Bain & Company projects a significant US$100 billion market opportunity in the United States for software-as-a-service (SaaS) companies leveraging agentic artificial intelligence (AI). This burgeoning market is primarily driven by the automation of complex coordination tasks within enterprise systems.
This valuation emerges from the second installment of Bain’s five-part research series focused on the evolving software landscape in the age of AI. The report delves into the potential for agentic AI to unlock new software market segments and outlines strategic approaches for SaaS providers to capitalize on these opportunities.
The Crucial Role of Coordination in Enterprise Workflows
Bain identifies the core of this market in the often manual, labor-intensive processes that employees undertake to bridge gaps between various enterprise applications. These workflows are frequently interwoven across Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), and support systems, and can also involve vendor management platforms and email communications.
Typical tasks within this domain include extracting data from one system, cross-referencing it with information from another, interpreting unstructured text communications, and then making decisions regarding approval, response, escalation, or waiting. Such processes are inherently complex and require nuanced understanding.
While traditional automation methods like rule-based systems and Robotic Process Automation (RPA) have limitations in scenarios involving ambiguity and dispersed information across multiple platforms, agentic AI offers a more sophisticated solution. It possesses the capability to interpret information from diverse sources, orchestrate actions across different systems, and operate within defined policy guardrails.
The report emphasizes that agentic AI is not positioned as a direct replacement for existing SaaS platforms. Instead, its market value is derived from transforming time-consuming, manual coordination efforts into quantifiable software expenditures.
Bain estimates that vendors are already capturing between US$4 billion and US$6 billion within the U.S. market. However, the firm points out that over 90% of this potential remains untapped, highlighting a substantial runway for growth.
Beyond the U.S., Bain forecasts that Canada, Europe, Australia, and New Zealand could collectively contribute a market of comparable size. This would bring the total addressable market across these regions and the U.S. to approximately US$200 billion.
Market Potential by Enterprise Function
The addressable market for agentic AI is not uniformly distributed across all enterprise functions. Bain’s analysis indicates that the sales function represents the largest single segment, estimated at around US$20 billion. This substantial share is largely attributed to the sheer volume of sales personnel, rather than an exceptionally high inherent automation potential within sales tasks themselves.
Cost of goods sold (COGS) and operations collectively account for an estimated US$26 billion. The extensive workforce engaged in operational roles means that even modest automation rates can translate into a significant addressable market. Research and development (R&D) and engineering, customer support, and finance each represent an estimated US$6 billion to US$12 billion in market size. These functions feature substantial workforces and demonstrate higher automation potential within specific workflow segments.
Customer support and R&D or engineering exhibit the highest automation potential, with an estimated 40% to 60% of workflow tasks being amenable to automation. Bain attributes this to the presence of structured data, standardized processes, and clearer output signals in these areas. Finance and human resources fall within the 35% to 45% range, with accounts payable and payroll offering higher automation prospects compared to financial planning and employee relations, which often involve more subjective judgment.
Sales and IT functions are estimated at 30% to 40% automation potential. Bain notes limitations in these areas due to factors such as the nuance of interpersonal relationships, the deal-by-deal variability inherent in sales, and the unpredictable nature of security incidents. Legal functions show a lower overall automation potential, ranging from 20% to 30%. While tasks like contract review and compliance are repeatable, the significant consequences of errors necessitate a higher degree of human oversight.
Bain’s Framework for Automation Feasibility
The Bain report outlines six critical factors that influence the realistic extent to which a workflow can be effectively managed by an AI agent. These include the verifiability of the output, the potential consequence of failure, the availability of digitized knowledge, and the variability of the process.
Workflows that provide clear verification signals are generally easier to automate than those requiring subjective judgment. Examples of highly automatable tasks include code compilation, reconciled invoices, and resolved support tickets.
Workflows carrying regulatory or financial risk, even if technically automatable, demand closer human supervision. This category encompasses tasks such as tax filings, legal compliance, and security incident response.
Bain also identifies the availability of digitized knowledge as a key constraint. AI agents require access to structured data and documented context. Furthermore, they need machine-readable inputs, including decision-making logic that is often held informally by experienced personnel.
Integration complexity poses a significant challenge when workflows traverse multiple systems and APIs. Authentication layers and intricate exception-handling processes further complicate automation, making end-to-end automation of such workflows more difficult than those confined within a single platform. The highest-value automation opportunities are concentrated in areas where no single system of record dictates the complete outcome, often spanning multiple core enterprise systems.
David Crawford, chairman of Bain’s global technology and telecommunications practice, asserts that while SaaS companies have spent the last two decades building their positions around systems of record, the next frontier for competitive advantage lies in “cross-workflow decision context”—the ability to interpret and act across workflows that dynamically move between various systems.
Illustrative Company Examples and Adjacent Workflow Expansion
The report references several companies at the forefront of agentic AI adoption, including Cursor, Sierra, Harvey, Glean, Salesforce, ServiceNow, and Workday. Bain reports that Cursor has achieved an average monthly revenue exceeding US$16.7 million, doubling its revenue in a single quarter. Sierra has surpassed US$150 million in annual revenue, Harvey has reached over US$190 million annually, and Glean has exceeded US$200 million per annum.
The report also highlights GitHub as an example of a company that has successfully leveraged data from its core workflow to expand into adjacent areas. GitHub’s primary business revolves around developer collaboration and source code control. By utilizing its extensive repository and workflow data, the company has effectively supported its expansion into AI-assisted developer productivity tools and security automation.
Bain suggests that SaaS companies can achieve expansion through two primary avenues of workflow automation. The first is by automating core workflows, where they already possess deep domain expertise and established customer trust. Existing system integrations can often facilitate this type of automation. The second approach involves automating adjacent workflows that the company does not currently serve directly. These areas can be more challenging to identify, as they necessitate a detailed mapping of customer workflows and the underlying data that informs decision-making.
The advent of agentic AI also signals a potential shift in pricing models. As AI agents deliver completed outcomes, outcome-based and usage-based pricing models may become more relevant, particularly when agents are resolving issues or processing invoices. This contrasts with traditional pricing structures that are typically based on user seats and login counts.
Strategic Recommendations for SaaS Companies
Bain advises SaaS companies to initiate their agentic AI strategy by meticulously identifying which customer workflows are currently amenable to automation. The firm emphasizes the importance of assessing automation potential at the subprocess level, rather than treating entire functions as uniformly automatable.
The report further recommends that companies rigorously assess the quality of their data. Key factors to consider include whether the data is comprehensive, directly linked to desired outcomes, and readily usable for automation purposes.
To bridge any capability gaps, Bain suggests that companies can pursue internal development, strategic acquisitions, or valuable partnerships. The report cites AppLovin’s in-house development of its Axon platform, ServiceNow’s acquisition of Moveworks, and Salesforce’s partnership with Workday as distinct examples of successful approaches to capability enhancement.
The firm also underscores the critical need for specialized AI engineering talent, robust cloud-native architectures designed for multi-agent orchestration, and sufficient funding for model training and inference. Companies are advised to align their pricing and sales incentives with AI-driven outcomes, moving away from legacy seat-based models.
Moreover, Bain highlights the necessity for SaaS companies to establish data and product foundations specifically engineered for agentic workflows. This includes implementing machine-readable hand-off mechanisms and systems capable of capturing detailed decisions and outcomes from each workflow execution.
Crawford concludes that the window of opportunity for SaaS companies is “measured in quarters, not years,” as AI-native competitors rapidly accumulate valuable deployment data with every customer workflow they successfully automate.
Original article, Author: Samuel Thompson. If you wish to reprint this article, please indicate the source:https://aicnbc.com/21581.html
