Major divergence in the tech industry

Data and process are being treated as increasingly distinct concepts, with different job roles, terminology and tools, when they are two sides of the same coin.

Specialisms versus silos

In the data space, data analyst, data architect and data scientist are prevalent. Process analyst, process improvement specialist and RPA engineer are increasingly widespread. This in itself is not an issue - we need specialists - and these areas of specialism are progressively more complex.

The issue arises when the concepts of data and process become so divorced that the relationship between them is no longer understood, acknowledged or discussed. The process-data divide risks becoming the new gap to be bridged, replacing ‘IT’ and ‘business’ as the dominant silos within organisations.

Symbiosis of data and process

Almost all data is created, either manually or automatically, as part of a process. Almost all processes use, create or transform data. How can we bring the integral relationship of data and process back to the forefront of discussion? By being curious.

Business analysts have a duty to their organisations to think holistically and consider all perspectives. This means asking questions about both data and process and creating appropriate models which prompt engagement and facilitate discussion.

Through a data lens:

• How and when is this data created?
• Who uses the data?
• What is the business purpose?

Through a process lens:

• What data is required for this process to succeed?
• What data is created during the process?
• How will data be transformed?

Why are so many BAs scared of data?

It is interesting and troubling that business analysis, which has its roots in systems analysis, has turned its back on data. Data is a vital asset for organisations. Those who make data-driven decisions and are able look for early warning signals provided by their business data, are the most successful in the long term.

For you

Be part of something bigger, join BCS, The Chartered Institute for IT.

Business analysts should be championing evidence-based decision making; to do this, we must be comfortable with data. This does not mean pivot tables and the ability to use R. That is analytics, not analysis. The two areas are related but are not the same. It does mean the ability to understand key data concepts such as relationships between data, interfaces between systems, and master data management.

Logical data models are a key tool in the BA toolkit. Being able to visualise data, understand the connections and hierarchies is a business imperative and class models or entity relationships should not be perceived as ‘too technical for business users’. When we use appropriate business language, business users and leaders are entirely capable of filling gaps in data models, highlighting missing concepts, assumptions and errors.

Process models were once ‘new’ to many organisations and business users. BAs are confident championing the value offered by process models: reduce hand-offs, remove bottlenecks, improve cycle times, enable consistency, retain knowledge, train staff and more. The same cannot be said for data models.

The future of data in business analysis

The national apprenticeship standard for business analysis was recently reviewed and updated. The most contentious area amongst BA leaders? Data.

Half the group stated that in a 20+ year career as a BA, they had never created a data model nor needed to consider data concepts. Half were incredulous and felt data was equally relevant to process for the BAs of the future. So, what’s the truth of the matter? Well, it depends.

Many BAs do not perform data analysis or see logical data models as valuable. If no one is ‘asking’ for these types of models, they are deemed unnecessary. BAs need to have the confidence to produce logical models and see the value in data modelling, to be able to identify when it is appropriate and will offer value to the organisation. Business initiatives which involve system upgrades and migrations, reporting, security, access, compliance or interfaces are all likely candidates.  

BAs must be confident championing the value offered by data models: understand impact of change, reduce errors, improve data quality, increase data ownership, optimise design, apply security, enable consistency, retain knowledge and more.

Process and data are inextricably linked and someone needs to have their eye on both concepts. If BAs are truly holistic thinkers, we must consider our accepted role in analysis of process (model the process, validate the model, improve the process) to be equally applicable to the world of data.