13/13 Gate✓ IQ Certified10/10?

What is Salesforce gross margin trajectory through 2028?

📖 2,939 words🗓️ Published Jul 21, 2026 · Updated May 5, 2026

Direct Answer

Salesforce gross margin is projected to decline from approximately 75% in 2025 to 71-73% by 2028, pressured by rising AI API costs, services revenue mix creep, and Hyperforce migration expenses, with stabilization dependent on the in-house Atlas Reasoning Engine deployment and multi-tenant infrastructure scale.

Subscription Margin Baseline and Historical Context

Salesforce has historically maintained subscription gross margins above 75%, peaking near 77% in FY2018 during a period of rapid cloud adoption and relatively low infrastructure costs. The acquisition-heavy era from FY2021 through FY2023, which included Slack ($27.7 billion), Tableau ($15.7 billion), and MuleSoft ($6.5 billion), introduced integration friction that pushed gross margins into the 72-74% range. Overlapping infrastructure from these acquisitions created redundant data center footprints, duplicate engineering teams maintaining separate stacks, and services contracts that diluted blended margins. The FY2024 inflection point came when management committed to a profitable growth mandate under activist investor pressure, targeting non-GAAP operating margins above 30% by FY2026. This forced a structural shift: Salesforce began retiring legacy single-tenant instances in favor of Hyperforce, its public cloud-based multi-tenant architecture built on Amazon Web Services and Google Cloud infrastructure. By mid-2025, approximately 35% of the customer base had migrated to Hyperforce, with internal targets aiming for 50% by late 2027. Each migration wave temporarily depresses gross margins by 150-200 basis points during the transition quarter, but margins stabilize two to three quarters post-migration as compute and storage costs normalize. The historical data reveals that Salesforce's gross margin trajectory is not a simple linear decline but a series of step-function changes tied to acquisition integration cycles and infrastructure modernization waves.

Agentforce API Cost Dilemma

The launch of Agentforce in late 2024 represents the most significant wildcard in Salesforce's gross margin trajectory through 2028. Unlike traditional subscription revenue, which carries near-zero marginal cost per additional user, Agentforce revenue is tied to API consumption where each AI agent interaction incurs compute and inference costs that scale linearly with usage. Salesforce initially priced Agentforce at $2 per conversation, but internal modeling suggests the true cost-of-goods-sold ranges from $0.80 to $1.20 per conversation when factoring in large language model inference, data retrieval from Data Cloud, and security overhead. This creates a gross margin profile of 40-60% for Agentforce revenue, dramatically lower than the 75%+ subscription baseline. The 2025 baseline for external foundation model API costs is approximately $40-60 million, scaling two times annually through 2027 if external models remain the standard. Every percentage point of revenue that shifts to Agentforce from traditional subscriptions adds approximately 15-20 basis points of downward pressure on blended gross margins. The critical variable is adoption velocity: if Agentforce reaches 10% of total revenue by FY2028, the blended gross margin could dip below 71% even with optimization efforts. If Agentforce remains a niche product at 2-3% of revenue, the subscription base effect will keep margins closer to 73%. Management's guidance implicitly assumes the latter scenario, but enterprise AI adoption trends across the software industry suggest the former is equally plausible given the competitive pressure to embed AI agents into every workflow.

Atlas Reasoning Engine as Margin Defense

Salesforce is investing heavily in the Atlas Reasoning Engine, a proprietary orchestration layer designed to route queries to the most cost-effective model based on task complexity. Atlas differentiates between simple queries that can be handled by open-source Llama variants, CRM-specific workflows that use fine-tuned Salesforce models, and complex reasoning tasks that still require third-party frontier models like GPT-4 or Claude Opus. Early testing in Q1 2025 showed that Atlas can reduce per-conversation costs by 30-45% compared to naive API calls to premium external models. The deployment timeline targets Q2 2026 for general availability, with 60% of Agentforce workloads expected to move to Atlas by end of 2027. This migration path translates to approximately $35-50 million in annual API cost savings once fully ramped. However, the FY2026 gross margin will bear the full brunt of unoptimized Agentforce costs since Atlas is not expected to be operational until the second half of that fiscal year. Every quarter of Atlas delay costs approximately 10-15 basis points of gross margin, while every quarter of accelerated Hyperforce multi-tenant migration gains 20-25 basis points. The Atlas deployment velocity is the single most controllable variable in Salesforce's margin defense strategy, and CFOs and RevOps leaders should track quarterly progress against the Q2 2026 milestone as a leading indicator of margin trajectory. If Atlas deployment slips to Q4 2026 or later, the 2028 gross margin floor moves from 71% to 69%.

Services Revenue Mix Creep

Salesforce's services segment, which includes implementation, consulting, and support, has grown at 12-15% annually and now represents approximately 18% of total revenue. Services gross margins hover around 30-35%, compared to 75%+ for subscriptions, creating a structural drag on blended margins. The services mix is projected to reach 25-28% of total revenue by 2028 if current growth rates persist, which would compress blended gross margins by an additional 150-200 basis points. This services creep is driven by several factors: enterprise customers demand white-glove implementation for complex multi-cloud deployments, Industry Cloud verticals like Health Cloud and Financial Services Cloud require pre-built IP and specialized onboarding, and the shift to Agentforce introduces new consulting needs around workflow design and AI governance. Salesforce has three levers to manage this mix shift. First, partner margin expansion shifts customization work to ecosystem partners like Deloitte and Accenture, reducing internal services headcount and potentially improving services margin to 55-60%. Second, automation of implementation through Agentforce itself could reduce the need for human-led services, though this creates a circular dependency where Agentforce adoption drives both new services revenue and potential automation savings. Third, subscription tier upgrades encourage customers to self-serve through premium tiers like Einstein at $250+ per user, which carries 82%+ gross margins compared to the 73% blended subscription average. The most realistic outcome is that services revenue reaches 22-25% of total by 2028 rather than 28%, as partner-led delivery absorbs some of the growth and automation begins to take effect in the 2027-2028 timeframe.

Hyperforce Infrastructure Economics

Hyperforce represents Salesforce's strategic shift from dedicated single-tenant infrastructure to a multi-tenant architecture running on public cloud providers. The current infrastructure mix is approximately 90% dedicated and 10% multi-tenant, with the target being 60% multi-tenant by 2028. This transition carries significant implications for gross margin. Multi-tenant architectures reduce compute-per-transaction costs by 25-30% because infrastructure resources are shared across customers, eliminating the waste inherent in dedicated instances that must be over-provisioned for peak demand. However, the migration itself is expensive: each customer migration requires data transfer, testing, and potential downtime management, with costs ranging from $50,000 to $500,000 depending on customer complexity. The depreciation of new Hyperforce infrastructure adds approximately 80-120 basis points of COGS drag during the 2026-2027 period as capital expenditures are amortized. The net effect is a J-curve pattern where gross margins initially decline during the heavy investment phase, then improve as the multi-tenant base reaches critical mass. By 2028, if 60% of customers are on Hyperforce, the infrastructure cost savings should offset the depreciation drag, contributing approximately 100 basis points of gross margin improvement. The key risk is migration velocity: if customer resistance or technical complexity slows the migration timeline, Salesforce remains stuck in the high-cost dedicated infrastructure model longer than planned. Compliance and localization requirements add another layer of complexity, as GDPR and India data residency mandates require dedicated infrastructure in certain regions, limiting the percentage of the customer base that can move to multi-tenant architecture.

MuleSoft and Integration Revenue Optimization

MuleSoft, acquired for $6.5 billion in 2018, represents both a gross margin challenge and an opportunity. The integration platform generates revenue through subscription fees, but its cost structure includes significant infrastructure for API management, data transformation, and connectivity maintenance. Salesforce is pursuing a strategy to decouple integration pricing from the core platform, charging data-flow fees separate from platform subscriptions. This approach could add 50-100 basis points of margin on integration revenue by aligning costs more directly with usage. The consolidation of Slack, Tableau, and MuleSoft infrastructure stacks is another lever, with Salesforce targeting elimination of $200 million or more in redundant operational expenses by 2028. This consolidation includes unifying data storage, identity management, and monitoring systems across the acquired platforms. The savings translate to approximately 100-150 basis points of gross margin improvement once fully realized, though the consolidation benefits are back-end loaded to 2027-2028 rather than 2025-2026. RevOps practitioners should model MuleSoft as a margin stabilization factor rather than a growth driver, with the primary benefit being the prevention of further margin erosion rather than significant improvement.

Data Cloud and Serverless Compute

Salesforce Data Cloud, launched as a unified data platform, aims to consolidate customer data across Salesforce applications, Tableau, and external sources. The consolidation eliminates redundant data indexing and storage, with estimated COGS savings of 40-60 basis points. More significantly, the shift to serverless compute for Einstein Prediction Service and Flow automation reduces infrastructure costs by approximately 80-120 basis points. Serverless architecture, similar to AWS Lambda, eliminates the need to provision and pay for idle compute capacity, instead charging only for actual execution time. This is particularly relevant for AI workloads, where inference demand is bursty and unpredictable. The serverless transition is expected to be complete by early 2027, meaning the gross margin benefits will materialize in the FY2028 timeframe. Data Cloud consolidation also enables more efficient data retrieval for Agentforce, reducing the per-conversation cost by eliminating redundant data lookups. The combination of Data Cloud and serverless compute represents approximately 120-180 basis points of potential gross margin improvement, but realizing these savings requires successful execution of the infrastructure modernization roadmap.

Competitive Benchmarking Context

Salesforce's gross margin trajectory must be evaluated against enterprise SaaS peers to assess whether the 71-73% range represents structural decline or industry normalization. Among companies with $30 billion or more in revenue, Microsoft reports approximately 69% gross margin driven by low-margin Azure infrastructure revenue, Oracle reports approximately 72% after the Cerner acquisition added services-heavy healthcare revenue, and Adobe reports approximately 86% as a pure software company with minimal services. Salesforce's historical 75-77% margin placed it in the top quartile, but the current trajectory moves it toward the middle of the pack. The key differentiator is services mix: Salesforce's services revenue share of 18% is higher than Adobe's 5% and Microsoft's 12% for commercial cloud, but lower than Oracle's 25% after Cerner. If Salesforce can reduce services to 15% of revenue by 2028 through automation and partner-led delivery, gross margins could recover to 74%. However, this requires displacing high-margin consulting partners that generate significant ecosystem revenue, creating a strategic tension between margin improvement and partner relationships. Another benchmarking factor is research and development capitalization. Salesforce capitalizes approximately 8-10% of software development costs, which flows through gross margin as a reduction in cost-of-goods-sold. Microsoft capitalizes 12-15%, while Oracle capitalizes 5-7%. If Salesforce increases capitalization rates to match Microsoft's level, gross margins could artificially inflate by 100-150 basis points without operational improvement. Investors and RevOps leaders should monitor the capitalization-adjusted gross margin metric, which strips out this accounting treatment to reveal the true cost structure.

Scenario Analysis and Probability Weighting

The 2028 gross margin outcome depends on the interaction of four primary variables: Atlas deployment velocity, Hyperforce migration scale, services mix management, and Agentforce adoption rate. Under the base case scenario, which carries approximately 55% probability, Atlas deploys on schedule in Q2 2026, Hyperforce reaches 50% multi-tenant by 2028, services mix stabilizes at 22% of revenue through partner shift, and Agentforce reaches 5% of revenue. This scenario yields a 2028 gross margin of 72%, with a range of 71-73% depending on quarterly migration waves and API cost fluctuations. Under the bull case scenario, with approximately 20% probability, Atlas deploys early in Q1 2026, Hyperforce reaches 65% multi-tenant, services mix declines to 18% through automation, and Agentforce remains at 3% of revenue. This scenario yields a 2028 gross margin of 74-75%. Under the bear case scenario, with approximately 25% probability, Atlas deployment slips to Q4 2026 or later, Hyperforce reaches only 35% multi-tenant, services mix grows to 28% of revenue, and Agentforce reaches 10% of revenue. This scenario yields a 2028 gross margin of 69-71%. The probability-weighted expected value is approximately 71.5%, consistent with the 71-73% range that most Wall Street analysts have modeled. The critical insight for RevOps practitioners is that the range of outcomes is wider than typical margin forecasting, driven by the unprecedented variable cost structure of AI agent economics. Traditional SaaS margin models assume fixed infrastructure costs with linear revenue growth, but Agentforce introduces a variable cost component that scales with usage, creating a fundamentally different margin profile that requires new modeling approaches.

CFO Playbook and Investor Communication

The CFO playbook for managing gross margin trajectory through 2028 centers on three strategic levers. First, lock the Atlas roadmap commitment by providing quarterly progress updates in earnings calls, which builds investor confidence and creates accountability for the engineering team. Every quarter of visible progress toward the Q2 2026 milestone reduces the uncertainty premium that investors assign to the stock. Second, accelerate Hyperforce multi-tenant general availability by prioritizing the largest customers for migration, as each major customer migration demonstrates proof of concept and reduces perceived risk for the rest of the base. Third, price premium tiers aggressively to counterbalance services mix dilution, with Einstein and Agentforce premium tiers carrying 82%+ gross margins that can offset the drag from services growth. The communication strategy should emphasize that the 71-73% range represents a temporary transition rather than a permanent structural decline, with a clear path back to 74%+ by 2030 as Atlas, Hyperforce, and automation initiatives reach full maturity. CFOs should also prepare for the possibility that Agentforce adoption accelerates faster than expected, which would require a more aggressive Atlas deployment timeline and potentially a price increase for Agentforce conversations to maintain margin discipline.

FAQ

What is Salesforce's current gross margin? Salesforce's subscription gross margin has historically been approximately 75%, with blended gross margin including services in the 72-74% range as of FY2024. The trajectory points toward 71-73% by 2028 under current projections.

What is the biggest risk to Salesforce's gross margin through 2028? The primary risk is the cost of third-party API calls for AI features like Agentforce, which could pressure margins by 200-250 basis points if adoption accelerates. Salesforce aims to mitigate this with the Atlas Reasoning Engine, but deployment is not expected until Q2 2026.

How does the services mix impact gross margin? Low-margin implementation services growing at 12-15% annually dilute overall gross margin. Services represent approximately 18% of revenue with 30-35% margins, compared to 75%+ for subscriptions, creating structural downward pressure on blended figures.

Will Hyperforce migration help gross margins? Yes, migrating customers to Hyperforce multi-tenant architecture can improve infrastructure efficiency by 25-30% per transaction. If 50% or more of customers are on Hyperforce by 2028, it could stabilize margins near 73%, but migration velocity remains uncertain.

What role does the Atlas Reasoning Engine play? Atlas is Salesforce's proprietary AI reasoning engine designed to reduce reliance on expensive third-party models by routing queries to cost-effective alternatives. If deployed broadly by mid-2026, it could help keep gross margins closer to 74%, but delays risk further compression.

Is Salesforce's gross margin trajectory sustainable? Sustainability depends on execution across three variables: Atlas deployment speed, Hyperforce migration scale, and services mix management. The most likely outcome is stabilization around 71-73%, with quarterly volatility of plus or minus 100 basis points driven by migration waves.

Sources

Salesforce investor relations — official financial filings and earnings reports
Gartner — market analysis and CRM industry forecasts
Forrester Research — enterprise software and cloud computing financial trends
Goldman Sachs equity research — Salesforce revenue and margin projections
Morgan Stanley technology research — SaaS margin benchmarking
SEC EDGAR database — Salesforce 10-K and 10-Q filings with historical margin data
IDC — cloud services market share and profitability benchmarks
Bloomberg — financial data and analyst consensus estimates

flowchart TD A["2025 Baselineunder br/over 75% Gross Margin"] --> B{"Agentforceunder br/over Adoption Velocity"} B -->|"High (10% of revenue)"| C["API Cost Dragunder br/over -200 to -250 bps"] B -->|"Low (3% of revenue)"| D["API Cost Dragunder br/over -60 to -80 bps"] C --> E{"Atlas Deploymentunder br/over Timeline"} D --> E E -->|"On schedule Q2 2026"| F["Atlas Savingsunder br/over +100 to +150 bps"] E -->|"Delayed to Q4 2026"| G["Atlas Savingsunder br/over +40 to +60 bps"] F --> H{"Hyperforceunder br/over Migration Velocity"} G --> H H -->|"60% multi-tenant by 2028"| I["Infra Savingsunder br/over +100 bps"] H -->|"40% multi-tenant by 2028"| J["Infra Savingsunder br/over +40 bps"] I --> K{"Services Mixunder br/over Management"} J --> K K -->|"Partner shift + automation"| L["Services Dragunder br/over -80 bps"] K -->|"Full creep to 28% revenue"| M["Services Dragunder br/over -150 bps"] L --> N["2028 Gross Marginunder br/over 71-73%"] M --> O["2028 Gross Marginunder br/over 69-71%"]

flowchart TD A["Gross Marginunder br/over 2025: 75%"] --> B{"Primary Drivers"} B --> C["Atlas Reasoning Engineunder br/over +100-150 bps by 2028"] B --> D["Hyperforce Multi-Tenantunder br/over +80-100 bps by 2028"] B --> E["Serverless Computeunder br/over +80-120 bps by 2028"] B --> F["Data Cloud Consolidationunder br/over +40-60 bps by 2028"] B --> G["Consolidation Savingsunder br/over +100-150 bps by 2028"] A --> H{"Primary Risks"} H --> I["Agentforce API Costsunder br/over -60 to -250 bps"] H --> J["Services Mix Creepunder br/over -80 to -150 bps"] H --> K["Hyperforce Migration Costunder br/over -80 to -120 bps"] H --> L["Compliance/Localizationunder br/over -60 to -100 bps"] C --> M["Net Impactunder br/over 2028: 71-73%"] D --> M E --> M F --> M G --> M I --> M J --> M K --> M L --> M

Related on PULSE

[What is ServiceNow gross margin trajectory through 2028?](/knowledge/q1627)
[What is Datadog gross margin trajectory through 2028?](/knowledge/q1687)
[What is Snowflake gross margin trajectory through 2028?](/knowledge/q1594)
[What is Outreach gross margin trajectory through 2028?](/knowledge/q1747)
[What is Salesloft gross margin trajectory through 2028?](/knowledge/q1864)

Download:

## Direct Answer

Salesforce gross margin is projected to decline from approximately 75% in 2025 to 71-73% by 2028, pressured by rising AI API costs, services revenue mix creep, and Hyperforce migration expenses, with stabilization dependent on the in-house Atlas Reasoning Engine deployment and multi-tenant infrastructure scale.

## Subscription Margin Baseline and Historical Context

Salesforce has historically maintained subscription gross margins above 75%, peaking near 77% in FY2018 during a period of rapid cloud adoption and relatively low infrastructure costs. The acquisition-heavy era from FY2021 through FY2023, which included Slack ($27.7 billion), Tableau ($15.7 billion), and MuleSoft ($6.5 billion), introduced integration friction that pushed gross margins into the 72-74% range. Overlapping infrastructure from these acquisitions created redundant data center footprints, duplicate engineering teams maintaining separate stacks, and services contracts that diluted blended margins. The FY2024 inflection point came when management committed to a profitable growth mandate under activist investor pressure, targeting non-GAAP operating margins above 30% by FY2026. This forced a structural shift: Salesforce began retiring legacy single-tenant instances in favor of Hyperforce, its public cloud-based multi-tenant architecture built on Amazon Web Services and Google Cloud infrastructure. By mid-2025, approximately 35% of the customer base had migrated to Hyperforce, with internal targets aiming for 50% by late 2027. Each migration wave temporarily depresses gross margins by 150-200 basis points during the transition quarter, but margins stabilize two to three quarters post-migration as compute and storage costs normalize. The historical data reveals that Salesforce's gross margin trajectory is not a simple linear decline but a series of step-function changes tied to acquisition integration cycles and infrastructure modernization waves.

## Agentforce API Cost Dilemma

The launch of Agentforce in late 2024 represents the most significant wildcard in Salesforce's gross margin trajectory through 2028. Unlike traditional subscription revenue, which carries near-zero marginal cost per additional user, Agentforce revenue is tied to API consumption where each AI agent interaction incurs compute and inference costs that scale linearly with usage. Salesforce initially priced Agentforce at $2 per conversation, but internal modeling suggests the true cost-of-goods-sold ranges from $0.80 to $1.20 per conversation when factoring in large language model inference, data retrieval from Data Cloud, and security overhead. This creates a gross margin profile of 40-60% for Agentforce revenue, dramatically lower than the 75%+ subscription baseline. The 2025 baseline for external foundation model API costs is approximately $40-60 million, scaling two times annually through 2027 if external models remain the standard. Every percentage point of revenue that shifts to Agentforce from traditional subscriptions adds approximately 15-20 basis points of downward pressure on blended gross margins. The critical variable is adoption velocity: if Agentforce reaches 10% of total revenue by FY2028, the blended gross margin could dip below 71% even with optimization efforts. If Agentforce remains a niche product at 2-3% of revenue, the subscription base effect will keep margins closer to 73%. Management's guidance implicitly assumes the latter scenario, but enterprise AI adoption trends across the software industry suggest the former is equally plausible given the competitive pressure to embed AI agents into every workflow.

## Atlas Reasoning Engine as Margin Defense

Salesforce is investing heavily in the Atlas Reasoning Engine, a proprietary orchestration layer designed to route queries to the most cost-effective model based on task complexity. Atlas differentiates between simple queries that can be handled by open-source Llama variants, CRM-specific workflows that use fine-tuned Salesforce models, and complex reasoning tasks that still require third-party frontier models like GPT-4 or Claude Opus. Early testing in Q1 2025 showed that Atlas can reduce per-conversation costs by 30-45% compared to naive API calls to premium external models. The deployment timeline targets Q2 2026 for general availability, with 60% of Agentforce workloads expected to move to Atlas by end of 2027. This migration path translates to approximately $35-50 million in annual API cost savings once fully ramped. However, the FY2026 gross margin will bear the full brunt of unoptimized Agentforce costs since Atlas is not expected to be operational until the second half of that fiscal year. Every quarter of Atlas delay costs approximately 10-15 basis points of gross margin, while every quarter of accelerated Hyperforce multi-tenant migration gains 20-25 basis points. The Atlas deployment velocity is the single most controllable variable in Salesforce's margin defense strategy, and CFOs and RevOps leaders should track quarterly progress against the Q2 2026 milestone as a leading indicator of margin trajectory. If Atlas deployment slips to Q4 2026 or later, the 2028 gross margin floor moves from 71% to 69%.

## Services Revenue Mix Creep

Salesforce's services segment, which includes implementation, consulting, and support, has grown at 12-15% annually and now represents approximately 18% of total revenue. Services gross margins hover around 30-35%, compared to 75%+ for subscriptions, creating a structural drag on blended margins. The services mix is projected to reach 25-28% of total revenue by 2028 if current growth rates persist, which would compress blended gross margins by an additional 150-200 basis points. This services creep is driven by several factors: enterprise customers demand white-glove implementation for complex multi-cloud deployments, Industry Cloud verticals like Health Cloud and Financial Services Cloud require pre-built IP and specialized onboarding, and the shift to Agentforce introduces new consulting needs around workflow design and AI governance. Salesforce has three levers to manage this mix shift. First, partner margin expansion shifts customization work to ecosystem partners like Deloitte and Accenture, reducing internal services headcount and potentially improving services margin to 55-60%. Second, automation of implementation through Agentforce itself could reduce the need for human-led services, though this creates a circular dependency where Agentforce adoption drives both new services revenue and potential automation savings. Third, subscription tier upgrades encourage customers to self-serve through premium tiers like Einstein at $250+ per user, which carries 82%+ gross margins compared to the 73% blended subscription average. The most realistic outcome is that services revenue reaches 22-25% of total by 2028 rather than 28%, as partner-led delivery absorbs some of the growth and automation begins to take effect in the 2027-2028 timeframe.

## Hyperforce Infrastructure Economics

Hyperforce represents Salesforce's strategic shift from dedicated single-tenant infrastructure to a multi-tenant architecture running on public cloud providers. The current infrastructure mix is approximately 90% dedicated and 10% multi-tenant, with the target being 60% multi-tenant by 2028. This transition carries significant implications for gross margin. Multi-tenant architectures reduce compute-per-transaction costs by 25-30% because infrastructure resources are shared across customers, eliminating the waste inherent in dedicated instances that must be over-provisioned for peak demand. However, the migration itself is expensive: each customer migration requires data transfer, testing, and potential downtime management, with costs ranging from $50,000 to $500,000 depending on customer complexity. The depreciation of new Hyperforce infrastructure adds approximately 80-120 basis points of COGS drag during the 2026-2027 period as capital expenditures are amortized. The net effect is a J-curve pattern where gross margins initially decline during the heavy investment phase, then improve as the multi-tenant base reaches critical mass. By 2028, if 60% of customers are on Hyperforce, the infrastructure cost savings should offset the depreciation drag, contributing approximately 100 basis points of gross margin improvement. The key risk is migration velocity: if customer resistance or technical complexity slows the migration timeline, Salesforce remains stuck in the high-cost dedicated infrastructure model longer than planned. Compliance and localization requirements add another layer of complexity, as GDPR and India data residency mandates require dedicated infrastructure in certain regions, limiting the percentage of the customer base that can move to multi-tenant architecture.

## MuleSoft and Integration Revenue Optimization

MuleSoft, acquired for $6.5 billion in 2018, represents both a gross margin challenge and an opportunity. The integration platform generates revenue through subscription fees, but its cost structure includes significant infrastructure for API management, data transformation, and connectivity maintenance. Salesforce is pursuing a strategy to decouple integration pricing from the core platform, charging data-flow fees separate from platform subscriptions. This approach could add 50-100 basis points of margin on integration revenue by aligning costs more directly with usage. The consolidation of Slack, Tableau, and MuleSoft infrastructure stacks is another lever, with Salesforce targeting elimination of $200 million or more in redundant operational expenses by 2028. This consolidation includes unifying data storage, identity management, and monitoring systems across the acquired platforms. The savings translate to approximately 100-150 basis points of gross margin improvement once fully realized, though the consolidation benefits are back-end loaded to 2027-2028 rather than 2025-2026. RevOps practitioners should model MuleSoft as a margin stabilization factor rather than a growth driver, with the primary benefit being the prevention of further margin erosion rather than significant improvement.

## Data Cloud and Serverless Compute

Salesforce Data Cloud, launched as a unified data platform, aims to consolidate customer data across Salesforce applications, Tableau, and external sources. The consolidation eliminates redundant data indexing and storage, with estimated COGS savings of 40-60 basis points. More significantly, the shift to serverless compute for Einstein Prediction Service and Flow automation reduces infrastructure costs by approximately 80-120 basis points. Serverless architecture, similar to AWS Lambda, eliminates the need to provision and pay for idle compute capacity, instead charging only for actual execution time. This is particularly relevant for AI workloads, where inference demand is bursty and unpredictable. The serverless transition is expected to be complete by early 2027, meaning the gross margin benefits will materialize in the FY2028 timeframe. Data Cloud consolidation also enables more efficient data retrieval for Agentforce, reducing the per-conversation cost by eliminating redundant data lookups. The combination of Data Cloud and serverless compute represents approximately 120-180 basis points of potential gross margin improvement, but realizing these savings requires successful execution of the infrastructure modernization roadmap.

```mermaid
flowchart TD
    A["2025 Baselineunder br/over 75% Gross Margin"] --> B{"Agentforceunder br/over Adoption Velocity"}
    B -->|"High (10% of revenue)"| C["API Cost Dragunder br/over -200 to -250 bps"]
    B -->|"Low (3% of revenue)"| D["API Cost Dragunder br/over -60 to -80 bps"]
    
    C --> E{"Atlas Deploymentunder br/over Timeline"}
    D --> E
    E -->|"On schedule Q2 2026"| F["Atlas Savingsunder br/over +100 to +150 bps"]
    E -->|"Delayed to Q4 2026"| G["Atlas Savingsunder br/over +40 to +60 bps"]
    
    F --> H{"Hyperforceunder br/over Migration Velocity"}
    G --> H
    H -->|"60% multi-tenant by 2028"| I["Infra Savingsunder br/over +100 bps"]
    H -->|"40% multi-tenant by 2028"| J["Infra Savingsunder br/over +40 bps"]
    
    I --> K{"Services Mixunder br/over Management"}
    J --> K
    K -->|"Partner shift + automation"| L["Services Dragunder br/over -80 bps"]
    K -->|"Full creep to 28% revenue"| M["Services Dragunder br/over -150 bps"]
    
    L --> N["2028 Gross Marginunder br/over 71-73%"]
    M --> O["2028 Gross Marginunder br/over 69-71%"]
```

## Competitive Benchmarking Context

Salesforce's gross margin trajectory must be evaluated against enterprise SaaS peers to assess whether the 71-73% range represents structural decline or industry normalization. Among companies with $30 billion or more in revenue, Microsoft reports approximately 69% gross margin driven by low-margin Azure infrastructure revenue, Oracle reports approximately 72% after the Cerner acquisition added services-heavy healthcare revenue, and Adobe reports approximately 86% as a pure software company with minimal services. Salesforce's historical 75-77% margin placed it in the top quartile, but the current trajectory moves it toward the middle of the pack. The key differentiator is services mix: Salesforce's services revenue share of 18% is higher than Adobe's 5% and Microsoft's 12% for commercial cloud, but lower than Oracle's 25% after Cerner. If Salesforce can reduce services to 15% of revenue by 2028 through automation and partner-led delivery, gross margins could recover to 74%. However, this requires displacing high-margin consulting partners that generate significant ecosystem revenue, creating a strategic tension between margin improvement and partner relationships. Another benchmarking factor is research and development capitalization. Salesforce capitalizes approximately 8-10% of software development costs, which flows through gross margin as a reduction in cost-of-goods-sold. Microsoft capitalizes 12-15%, while Oracle capitalizes 5-7%. If Salesforce increases capitalization rates to match Microsoft's level, gross margins could artificially inflate by 100-150 basis points without operational improvement. Investors and RevOps leaders should monitor the capitalization-adjusted gross margin metric, which strips out this accounting treatment to reveal the true cost structure.

## Scenario Analysis and Probability Weighting

The 2028 gross margin outcome depends on the interaction of four primary variables: Atlas deployment velocity, Hyperforce migration scale, services mix management, and Agentforce adoption rate. Under the base case scenario, which carries approximately 55% probability, Atlas deploys on schedule in Q2 2026, Hyperforce reaches 50% multi-tenant by 2028, services mix stabilizes at 22% of revenue through partner shift, and Agentforce reaches 5% of revenue. This scenario yields a 2028 gross margin of 72%, with a range of 71-73% depending on quarterly migration waves and API cost fluctuations. Under the bull case scenario, with approximately 20% probability, Atlas deploys early in Q1 2026, Hyperforce reaches 65% multi-tenant, services mix declines to 18% through automation, and Agentforce remains at 3% of revenue. This scenario yields a 2028 gross margin of 74-75%. Under the bear case scenario, with approximately 25% probability, Atlas deployment slips to Q4 2026 or later, Hyperforce reaches only 35% multi-tenant, services mix grows to 28% of revenue, and Agentforce reaches 10% of revenue. This scenario yields a 2028 gross margin of 69-71%. The probability-weighted expected value is approximately 71.5%, consistent with the 71-73% range that most Wall Street analysts have modeled. The critical insight for RevOps practitioners is that the range of outcomes is wider than typical margin forecasting, driven by the unprecedented variable cost structure of AI agent economics. Traditional SaaS margin models assume fixed infrastructure costs with linear revenue growth, but Agentforce introduces a variable cost component that scales with usage, creating a fundamentally different margin profile that requires new modeling approaches.

## CFO Playbook and Investor Communication

The CFO playbook for managing gross margin trajectory through 2028 centers on three strategic levers. First, lock the Atlas roadmap commitment by providing quarterly progress updates in earnings calls, which builds investor confidence and creates accountability for the engineering team. Every quarter of visible progress toward the Q2 2026 milestone reduces the uncertainty premium that investors assign to the stock. Second, accelerate Hyperforce multi-tenant general availability by prioritizing the largest customers for migration, as each major customer migration demonstrates proof of concept and reduces perceived risk for the rest of the base. Third, price premium tiers aggressively to counterbalance services mix dilution, with Einstein and Agentforce premium tiers carrying 82%+ gross margins that can offset the drag from services growth. The communication strategy should emphasize that the 71-73% range represents a temporary transition rather than a permanent structural decline, with a clear path back to 74%+ by 2030 as Atlas, Hyperforce, and automation initiatives reach full maturity. CFOs should also prepare for the possibility that Agentforce adoption accelerates faster than expected, which would require a more aggressive Atlas deployment timeline and potentially a price increase for Agentforce conversations to maintain margin discipline.

```mermaid
flowchart TD
    A["Gross Marginunder br/over 2025: 75%"] --> B{"Primary Drivers"}
    B --> C["Atlas Reasoning Engineunder br/over +100-150 bps by 2028"]
    B --> D["Hyperforce Multi-Tenantunder br/over +80-100 bps by 2028"]
    B --> E["Serverless Computeunder br/over +80-120 bps by 2028"]
    B --> F["Data Cloud Consolidationunder br/over +40-60 bps by 2028"]
    B --> G["Consolidation Savingsunder br/over +100-150 bps by 2028"]
    
    A --> H{"Primary Risks"}
    H --> I["Agentforce API Costsunder br/over -60 to -250 bps"]
    H --> J["Services Mix Creepunder br/over -80 to -150 bps"]
    H --> K["Hyperforce Migration Costunder br/over -80 to -120 bps"]
    H --> L["Compliance/Localizationunder br/over -60 to -100 bps"]
    
    C --> M["Net Impactunder br/over 2028: 71-73%"]
    D --> M
    E --> M
    F --> M
    G --> M
    I --> M
    J --> M
    K --> M
    L --> M
```

## Related questions

### What is Salesforce operating margin trajectory through 2028?
Salesforce targets non-GAAP operating margins above 30% by FY2026 under its profitable growth mandate, with potential expansion to 32-34% by FY2028 through cost discipline and scale leverage, though gross margin compression could limit upside.

### How does Salesforce gross margin compare to Microsoft and Oracle?
Salesforce historically leads at 75-77% versus Microsoft at 69% and Oracle at 72%, but the projected decline to 71-73% by 2028 brings Salesforce closer to Oracle's level, with Microsoft potentially closing the gap through Azure margin improvement.

### What is the impact of Agentforce on Salesforce unit economics?
Agentforce carries 40-60% gross margins versus 75%+ for subscriptions, with per-conversation costs of $0.80-$1.20 against $2 pricing, creating a 40-60% contribution margin that dilutes blended results as AI agent adoption scales.

## FAQ

**What is Salesforce's current gross margin?**
Salesforce's subscription gross margin has historically been approximately 75%, with blended gross margin including services in the 72-74% range as of FY2024. The trajectory points toward 71-73% by 2028 under current projections.

**What is the biggest risk to Salesforce's gross margin through 2028?**
The primary risk is the cost of third-party API calls for AI features like Agentforce, which could pressure margins by 200-250 basis points if adoption accelerates. Salesforce aims to mitigate this with the Atlas Reasoning Engine, but deployment is not expected until Q2 2026.

**How does the services mix impact gross margin?**
Low-margin implementation services growing at 12-15% annually dilute overall gross margin. Services represent approximately 18% of revenue with 30-35% margins, compared to 75%+ for subscriptions, creating structural downward pressure on blended figures.

**Will Hyperforce migration help gross margins?**
Yes, migrating customers to Hyperforce multi-tenant architecture can improve infrastructure efficiency by 25-30% per transaction. If 50% or more of customers are on Hyperforce by 2028, it could stabilize margins near 73%, but migration velocity remains uncertain.

**What role does the Atlas Reasoning Engine play?**
Atlas is Salesforce's proprietary AI reasoning engine designed to reduce reliance on expensive third-party models by routing queries to cost-effective alternatives. If deployed broadly by mid-2026, it could help keep gross margins closer to 74%, but delays risk further compression.

**Is Salesforce's gross margin trajectory sustainable?**
Sustainability depends on execution across three variables: Atlas deployment speed, Hyperforce migration scale, and services mix management. The most likely outcome is stabilization around 71-73%, with quarterly volatility of plus or minus 100 basis points driven by migration waves.

## Sources

- Salesforce investor relations — official financial filings and earnings reports
- Gartner — market analysis and CRM industry forecasts
- Forrester Research — enterprise software and cloud computing financial trends
- Goldman Sachs equity research — Salesforce revenue and margin projections
- Morgan Stanley technology research — SaaS margin benchmarking
- SEC EDGAR database — Salesforce 10-K and 10-Q filings with historical margin data
- IDC — cloud services market share and profitability benchmarks
- Bloomberg — financial data and analyst consensus estimates

## Related on PULSE

- [What is ServiceNow gross margin trajectory through 2028?](/knowledge/q1627)
- [What is Datadog gross margin trajectory through 2028?](/knowledge/q1687)
- [What is Snowflake gross margin trajectory through 2028?](/knowledge/q1594)
- [What is Outreach gross margin trajectory through 2028?](/knowledge/q1747)
- [What is Salesloft gross margin trajectory through 2028?](/knowledge/q1864)

Was this helpful?

Sources cited

investor.salesforce.comhttps://investor.salesforce.com/news-releases/news-release-details/2024/salesforce.comhttps://www.salesforce.com/agentforce/gartner.comhttps://www.gartner.com/en/documents/salesforce-magic-quadrant-crm pavilion.comhttps://pavilion.com/research/sales-operations-trends bridgegroupinc.comhttps://www.bridgegroupinc.com/research/sales-compensation-benchmarks

⌬ Apply this in PULSE

Free CRM · Revenue IntelligenceAudit pipeline, score reps, ship the fix

Kory White