Consider Data & Process as Two Sides of the Same Coin

Which came first the data or the process? In most large enterprises today, business process and data management professionals tactfully ignore one another. Business process professionals – primarily business subject-matter experts representing functional groups or business units – focus on the business problem and the human-centric dependencies, such as defining roles and responsibilities for getting processes done and facilitating organizational change management. Alternatively, data management remains an IT-driven competency and focuses on technical solutions with an unrealistic “if we build it, they will come” mentality regarding business adoption.

In reality, business processes based on the invalid assumption that trustworthy, insightful data will magically be made available are at risk of breakdown – causing significant employee inefficiencies, poor customer experiences, and increased operational risks. And data management solutions, like data warehousing, business intelligence, and master data management, always face the truism of “Put garbage in, get garbage out” – meaning that if the processes that capture and update the raw data feeding these systems generate untrustworthy data, no amount of technology investment will be able to clean it all.

Some business process professionals get it. Forrester recently talked with 28 business process pros, many with substantial business process management (BPM) expertise, and a number of them voiced awareness that MDM and BPM should be aligned, process modeling and data modeling should be interrelated, and data governance and process governance should be on similar, integrated tracks.

High-Quality Data Requires Standardized Business Processes . . .Data management technologies, such as extract, transform, and load (ETL) tools, data quality (DQ) platforms, and master data management (MDM) solutions, deliver impressive capabilities to effectively transform raw data into useful information. But this technology has hit a glass ceiling in its ability to ensure that enterprises embrace it as critical infrastructure. Why? Because MDM can only deliver a single view of critical enterprise information if it can trust the processes and systems that feed it.

MDM and data quality efforts often focus on collecting data – such as customer name and contact information – from multiple sources like ERP and CRM apps, eCommerce websites, order management systems, and external partners. The DQ or MDM solution will then apply complex data standardization, cleansing, matching, and merging logic to attempt to reconcile and deliver a master version of the customer data. But what happens when the upstream business processes capturing and updating this customer information ignore data quality and completeness? For example, when call center processes don’t require reps to ask inbound callers to confirm and update their preferred contact information (phone, mail, email), downstream direct marketers trying to generate successful outbound campaigns suffer.

Inhibitors to delivering – and deriving value from – MDM include:

  • Business process efficiency prioritized over data quality. Business leaders that own these processes often measure success as their ability to complete their processes as efficiently as possible and move on to the next transaction or the next step of the process. Rarely do process owners use data quality as a measure of process quality. But they should.
  • Data governance and stewardship processes deemed nice to have, not must have. Too often, organizations view data quality as a one-time cleanup project, not an ongoing organizational responsibility. But the only way to earn business confidence in critical enterprise data is to adopt a formal data governance program that defines business and IT processes for monitoring and mitigating exceptions to data quality rules. For example, a CRM system may have multiple records with the customer name “John Smith” but may lack sufficient information to automate a match and merge of these records into a single, unique individual. Instead, these potential matches must be routed to a data steward who can manually analyze and determine whether these records should remain unique or be merged.
  • A lack of focus on cross-enterprise dependencies. Corporate prioritization processes often put data quality investments into the hands of functional or business unit leads who are given the incentive to address specific needs – like improving direct marketing campaign effectiveness or reducing average handling time in the call center – rather than fixing more holistic enterprise issues. As a result, an ERP-centric data quality project to standardize bill of materials (BOM) data, for example, could in fact cause significant confusion and frustration for end users of the downstream data warehousing and business intelligence financial reporting capabilities if their definitions of BOM quality and consistency don’t match those of the ERP end users.

. . . And Critical Business Processes Demand High-Quality DataAsk any business process professional and she’ll tell you that data just isn’t sexy; it simply rides along the process train doing nothing special. However, savvy process leaders and process analysts understand the critical role that data plays in process execution – driving automated decisions, routing rules, and placing process in the proper context for end users. In many ways, data is the fuel that drives the process engine forward.

Unfortunately, most process automation and process improvement initiatives place a low priority on data quality from the very outset – sometimes setting off a vicious cycle of process and data attrition that can doom fledgling BPM projects to failure.

  • Garbage in/garbage out creates process confusion. Ultimately, BPM is about aggregating and presenting relevant data to business users at key points in the process, empowering users to make informed decisions on how the process should proceed. When process architects pull process data from low-quality data sources, decision quality takes a hit, leading to significant rework and frustration for business users and stakeholders.
  • Lack of process trust slows BPM momentum and adoption. When data inconsistency and data quality issues are left unchecked on BPM projects – or when data issues are dealt with in triage mode – business users quickly begin to question the validity of underlying business processes. “Our processes drive salary and compensation decisions, and we have to trust the integrity of the data, otherwise we will be forced to fall back on the manual process,” reported a senior manager at a large retail bank.
  • Data conflicts reinforce a siloed approach to process management. Department managers and staff often see data as a valuable asset to be guarded at all costs, and business process professionals often overlook this key political issue. When cross-departmental data conflicts are not addressed or left unresolved in end-to-end business processes, users are more likely to revert to their original, disconnected departmental data and processes.

Business process professionals and data management professionals specializing in data management and business process improvement projects must bridge their currently siloed initiatives.

To start, you should:

  1. Frame your MDM business case as a process improvement initiative. Recognize data as the means, not the end, to improving the critical business processes and decisions that drive your organization. Your MDM business case must demonstrate how improvements to these processes will reduce costs or risks, increase revenues, or create strategic differentiation. It can’t simply focus on how many duplicate records will be eliminated – because without the appropriate business context, who cares about the duplicates?
  2. Align data and process governance organizations. Establish formal lines of communication between your BPM and MDM initiatives by bringing business process professionals into data governance and stewardship efforts and by bringing data management professionals into process governance activities. Be proactive – don’t wait until you hit data quality challenges on your BPM initiative. Develop your strategy and game plan for synchronizing process data from the very beginning.
  3. Look to MDM vendors to take the lead on evangelizing business process convergence. MDM and DQ vendors evangelize data governance best practices and deliver stewardship functionality to manually review, validate, and approve questionable data. Unfortunately, they can only control the approach that masters the data and delivers a ”single view.” But these capabilities do not ensure that the master data is consumed by the right stakeholders, at the right time, and in the right context. That’s why building the business case for MDM is so difficult – and why these vendors need BPM specialists to embrace MDM to bridge this gap.
  4. Force your BPMS provider to take a position on data management. Realize that your BPMS vendor is happy to relegate data management to a component outside of its product’s scope. Counteract by pushing your BPMS vendor to explain how its offering minimizes the time and resources needed to keep process data in sync with trusted data sources.

Share this post

scroll to top