Do Agentforce and Data 360 Share One Flex Credit Pool With No Upfront Allocation?

The buyer question, stated cleanly for RevOps, GTM engineering, and revenue finance: do Agentforce and Data 360 draw from the same Flex Credit pool, and do we have to allocate credits upfront between them? The question matters because a pilot business case built on Agentforce action counts assumes the credits it forecast are still there when the agent runs.

The answer is yes, they share one currency, and no, you do not allocate upfront. Source: Salesforce Flexible Buying Models.

This page is a decoder, not a price list. Order form beats any blog math. Every figure below is labeled with its source.

What Salesforce actually said

The Flexible Buying Models FAQ answers the allocation question directly.

Do I have to decide how my Flex Credits will be used when I buy them? No. Flex Credits can be spent across Agentforce and Data 360 based on where you need them, with no upfront allocation required.Salesforce, Flexible Buying Models

• Flex Credits are available today for Agentforce and Data 360, and Salesforce positions them as spend flexibly across services. Source: Salesforce Flexible Buying Models.
• The Flex Credits Rate Card updated 08.31.2026 lists Agentforce and Data 360 usage types against the same credit currency, and includes sandbox multiplier notes for pre-production, with Data 360 sandbox not using tiered multipliers. Source: Flex Credits Rate Card 08.31.2026.
• Salesforce's illustrative Agentforce Flex Credit examples carry a disclaimer that they do not include other costs such as Data 360 credits or other consumption services. Source: Agentforce pricing.

Read those three together. The vendor tells you the pool is fungible, meters both product families against it, and then warns you that its own Agentforce example is not the whole bill.

Why this opens a Proof Gap

The Proof Gap is the distance between what a team claims an AI deployment cost or produced and what it can evidence at close. A shared credit pool widens it in a specific way: the Agentforce forecast is built as if the pool were private.

1. The pilot model counts agent actions and stops there, because that is the only number the Agentforce business case owner controls.
2. Data 360 work draws on the same credits. Unification, segmentation, activation, and query usage are metered usage types on the same rate card, and they run on a data team's schedule, not the agent rollout's schedule.
3. The two teams do not share a forecast. Nobody is wrong on their own line. The pool still empties earlier than either plan predicted.
4. At close, the Agentforce cost claim cannot be evidenced, because the invoice reflects a pool that two functions drew from without a split.

The failure is not fungibility. Fungibility is the feature buyers asked for. The failure is a single-product forecast run against a multi-product meter.

What Digital Wallet is for here

Digital Wallet gives visibility by usage type. It is a reporting surface, not automatic product fencing. It does not hold Agentforce credits back from Data 360, and it does not hold Data 360 credits back from Agentforce. If you need a boundary between the two, you build it operationally.

• Use it weekly, not at close. A monthly read tells you what already happened.
• Break burn by usage type. One org-level credit number cannot answer whether the agent rollout or the data build consumed the quarter.
• Pair it with a written threshold. Alerting is not stopping. That distinction is the whole point of Digital Wallet has no kill switch.

The decision rule and the operator action

Decision rule: do not scale a new Agentforce production cohort on a Flex Credits buy until Digital Wallet shows the Agentforce versus Data 360 usage-type split for the last full week, and both an Agentforce owner and a Data 360 owner are named on that pool.

1. Open Digital Wallet and export or screenshot usage by type for the last full week, Agentforce versus Data 360.
2. Set one shared Flex Credit budget line rather than two imaginary sub-pools, and write both owner names against it.
3. Give the Data 360 owner the same weekly review slot the Agentforce owner already has. A shared meter needs a shared review.
4. Re-run the Agentforce pilot cost model with Data 360 burn included, then take that number, not the original one, into the next commit conversation.
5. Put the split in the close pack. If the pack cannot separate the two, the pilot cost claim is unverifiable.

What this page is not

• It is not the arrears forecast page. That model lives at Flex Credits forecast and close.
• It is not the chargeback page. Splitting one org-level pool across business units is at BU chargeback.
• It is not the sandbox rate page. Pre-production multipliers are at the sandbox rate decoder.
• It is not a price list. Nothing here states a credit multiplier or dollar rate that Salesforce has not published, and your contracted rate governs.

Related: Flex Credits forecast and close · BU chargeback · Digital Wallet has no kill switch · Sandbox rate decoder · Who audits the agent meter · The Proof Gap