Adobe 2Q26 Earnings Update: Not Out Of The Woods; Maintain BUY

Key Points

• Adobe beat 3Q26 expectations with revenue of US$6.76 billion (+13% YoY) and non-GAAP EPS of US$6.13, and raised FY2026 guidance. The stock still fell about 2.3% because the Q4 guide only matched consensus.
• About half of the 15% non-GAAP EPS growth came from share buybacks, with around US$2.2 billion repurchased in the quarter, so underlying earnings growth is weaker than the headline.
• AI-first ARR passed US$650 million (+150% YoY) but is still only about 2.4% of the US$27.5 billion ARR base, and Adobe has not said how many freemium users convert to paid.
• Total ARR growth slowed to 11.2% from 12.5% and RPO grew just 8%, while gross margin fell 60bps to about 88.7% as AI compute costs rise.
• At under 10x forward P/E we maintain BUY with a US$320 target price, though the margin of safety is now narrower.

Source: Adobe quarterly results. iFAST estimates.

AI monetisation: the milestone arrives, but the base stays small

AI-first ending ARR exceeded US$650 million, up more than 150% YoY, compared with the US$500 million reported in 2Q26. This represents a 30% sequential increase and continues the pattern of Adobe achieving its own AI targets ahead of schedule — the same dynamic that saw the previous US$250 million year-end target reached a full quarter early.

The supporting funnel metrics moved in the same direction:

Table 2: Key AI metrics

The 1 billion MAU milestone is the headline Adobe wanted, while the sequential doubling of Acrobat AI Assistant usage is the most encouraging single datapoint in the release — it is the product where AI sits closest to an existing paid workflow.

But the proportion has not changed materially since last quarter. AI-first ARR of US$650 million is roughly 2.4% of the US$27.50 billion ARR base, compared with about 1.8% in 2Q26. Even at 150% growth, the absolute contribution remains small relative to Adobe’s consolidated ARR base. At Adobe’s US$1 billion target for the year, AI-first ARR would still account for only around 3.6% of total ARR; whether this is sufficient to offset slower core growth will depend on the extent of that slowdown and how much AI revenue is incremental.

There is also a disclosure limitation that we flagged last quarter, and which persists. Adobe’s “AI-first” ARR only captures revenue from products that are explicitly sold as AI offerings. It does not capture the potentially larger benefit from AI features built into existing Creative Cloud plans, such as helping Adobe retain customers or encouraging them to upgrade. As a result, investors cannot clearly determine how much of Adobe’s AI growth is coming from new AI-related revenue versus AI helping to protect and expand its existing business. This lack of visibility is one reason the market remains cautious about the headline 150% growth figure.

Taken together, the 1 billion MAU milestone represents a user base expanding at more than 20% YoY, with creative freemium users growing around 70% YoY while contractually committed revenue (RPO) is growing at just 8% and declined sequentially. Adobe is acquiring users far faster than it is acquiring contractual revenue commitments. On the bright side, Adobe now has a large pool of freemium users to convert into paying customers; the tougher question is how effectively it can do so – a question Adobe’s disclosures have yet to answer clearly.

Figure 2: RPO vs ARR

The key risk, marked to market: the recurring revenue drag has arrived

In our previous update, we wrote that the freemium pivot “delays the point at which user growth translates into annual recurring revenue” and that we therefore expected “pressure to build in Adobe’s second-half ARR performance”.

Total ARR growth slowed to 11.2% from 12.5% in Q2, consistent with that concern, although the slowdown alone does not establish how much is attributable to the freemium pivot.

RPO of US$22.16 billion grew 8% YoY, against 13% revenue growth. The gap suggests slower growth in contracted revenue visibility, although RPO is also affected by contract duration and renewal timing. If this weakness persists, it could weigh on future revenue growth,  consistent with the 11–12% implied Q4 growth rate in the guidance.

The cost pressure arrives faster than the revenue. One hundred million creative freemium users, growing 70% YoY, consume compute regardless of whether they convert, which helps explain the pressure on gross margins.

This is because expanding the freemium base can increase compute costs before users convert, potentially adding to gross-margin pressure. Non-GAAP operating margin guidance of around 45% for the next quarter suggests that the cost is being absorbed for now, but sustaining margins as usage grows will depend on monetisation and compute efficiency.

As such, we argue that until Adobe discloses what proportion of the freemium base converts, and at what price point, the 100 million-user figure remains difficult to assess without clearer evidence of paid conversion and the cost of serving those users.

Figure 3: Gross margin declining

Leadership transition: the overhang we flagged is now resolved

The succession question that dominated our 2Q26 review now has an answer. Anil Chakravarthy will become CEO on 1 December 2026, the first day of FY2027, with Shantanu Narayen moving to Executive Chair after 29 years leading the company. Chakravarthy is currently President of Customer Experience Orchestration and was previously CEO of Informatica.

This removes one of the three factors we identified as weighing on the stock after Q2. The handover is relatively orderly: an internal candidate with segment P&L responsibility, a start date aligned with the fiscal year, and the outgoing CEO remaining on the board rather than exiting outright. The CFO seat, vacated by Dan Durn in June, remains the unfinished part of the leadership picture.

The substantive signal is Chakravarthy’s stated priority of agentic software as the next growth vector. This points towards Digital Experience and enterprise workflow orchestration rather than the Creative franchise, and towards consumption- or outcome-based revenue rather than seat expansion. His Informatica background is consistent with that emphasis and could imply a different capital allocation posture from the one that built Creative Cloud.

One practical consequence for the near term: FY2027 guidance, due with Q4 results in December, will be set by an incoming CEO in his first month. New chief executives often use the initial guidance cycle to establish a more conservative baseline. Investors should therefore be prepared for an initial FY27 guide below current Street expectations and should be careful not to interpret such a reset as fresh evidence of structural deceleration.

Not out of the woods; Maintain BUY

While the share price has increased by more than 20% since our last call, as of 18 September 2026, several issues remain. Reported revenue growth of 13% is respectable, but the 15% growth in non-GAAP EPS has been meaningfully supported by share buybacks. Moreover, three quarters into the pivot, Adobe has delivered strong user growth without disclosing the corresponding conversion rate. With RPO growing at 8% and implied Q4 revenue growth of 11–12%, the underlying trajectory remains weaker across the key operating indicators, particularly when excluding the benefit from buybacks.

Our previous positive view stemmed from a significant disconnect between Adobe’s share price and its underlying fundamentals, with the market pricing in a degree of terminal risk that we believed was unlikely to materialise within the next two years. In simple terms, we do not see much downside in earnings estimates for the next two years at the moment. With that in mind, at forward P/E of less than 10x is an attractive opportunity.

That view remains. However, following the share-price rally, the margin of safety is narrower than at our June call, while AI monetisation remains unproven. As such, we maintained our target price at USD 320 with a BUY call.

Figure 4: Adobe next quarter sales growth vs peers

Table 3: Valuation Table