The Adoption Gap: Why most process repositories have an audience problem

The Adoption Gap is rarely caused by a lack of effort from the process team. In most organisations, it is the result of structural decisions that were reasonable at the time but were never revisited as the programme matured. 

The tool was designed for modellers, not readers. 
Enterprise process platforms have traditionally been built for specialists who create and maintain process content. They are evaluated on modelling depth, notation support, and methodological rigour – all of which remain important. However, the experience that works well for the people producing process knowledge is not automatically suitable for the thousands of employees who need to access and apply it. 

Access was treated as a licensing decision rather than an adoption decision. 
When access is priced per named user, expanding usage becomes a budget discussion. Additional departments are added one business case at a time, and the value of broader adoption is weighed against incremental licence costs. The programme may demonstrate clear business value, yet its enterprise-wide adoption remains constrained by the commercial licensing model.  

Maintenance remained centralised. 
When only trained specialists can update process content, updates naturally happen through scheduled review cycles rather than when changes occur in the business. Over time, the repository and operational reality begin to drift apart. Once users encounter outdated content repeatedly, trust declines – and with it the feedback needed to keep the repository accurate. 

Together, these factors create a pattern that is easy to miss: organisations invest in building process knowledge, but the conditions required to make that knowledge widely accessible and continuously relevant are not always in place. 

Without regular use, it becomes harder to keep process knowledge accurate, current, and trusted. The question is no longer how comprehensive the repository is, but whether people can rely on it when decisions need to be made. 

Documentation coverage alone does not show whether process knowledge is creating value. The following four measures provide a much clearer view of whether a BPM programme is reaching the people it was designed to support. 

• Reach: What proportion of employees whose work is described in the repository accessed process content in the last quarter? Measure actual usage, not the number of licensed users. 
• Contribution breadth: How many distinct people suggested a change, flagged an issue, or contributed to process content in the last quarter? If contribution is limited to a small central team, maintenance becomes a bottleneck. 
• Content freshness: What proportion of models were last reviewed more than twelve months ago? A high share of outdated content indicates that users may begin to question the reliability of the repository. 
• Time to first useful contribution: How long does it take an untrained business user to find a process relevant to their role, understand it, and provide useful feedback or suggest an improvement? Measure it with a real employee and a real scenario – the result is often different from expectations. 

Organisations measuring these indicators for the first time often discover that reach is significantly lower than expected. This is not necessarily a reflection of the effort invested by process teams. It is usually a signal that the conditions for broad adoption have not yet been created. 

The Adoption Gap has always represented a value risk: process knowledge that is not used cannot deliver its intended benefits. In an era of increasing AI adoption, it creates an additional challenge – a readiness risk. 

AI-supported workflows, decision-support systems, and autonomous agents depend on process information that is structured, current, and connected to the wider operational context of the organisation. Process repositories that are outdated, inconsistently maintained, or rarely used provide a weak foundation for these initiatives. 

The challenge compounds over time. Process content that nobody uses is less likely to be challenged and improved. Errors remain unnoticed, models drift away from operational reality, and documentation increasingly reflects how work was intended to happen rather than how it happens today. When AI solutions rely on this information, they inherit those limitations. 

There is also a structural challenge. In many organisations, process knowledge, architecture information, and governance controls are managed in separate tools by separate teams. People can often bridge these gaps through experience and manual interpretation. Automated systems cannot. They require a connected, consistent view of how the organisation operates. 

A repository with an adoption problem is therefore not only a usability challenge. It can become a barrier to building reliable, AI-enabled capabilities. 

AI-ready process data has four characteristics that distinguish it from documentation created primarily for compliance or modelling completeness. 

• It is current: Models reflect how work is actually performed today, not only how processes were originally designed. Achieving this requires ways for the people closest to the work to identify and improve content when changes occur, rather than relying solely on periodic review cycles. 
• It is structurally consistent: Processes are captured in a structured format with defined relationships, terminology, and modelling standards. While written procedures often rely on interpretation, structured process information provides a more reliable foundation for analysis, automation, and AI-supported use cases. 
• It is connected: Processes are linked to the applications that support them, the data they depend on, and the governance requirements that apply to them. Without these connections, AI systems operate with incomplete context, increasing the likelihood of inaccurate outputs or missed dependencies. 
• It is actively used: A repository that people regularly consult is more likely to be corrected and improved when reality changes. Adoption and accuracy are therefore closely connected: the Adoption Gap and the AI-readiness challenge are two perspectives on the same underlying issue – whether process knowledge remains relevant and usable. 

Closing the Adoption Gap is not primarily a training challenge. Training can help people understand processes, but it cannot compensate for barriers in access, usability, or governance. 

Three changes have the greatest impact: 

• Make consumption effortless. 
  Business users should be able to find and understand the process knowledge relevant to their work without specialist training or access to modelling tools. Process information needs to be available where people already work and in a format designed for everyday use. 
• Make contribution possible. 
  The people closest to operational reality should be able to identify outdated or incorrect content when they encounter it. Continuous feedback loops are what keep process knowledge accurate over time. 
• Make adoption scalable. 
  If expanding access increases cost significantly, licensing constraints can become a barrier to adoption. Access to process knowledge should be guided by business value, not limited by the cost of adding more users.  

None of this requires organisations to abandon the process knowledge they have already built. Migration approaches and technologies have evolved significantly, and moving an established repository to a platform designed for broader adoption can now be planned in weeks rather than years. 

But the first step is not a tooling decision. It is understanding the current situation: How many people actually use the process knowledge your organisation has invested in creating? 

Our ARIS Migration Handbook explains what happens when you move from ARIS to GBTEC Platform – from object transfer and conversion decisions to realistic timelines and migration phases. 

Understand what can be automated, what requires review, and how to approach migration with confidence. 

Download the Migration Handbook