single most common source of confused support tickets when a company first turns controls on.

Limit 1

Company credit line

The ceiling on everything, set by the issuer against your cash or credit profile. Every other limit lives inside it, and it is the one you cannot configure.
Limit 2

Cardholder limit

What one person or one card can spend in a period. The limit most teams think of as "the" limit, and the only one many companies ever set.
Limit 3

Per-transaction limit

A ceiling on any single purchase. This is what stops one unapproved annual renewal from consuming a whole quarterly allowance in one charge.
Limit 4

Merchant category restriction

An allow list or block list of merchant category codes. Powerful and blunt: MCCs are assigned by the acquirer, so a legitimate vendor occasionally sits in a category you blocked.
Limit 5

Velocity limit

How many times a card can be used in a window. The control that catches card testing and repeated small charges that each pass the per-transaction limit.
And then

Budget thresholds

Not a limit at all. A threshold watches an aggregate after the fact and notifies somebody. Worth having, but it is reporting rather than control, and it is where the arithmetic above applies.

Distribute cards by recurring spending need rather than by job title. A field technician who buys parts weekly needs a working limit; a director who travels twice a year does not need a standing one. Titles are a poor proxy for spending patterns and produce both blocked purchases and idle exposure. There is more detail in our guide to setting corporate card spending limits.

Choosing

Which control do you actually need?

Spend happens on cards, on many small purchases

Buy issued cards with real limits. This is the one case where the category delivers exactly what it advertises: rules enforced at authorization, nothing to reconcile afterwards because the bad charge never happened. Ramp Free, Brex Essentials and BILL Spend and Expense are all $0 per user per month here, funded by interchange on the card.

Corporate card monitoring

Spend happens on cards you already have

Do not buy a control you cannot apply. If your spend is on existing Chase, Amex or Citi cards and you are not going to reissue them, no platform can decline anything for you. What you can improve is the delay: get the feed watched continuously so you find out in days rather than at month end, and route the alert to the budget owner rather than to a shared inbox.

Budget alerts

Spend is a handful of large commitments

Card controls are close to irrelevant. Your overruns come from signatures, so the control has to sit at the point of commitment: a request checked against remaining budget before the SOW or renewal is signed. That is a procurement or intake workflow, and the card platforms do not do it. Watch renewals as a separate calendar.

Software renewal management

Most companies of any size are in all three situations at once, in different departments, and that is the real reason spend control projects disappoint. A single platform is bought to solve the whole problem, it solves the first column properly, and the overspend that triggered the project in the first place was in the third. Before you shortlist anything, split last quarter's actuals by the six channels in the table above. The column that holds the most money tells you which control to buy first, and it is rarely the cards.

Where we fit

Spendnotify does not block spend, and you should know that before you read further

We issue no cards, so we cannot decline a transaction. Anything on this page about authorization-level controls describes what Ramp, Brex, BILL, Emburse and Navan do, not what we do. If a hard control is what you need, buy one of them; several are free.

What we do is the row those products are weakest on. Spendnotify connects read-only to the accounts, cards and vendors you already use, holds the budgets you already set by department or cost center, and watches them continuously. It runs the trend projection from the table above rather than a bare percentage threshold, so the budget owner hears about a 130 percent month in the first week rather than the third, and it messages the person who owns the budget instead of posting into a dashboard nobody opens. It sits alongside whatever cards and accounting system you already have, and it changes nothing about them.

It is honest about the last row too. A commitment nobody has recorded anywhere is invisible to us as well. What we can do is notice the first invoice against it on the day it posts rather than at close, which on the arithmetic above is the difference between reallocating inside the month and explaining a variance after it.

Questions buyers ask

Frequently asked questions

Which budget management tools allow automated alerts for procurement overspend?

They fall into two groups that alert at different moments. Procurement platforms such as Coupa, ProcureDesk, TeamProcure and Opstream check a purchase request against remaining budget and warn the requester and the approver before the order is released. Card and expense platforms such as Ramp, Brex, BILL Spend and Expense and Emburse alert on transactions once they are authorized. The gap that catches most teams is the signed commitment that has not been invoiced yet, which appears in neither stream.

How can I set department budgets and get alerts when spend is trending over budget inside the platform?

You need two things that are usually sold separately: a budget structure that mirrors your cost centers, and a trend alert rather than a percentage threshold. A threshold fires when a share of the budget is already gone. A trend compares spend so far against the share of the period elapsed and projects the finish, which is what lets it flag a department in the first week of the month instead of the third.

How can I get alerts when I am trending over budget before the month is over?

Use a run-rate projection rather than a threshold. On a $60,000 monthly budget spending $2,600 a day, the projection reaches $78,000, or 130 percent, after only five days of data. A 70 percent threshold on the same budget stays silent until day 17, and a hard stop at 100 percent arrives on day 24, by which point the money is spent and there is nothing left to reallocate.

Is there a way to get notified when I am close to spending too much in a category?

Yes, and how well it works depends entirely on where the spend lands. On cards a platform issued itself, category rules are enforced at authorization, so an out-of-category charge is declined outright and the cardholder is told immediately. On your existing bank cards the same spend can only be imported from a feed, and pending transactions are generally not imported, so the earliest realistic notice is 1 to 3 business days after the charge.

What is the difference between spend control and spend visibility?

A control changes what happens; visibility changes what you know. A spending limit that declines a card at authorization is a control, because the money never leaves. A dashboard, a report and an alert are visibility, because the transaction has already cleared by the time you read them. Most software sold as spend control software is visibility on five of the six channels a department actually spends through.

Can spend control software block a transaction before it happens?

Only on cards the platform issued. Ramp, Brex, BILL Spend and Expense, Emburse and Navan all evaluate spend limits, per-transaction limits and merchant category rules at the authorization message and decline anything that violates them. Nothing in the category can block a charge on a Chase or Amex card you already had, an employee paying out of pocket, or a statement of work somebody has already signed.

How much does spend control software cost?

Published rates in the adjacent card and expense category run from $0 to $15 per user per month. Ramp Free, Brex Essentials and BILL Spend and Expense are $0 because interchange on the card funds them. Expensify Collect is $5 per member, Emburse Spend $8 with a 15-user minimum, Brex Premium $12, and Ramp Plus $15 plus a platform fee it does not publish. Dedicated procurement platforms publish nothing at all and quote per deal. There is a full breakdown in our note on what spend management software costs.

What spending limits can you set on a corporate card?

Five limit types nest inside each other. The company credit line caps everything. An individual or cardholder limit caps one person. A per-transaction limit caps a single purchase. A merchant category restriction blocks whole classes of merchant. A velocity limit caps how often a card can be used in a window. A charge can sit comfortably inside the cardholder limit and still be declined by any of the other four.

Do you need spend control software if you already have accounting software?

Usually yes, because of timing rather than features. QuickBooks Online does hold budgets, but only on the Plus plan at $140 a month and Advanced at $340, checked on 8 September 2026, and a budget in the general ledger updates when transactions are recorded. Your ledger is an accurate picture of last month. A control has to act on this week.

Can you do spend control in a spreadsheet?

You can hold the budget in one and many companies should. What a spreadsheet cannot do is notice anything. It reports what somebody typed into it, on the day they typed it, so the run rate that would have flagged a 130 percent month on day 5 only exists if a human calculates it every morning. In practice that happens for two weeks after somebody builds the file, and then it does not.

Keep reading

The control fired on day 24. The decision was needed on day 5.

Whichever platform you buy, most of your spend will be reported to you rather than blocked for you. Spendnotify watches the budgets you already set, across the cards and accounts you already have, runs the projection instead of waiting for a percentage, and messages the budget owner while there is still a month left to fix. Read-only, and it changes nothing about your cards or your ledger.

Run the demo