The automation and low-code market spans several overlapping categories. General integration platforms connect existing apps without code. Robotic process automation tools mimic human actions on legacy systems. Conversational and agentic AI builders create assistants that reason over data and take action. Workspace tools increasingly bundle lightweight automation directly into the systems people already use daily. Understanding which category a platform sits in is the first step in a fair comparison, since tools built for different jobs cannot be judged on the same criteria (Lamanna, 2025).
Microsoft Power Automate is the automation surface built into the Microsoft 365 and Power Platform ecosystem. It connects natively to SharePoint, Teams, Outlook and Dataverse, and extends to robotic process automation through desktop flows and AI capability through Copilot connectors. Pricing ranges from a free trial through a per-user Premium tier, with separate bot licences for unattended automation (Microsoft, 2026a). It suits organisations already licensed for Microsoft 365 that need governed automation, though its advantages are less pronounced outside that ecosystem.
Zapier was the original integration-as-a-service platform and remains the most accessible option for non-technical teams, connecting more than 9,000 apps through a visual, no-code editor (Zapier, 2026). Pricing is based on completed tasks rather than workflow complexity, so costs can rise quickly with volume. Zapier suits simple, linear workflows that need to go live fast, but its logic is more limited than platforms built for complex branching.
Make, formerly Integromat, offers a more visual and flexible canvas than Zapier, supporting branching, iteration and data transformation at a lower cost per operation (Make, 2026). It appeals to teams that need more logic than Zapier provides without moving to a fully technical platform, though the interface has a steeper learning curve for new users.
n8n is a fair-code, self-hostable platform aimed at technical teams. Unlike task-based competitors, it bills by workflow execution regardless of the number of steps involved, and can be self-hosted so workflow data never leaves an organisation’s own infrastructure (n8n, 2026). It supports custom code and advanced AI agent building, suiting organisations with development capability and strict data residency needs, but less suited to non-technical citizen developers.
Microsoft Copilot Studio is a graphical environment for building conversational and autonomous AI agents that can call APIs, retrieve knowledge and take action within Microsoft 365. Its governance controls, including connector restrictions, data loss prevention policies and a central agent control plane, are designed for organisations that need to monitor many agents at scale (Microsoft, 2026b). It suits agent-specific use cases rather than general workflow automation.
Notion’s native database automations let teams trigger notifications, status changes and property updates without leaving their existing workspace (Notion, 2026). They are useful for lightweight, single-database triggers, but cannot support multi-step conditional logic or connect meaningfully outside Notion, so they complement rather than replace a dedicated automation platform.
Other platforms are worth watching depending on context: Workato and Tray.io target integration-heavy mid-market and enterprise use cases, Boomi and MuleSoft serve larger integration platform as a service requirements, and dedicated robotic process automation tools such as UiPath and Automation Anywhere remain relevant where legacy, non-API systems must be automated. The low-code and automation market continues to grow quickly, so ongoing awareness of the wider market matters more than loyalty to a single vendor (Forrester, 2026).
A structured comparison avoids decisions based on marketing or familiarity. A recent evaluation framework scores platforms against process orchestration, integration, governance and security, and AI-enhanced automation, then weights each criterion by organisational priority before ranking candidates (Lamanna, 2025). Risk and governance deserve equal weight to features: the NIST AI Risk Management Framework’s map, measure, manage and govern functions transfer well to platform decisions, prompting a clear view of what could go wrong and who remains accountable (NIST, 2023). The UK Technology Code of Practice adds useful prompts around integrating with existing technology and defining a purchasing strategy, relevant well beyond the public sector (Government Digital Service, 2025).
Action Point
Choose one process being considered for AI or automation. Break it into smaller tasks and describe the outcome, inputs, rules, variation, judgement and risk for each one. Then compare these needs with what the technology has proved it can do. Identify one task that is a strong fit, one that needs human-technology collaboration and one that should remain human-led.