Business Process Automation Market Seen Reaching $35.5B by 2030

A big number arrives during a messy transition

The business process automation (BPA) market is getting another forecast, this time from a 473-page strategic report profiling 122 companies. The report puts the global BPA market at $18.7 billion in 2024 and expects it to reach $35.5 billion by 2030, a compound annual growth rate of 11.3%.23 Research and Markets announced it on March 4, 2026. Its profiles include ABB, IBM, Infosys, Polycom, Rockwell Automation, SAP, Schneider Electric, Siemens, Tata Consultancy Services and Yokogawa.23

The headline figure matters less than when it lands. Enterprise automation is moving from scripted bots toward AI agents that are supposed to reason through messy work. Many of those agent projects are stalling between pilot and production. Read together with that shift, the report tells a cautious story: demand for automation is real and lasting, but the tools buyers will use, and how the market is even defined, are changing quickly.

What the report says

The report names familiar growth drivers: pressure on costs, digital transformation programs, and AI and machine learning being built into BPA tools. That last factor, it argues, pushes automation from repetitive tasks into processes that involve decisions and predictive analytics.23 It also expects cloud delivery to bring automation within reach of small and midsize businesses that could not previously afford large platforms.23

The segment figures are more telling. On-premises deployment is projected to reach $16.9 billion by 2030 at a 9.2% CAGR, while cloud deployment grows faster at 13.4%.23 By region, the U.S. market is valued at $4.8 billion in 2024. China is forecast to grow 14.7% a year to $8.2 billion by 2030.23

The trend list reads like the automation industry's current talking points. It includes robotic process automation (RPA) for repetitive work, low-code and no-code platforms that let business users build their own workflows, process mining to find automation candidates, intelligent document processing for unstructured data, and "hyperautomation" for end-to-end process redesign.23 Coverage of the report highlights the same opportunities: digital transformation, cloud tools for smaller firms, better customer experience, and regulatory compliance.24

The forecasts don't agree

Put this report next to other market estimates and the numbers stop lining up.

The Business Research Company sizes BPA at $16.32 billion in 2025 and $18.83 billion in 2026. It expects $33.43 billion by 2030 at a much faster 15.4% CAGR.21 Both firms land in roughly the same place for 2030 but take different routes to get there. The bigger surprise is the publisher's own listing: a June 2026 edition of the same 473-page strategic report estimates the market at just $9.4 billion in 2025, rising to $20.4 billion by 2032 at 11.7%.25 That is about half the 2024 baseline announced in March.

The related reports released the same morning make the picture harder to read. One values RPA at $5.3 billion in 2024 and forecasts $35.2 billion by 2030, a 36.9% annual pace.30 By that estimate, RPA alone would match the whole BPA market. Digital process automation is forecast to grow from $17.5 billion to $33.2 billion over the same period.26 Intelligent process automation starts at $51.8 billion and grows only 5.9% a year to $72.9 billion.29

The most likely explanation is that these categories overlap heavily and are drawn differently in each report. That is not a scandal, but readers should take it seriously. Any single forecast line describes how one analyst draws the category boundaries, not the size of the economic activity. Buyers and investors should treat the $35.5 billion figure as a sign of direction, not something to plan around.

An industrial-looking list for a software market

The list of profiled companies stands out too. Half of the named companies (ABB, Rockwell, Schneider Electric, Siemens and Yokogawa) are best known for factory and process-industry automation. The list also includes Polycom, a communications brand.23 The Business Research Company's list is almost entirely software vendors: Microsoft, IBM, Oracle, SAP, Salesforce, ServiceNow, Pegasystems, UiPath, Celonis and OutSystems, among others.21

The difference suggests the strategic report defines BPA broadly enough to include operational technology. The second list is closer to what most enterprise buyers mean by BPA: workflow, RPA, process mining and low-code tools built for back-office and customer-facing work. For readers following AI workflow automation, the software-heavy list is the more useful guide to who is competing.

The real story: agents meet RPA

The report treats RPA as a growth driver.23 Much of the industry, though, is busy changing what RPA means. Analysts and vendors broadly agree on the problem. Traditional bots depend on hard-coded screen elements and break when an interface changes. They also struggle with emails, PDFs and contracts, the unstructured material that makes up much of enterprise work.11 AI agents are pitched as the answer because they can read documents, interpret intent and recover from changes.1113

The large RPA vendors are moving in that direction. UiPath has expanded into agentic orchestration, with agents that work across processes and hand off to humans when needed.12 Partners describe tools that monitor workflows and fix them automatically, and orchestration that spans agents from Microsoft, Google, OpenAI and others.14 Microsoft offers its own combination: Power Automate for structured flows, AI Builder for document extraction and Copilot Studio for agent behavior.16 No-code agent builders and open-source tools such as n8n are adding AI nodes so non-developers can build context-aware automations.12

On the main architectural question, the coverage mostly agrees: agents add a layer on top of RPA rather than replacing it. Several analyses describe a hybrid model. Agents interpret goals and handle exceptions, while deterministic bots carry out the structured system actions.2016 Even a vendor that markets RPA replacement recommends keeping stable bots running and retiring fragile ones over 12 to 24 months.11 Not everyone agrees. One consultancy says most 2026 deployments are moving from RPA to agentic designs because old bots cannot cope with the volume of exceptions.17 On balance, the hybrid view is the safer bet. It fits how slowly enterprises actually retire systems that already work.

Why the agent forecasts need a discount

The adoption numbers depend heavily on who is counting. One cited PwC survey found that 79% of organizations run AI agents.17 The same article quotes Deloitte research showing that only 11% operate agents at full production scale.17 An analysis of Gartner's 2026 Hype Cycle for Agentic AI puts production deployment at 17%. It also says most agentic innovations, including multi-agent systems, sit at the Peak of Inflated Expectations.9

Gartner's best-known warning frames the whole sector. In June 2025 the firm predicted that more than 40% of agentic AI projects would be canceled by the end of 2027. It cited rising costs, unclear business value and weak risk controls.48 It also estimated that only about 130 of the thousands of vendors claiming agentic capabilities are genuine. The rest, it said, are "agent washing": rebranding chatbots, assistants and RPA bots.81 A year on, commentary in Forbes argues the forecast now describes the present more than the future. That commentary blames the failures on missing success metrics, poor data access and unclear ownership, not on weak models.1

Gartner's longer-term view is still positive. It expects at least 15% of day-to-day work decisions to be made autonomously by agentic AI by 2028, and a third of enterprise software applications to include agentic AI by then.4 Both predictions can be true. Many weak projects get cut, and the use cases that prove themselves keep growing.5

Readers should be especially skeptical of vendor return-on-investment (ROI) claims. One automation vendor's blog, for example, says AI automation delivers 250–300% ROI against 10–20% for traditional automation.19 Figures like these come from people selling the product and are not independent evidence.

What it means

The strategic report's core point holds up: companies will keep spending to automate processes, and AI is widening the range of work that can be automated.23 What the report's segment tables cannot show is that the dollars are moving between categories. Money is shifting from standalone RPA licenses toward orchestration platforms, document intelligence and agent tools, which may get counted under different labels in next year's forecasts.

For buyers, the practical lessons come from the cancellation data more than the market sizing. Successful agent projects set measurable targets before any code is written. They model costs at production scale and build in governance from the start: human approval points, decision logging and limits on which tools an agent can use.203 Analysts note that companies with modern, API-accessible systems deploy agents more effectively. Those with fragmented legacy systems still lean on RPA to connect them.20 In practice, the speed of the BPA market's shift to AI agents will depend on how quickly companies modernize their underlying systems.