Salesforce's AI Agent Business Is Growing Over 200%, But Will It Show Up in Its Order Book?

Salesforce’s Agentforce reached $1.2B in annual recurring revenue in the last quarter (+205% YoY) and has completed 3.8B tasks to date, but signed customer contracts have not accelerated beyond a steady pace over the past two quarters. Management said growth would accelerate nearly a year ago, yet bookings/contract-value momentum remains limited versus usage metrics. With shares up ~40% from the June low and trading at 15x forward earnings ahead of the quarterly report, investors will watch whether renewed order book/cRPO converts agent adoption into dollars within the next 12 months.

Analysis

CRM is in a classic transition-stock setup: the market is paying for AI optionality before the evidence shows up in contracted dollars. The key mechanism is whether agents expand net wallet share via premium tiers and consumption, or instead let customers buy fewer human seats and delay spend recognition; right now the second-order risk is that usage growth can coexist with flat-to-weak cRPO, which is exactly how a “platform story” turns into a multiple trap.

Near term, the stock is vulnerable to an earnings print that confirms adoption but not monetization. That would force investors to separate activity metrics from revenue durability and could compress the forward multiple back toward slower-growth SaaS peers; the real inflection window is 1-3 quarters for bookings optics and 6-18 months through renewal cycles, when seat rationalization can actually hit ARR. If AI is truly additive, you should see premium-seat penetration and contract value re-accelerate together; if not, the stock likely trades like a mature workflow vendor rather than a category-defining AI winner.

Contrarian take: consensus may be overestimating near-term cannibalization. Agents can increase total spend by adding consumption credits on top of seats, and hidden usage inside flat contracts means reported contract value can lag true adoption. The falsifier is simple: if cRPO and organic cRPO inflect higher on the next 1-2 prints, the bear case on AI-driven seat loss is too early; if they do not, the rally is likely on borrowed time.

Key Decisions for Investors

• Do not add to CRM into the print; treat it as a prove-it event and wait for cRPO/organic cRPO confirmation over the next 1-2 quarters before paying up for the AI narrative.
• If you need enterprise software exposure, favor MSFT or NOW over CRM: they have broader platform pricing power and are less exposed to a single-seat monetization transition.
• Pair trade idea: short CRM vs long IGV on any post-earnings pop if bookings/contract value fail to accelerate; the short thesis improves if the market starts pricing CRM like a legacy SaaS multiple rather than an AI compounder.
• For event-driven accounts, consider a short-dated downside hedge on CRM into Wednesday only if implied vol is not already extreme; the risk/reward is best if management again talks up adoption without a matching guide raise.
• Falsify the bearish thesis if premium-seat bookings keep compounding and cRPO turns up materially; at that point, cover any short CRM exposure quickly because the market will re-rate the name on evidence, not usage.