Salesforce 2Q FY27 Earnings: AI Traction Builds Despite Slower Organic Growth, Downgrade to HOLD

Salesforce delivered a strong 2Q FY27 result, with revenue rising 11% y/y to USD11.3 billion, cRPO growth accelerating to 14% in constant currency and EPS more than doubling to USD5.90. That said, the quality of the growth was more mixed than the headline suggests, with Informatica contributing USD440 million of revenue, leaving organic growth at around 6%, down from approximately 9% in 1Q. On top of that, exclude Anthropic investment gain, EPS was USD3.37, up 16% y/y.

Overall, post earnings, we note that Salesforce share price rose about 23% (47% since our last update), which we believe the rally was driven mainly by a rerating in SaaS and application software valuations after the company announcing the collaboration with Claude, where the Anthropic's decision to build Claudeforce on top of Salesforce, rather than launching a competing CRM product, which reduces concerns that frontier models will quickly displace the application layer.

On top of that, share price was further supported by continued grow in seats-based revenue across Sales, Service and Slack, with attrition remained near record lows, and Agentforce customers returned to purchase additional credits after their initial deployments.

However, the rally has not been accompanied by a comparable improvement in the earnings outlook. Organic growth continued to slow, Integration and Analytics contracted, and bottom line guidance remain unchanged despite top line growth guided slight expansion, contributed by Informatica. We raise our target price from USD260 to USD277 after rolling forward our valuation to FY29, but the share price recovery leaves only 6.5% upside. We therefore downgrade the rating from BUY to HOLD.

Figure 1: Salesforce’s share price movement.

Organic growth slowed despite better bookings

In 2Q FY27, subscription and support revenue increased 12% y/y to USD10.8 billion, but the acquired Informatica revenue accounted for much of the sequential improvement. Excluding the acquisition, organic revenue growth slowed to around 6% from approximately 9% in 1Q. This remained well below software peers, where ServiceNow's subscription revenue growth and SAP's cloud revenue growth of around 24% during the same period. We believe the growth differences largely reflect Salesforce’s more mature revenue base and dominant position in enterprise CRM, which limit its growth runway relative to smaller, faster growing peers.

The product mix was also uneven. Sales grew 9% in constant currency and Service grew 5%, with both continuing to add seats. Meanwhile, Integration and Analytics moved from 1% growth in 1Q to a -6% decline, while Marketing and Commerce continues to see deteriorating growth at -1% y/y. That said, Sales and Service continued to hold up, but the contraction in MuleSoft and Tableau leaves the recovery increasingly dependent on the data, platform and AI-related businesses.

Nevertheless, bookings were better than the revenue mix, with net-new AOV growth during 1H FY27 was the strongest in four years, attrition remained near record lows, and contract terms increased across new business and renewals. On top of that, cRPO growth accelerated to 14% in constant currency, while management guided the same growth for 3Q. As such, with a stronger near term contract visibility, we expect organic revenue growth to improve modestly in 2H, though weakness in Integration and Analytics and Marketing and Commerce should keep the recovery within the high single digit range.

That said, we note that the stronger cRPO did not translate into better long term visibility, as total RPO grew 11%, while noncurrent RPO declined sequentially from USD34.3 billion to USD32.8 billion despite the longer contract terms. Moving forward, as mentioned in the previous earnings results, we expect the gap between current and total RPO growth to remain as Flex Credits and usage based agreements account for a larger share of bookings. Unlike multi year seat commitments, future consumption and overages only enter revenue as customers use the platform, shortening the backlog even as near term bookings improve.

Figure 2: Salesforce's 2Q FY27 key financial performance

Table 1: Subscription & Support Revenue.

Source: Salesforce, iFAST compilations. Data as of 31 July 2026.

Figure 3: cRPO and RPO growth.

Agentforce usage is turning into repeat spending

Salesforce’s AI business continued to scale, with Agentforce ARR exceeding USD1.5 billion and combined Agentforce and Data 360 ARR reaching USD3.9 billion. However, the Agentforce figure now includes Slackbot and Headless 360, while the combined ARR also includes USD1.1 billion of Informatica cloud ARR, weakening the comparability with previous quarters. The stronger evidence came from actual usage, with customers consuming 3.2 billion Agentic Work Units during the quarter, up 97% q/q, while paying customers in production increased 70% q/q to more than 2,000.

Moreover, greater Agentforce adoption can be seen in the quarter, with half of Agentforce bookings came from customers purchasing additional credits after their initial deployments. Despite only 5% of Sales and Service knowledge workers are currently on premium editions (a 60% to 80% price premium SKU), management note that customers that have started their agentic transition are already spending around twice their previous AOV, while premium Agentforce bookings more than doubled q/q. However, with seat growth becoming more mature, we expect Agentforce to contribute mainly by increasing spending across Salesforce’s installed base, which should partially offset slower growth across the core CRM businesses.

Notably, Salesforce announced Claudeforce during the quarter, bringing Salesforce data, workflows and actions into Claude through 37 prebuilt sales skills. Overall, the product builds on Headless 360, which was introduced in the previous quarter to allow third-party AI agents to access Salesforce through MCP and APIs. Yet, the key difference is distribution: rather than merely opening Salesforce to external agents, Claudeforce places its capabilities directly within one of the leading enterprise AI interfaces. As Claudeforce will be available through Salesforce’s premium editions, adoption could also drive the same upgrade opportunity across Sales and Service.

Overall, the partnership shows that Anthropic is using Salesforce as the underlying enterprises data, workflow and governance layer rather than developing a competing CRM product, and we believe this reduces the risk that frontier models would bypass Salesforce entirely and reinforces its role as the system of record beneath external AI interfaces, explaining much of the broader SaaS and application software rerating following the result.

That said, customer AI investment ROI remain uncertain. And over the longer term, allowing third party models to access Salesforce data through headless platform initiatives may also weaken the company’s control over the user interface and shift part of the value towards model providers. The key question is whether wider distribution through Claude can generate sufficient premium upgrades and consumption revenue to offset the potential commoditisation of the application layer.

Figure 4: AWUs surged to 3.2b in 2Q FY27.

Consumption costs are now visible

While headline top-line growth remained steady, margin was the weaker part of the quarter, with operating margin edged down 20bps y/y to 34.1%, mainly as subscription and support cost of revenue increased 23%, more than twice the corresponding revenue growth. This compressed subscription gross margin by around 170bps, with higher hosting and generative-AI spending contributing to the increase. Unlike the original seat-based applications, Agentforce carries model, hosting and data-processing costs that rise alongside consumption, hence, while premium pricing and repeat credit purchases should support gross profit growth as usage scales, we believe they are unlikely to prevent some dilution to subscription gross margin.

That said, FY27 guidance remained at 34.3%, unchanged from the previous guidance, as Salesforce is reinvesting its internal AI productivity gains into product development and sales capacity, while higher consumption costs are beginning to weigh on gross margin, offsetting operating leverage came from lower G&A growth. We therefore expect operating margin to remain around the mid-30% level over the next two years, and any improvement would require Agentforce consumption and premium upgrades to scale materially faster than the associated delivery and go to market costs, leaving Salesforce’s Rule of 50 ambition increasingly dependent on stronger organic revenue growth.

On top of that, the debt funded accelerated share repurchase has raised the financing burden, with quarterly interest expense increased to USD473 million from USD67 million, absorbing part of the EPS benefit from the lower share count. While the buyback remains accretive to per-share earnings, it does not improve operating profit or cash generation, and the higher interest expense will become more visible once the FY27 strategic investment gains roll off.

Figure 5: Salesforce’s operating margin.

Figure 6: Salesforce’s operating cost.

Figure 7: Management guidance.

Valuation has caught up; downgrade to HOLD with USD277 target price.

Overall, the quarter eased concerns over AI displacement. Sales, Service and Slack continued to add seats, attrition remained near record lows, and Anthropic chose to bring Claude into Salesforce rather than launch a competing CRM product, suggesting enterprise applications are more resilient than the market previously feared and helped drive the post earnings recovery from the de-rating earlier this year.

That said, we see the recent rally as multiple rerating rather than improving in earnings outlook, with key questions of Agentforce ROI remain unclear, and longer term moat erosion in system of record remain. On top of that, we don’t see further re-rating potential if organic revenue growth remains in the 7-9% range. Therefore, despite Salesforce now trades at a material discount to the faster growing application software peers, we view the discount as justified while organic growth remains within the mid- to high single digit range. Applying a fair P/E of 15 times, we derived a target price of USD277. With 6.5% upside after the post-results rally, we downgraded our investment rating from BUY to HOLD.

Side note: we forecasted a low double digits revenue growth over the next 3 years, that included the Informatica contribution and should not be read as the same rate of organic expansion. Meanwhile, our FY27 EPS estimate of USD16.31 includes the strategic investment gains recognised during the year, while the subsequent year declining reflects the normalisation of those gains and the higher interest burden, rather than a decline in operating earnings.

Table 2: Valuation Summary.

	FY26	FY27 E	FY28 E	FY29 E
Revenue (USD million)	41,525	46,100	51,058	56,688
Growth %, y/y	9.6%	11.0%	10.8%	11.0%
EPS	12.52	16.31	15.96	18.46
Growth %, y/y	22.7%	30.3%	-2.1%	15.6%
P/E	20.45	15.94	16.29	14.09
Fair P/E				15
Upside Potential				6.5%
Target Price				277
Source: Bloomberg Finance L.P., iFAST compilations. Data as of 31 August 2026.

Investment risks

1. Slower AI monetisation. Agentforce usage and credit refills may take longer than expected to become material relative to Salesforce's existing revenue base.

2. Further weakness in the original applications. Continued contraction in Integration and Analytics or Marketing and Commerce could offset the contribution from Agentforce, Data 360 and Slack.

3. Higher consumption and financing costs. Faster Agentforce usage could place additional pressure on subscription gross margin, while the higher interest burden reduces the earnings benefit from the lower share count.