Salesforce Data Cloud Rate Multipliers — Explained for 2026

A Salesforce Data Cloud sales-engineering deck typically frames consumption as "1 Credit equals roughly X events". That framing is misleading. Every event class has its own multiplier. The five main event classes have wildly different rates, and the dominant consumer differs between B2B and B2C programmes. See the broader Credit economics in our Data Cloud pricing guide.

Per Salesforce's official Data Cloud documentation, rate multipliers are published on the Salesforce documentation portal and updated each major release. The numbers below reflect 2026 published rates and the recent shifts; treat any pre-deal model as needing a sanity check against the current published table at signing.

The Master Rate Multiplier Table

Below is the consolidated view of 2026 published rate multipliers. Treat as indicative — Salesforce updates the table per major release and bundling tier.

The bold rows are where most enterprise programmes leak Credits. Streaming ingestion at 4×. Account-level B2B unification at 5×. Real-time activation at 4×. Multi-channel real-time at 6×+. Pre-deal modelling that does not weight these correctly will under-forecast Credit consumption by a factor of 2–3.

Ingestion Rate Multipliers

Ingestion is the first stage of any Data Cloud pipeline and the most easily over-spent. Salesforce supports three ingestion modes with materially different rate multipliers.

Batch ingestion (the cheap path)

Batch ingestion — CSV, scheduled DB pulls, S3 object reads — runs at roughly 1 Credit per 1,000 rows. For most non-real-time data sources, batch is the right choice. Transactional history, profile attributes, loyalty data, weekly sales feeds: all should be batch.

Streaming ingestion (3–5× cost)

Streaming ingestion via Kafka, webhook listeners, or always-on connectors runs at roughly 4 Credits per 1,000 rows. Use only where the downstream activation requires sub-minute freshness. Streaming a daily order feed is a common waste — daily orders almost never drive real-time downstream activation.

Real-time event API (5× cost)

Direct real-time event API ingestion runs at roughly 5 Credits per 1,000 events. Reserved for genuine real-time data (clickstream, IoT, mobile-app events) where the latency requirement is sub-second.

The single largest "easy win" in Credit reduction is converting non-time-sensitive sources from streaming to batch. A B2C client we worked with cut Credit consumption 38% in a single quarter by converting six of nine streaming connectors to nightly batch.

Profile Unification Rate Multipliers

Profile unification (identity resolution) consumes Credits per record resolved. The rate depends on the data class.

B2C profile resolution

Standard B2C identity resolution — matching email, phone, address across sources — runs at roughly 2 Credits per 1,000 records. The dominant driver is match-rule complexity. A simple email-only match is cheaper than a probabilistic match across 6 attributes.

B2B account / contact resolution

B2B identity resolution — matching contacts to accounts, resolving account hierarchies — runs at roughly 5 Credits per 1,000 records, more than double B2C. Account hierarchy depth and the number of subsidiary / parent relationships drive the rate higher.

B2B unification is the most common dominant Credit consumer in B2B programmes. If the programme is account-led, expect unification to be 40–60% of total Credit consumption.

Segmentation Rate Multipliers

Segmentation consumes Credits per record scanned, weighted by segment complexity and execution frequency.

Simple vs complex segments

A simple segment with three rules and no lookups runs at roughly 1 Credit per 1,000 records scanned. A complex segment with 4+ rules, joins, or nested lookups runs at roughly 3 Credits per 1,000 scanned. Pre-computing common segment building blocks reduces complex-segment cost.

Cadence is the multiplier on the multiplier

A weekly segment run consumes 1×; daily 7×; hourly 168×. The cadence question — "do we genuinely need daily, or does weekly carry the use case?" — is the single largest cost lever for segmentation.

Activation Rate Multipliers

Activation is the most heavily-multiplied event class — and the dominant consumer in B2C programmes.

Batch single-channel

Roughly 1 Credit per 1,000 audience members. The base rate; everything else compounds on this.

Real-time single-channel

Roughly 4 Credits per 1,000 audience members. The latency premium is 4×. Use only where downstream channel responsiveness requires it.

Multi-channel real-time

Roughly 6+ Credits per 1,000 audience members, scaling with channel count. A simultaneous push to email + paid social + web personalisation can run 8–10× the batch single-channel rate.

AI Inference Rate Multipliers

AI inference Credits scale per Einstein scoring event. Propensity scoring, content recommendation, send-time optimisation — each is one inference per scored record. Roughly 2 Credits per 1,000 inferences for standard Einstein scoring.

Agentforce action grounding consumes Credits separately, varying by the tier of Agentforce action. A simple Q&A grounding consumes less than a multi-step autonomous workflow grounding. See the Agentforce credit mechanics in our Editions That Include Data Cloud guide.

Five Tactics to Reduce Credit Consumption

1. Convert non-time-sensitive sources from streaming to batch. Typical 3–5× cost reduction on that ingestion surface. The largest single Credit-reduction lever for most programmes.
2. Reduce segmentation cadence where business value is marginal. Daily-to-weekly cuts cadence cost 6–7×. The question is "what business decision uses this segment?" — if no decision is made daily, the cadence doesn't need to be.
3. Consolidate activations. Three separate audience pushes can sometimes be a single union push. Each consolidated push reduces the per-activation Credit overhead.
4. Tune identity resolution match rules. Reduce match attributes where the precision lift is marginal. Avoid re-resolving the same records on every refresh — incremental resolution where supported.
5. Cache AI inferences where stable. Propensity scoring that is stable for 72 hours does not need re-scoring every hour. Cache, expire, refresh.

Our advisors handle Data Cloud Credit re-baselining inside Salesforce advisory engagements and as standalone consumption audits. See documented outcomes in our case studies.

What Buyers CAN Negotiate

The published rate multipliers themselves are usually not negotiable. They are Salesforce's standard rate card and apply equally across customers. What IS negotiable, and where buyer leverage matters:

Related Reading

• Pillar: Salesforce Data Cloud Pricing Guide
• Salesforce Editions That Include Data Cloud
• Data Cloud for Marketing Pricing
• Salesforce Contract Negotiation Guide
• Hidden Costs Comparison
• White Paper: SaaS True Cost

Frequently Asked Questions