Overlooked Asset for Enhanced Decision-Making and Value Creation
Organisations often overlook a critical factor that can significantly influence their strategic success or prevent them from implementing planned strategies: data quality. Timely, high-quality data serves as the foundation for informed decision-making and is essential for the effective deployment of digital enablement strategies, including those containing artificial intelligence (AI).
Latest technologies have advanced significantly, enabling organizations to deliver actionable contextual information that supports timely decision-making. These technologies provide only the necessary documentation at the required materiality level for authorized recipients, and also ensure that comprehensive compliance related details are accessible to process owners. Processes can be executed at application or at individual process level, utilizing cloud(s), edge, or hybrid deployments to achieve high availability, low latency, and enhanced privacy. Below we explore practical ways associated with iteratively enhancing data quality that will save time and add value.
Data quality refers to the degree to which data is accurate, complete, consistent, valid, timely, and unique, making it suitable for its intended purpose. High-quality data is essential for effective decision-making, as it ensures that different functional areas within organizations can rely on the information they use for analysis and reporting. Poor data quality can lead to operational issues, misguided strategies, and financial losses. With regards to the latter, it should not be a surprise that many digital enablement and / or AI projects fail in the first instance due to a lack of sufficient data quality.
Key Benefits of Enhancing Data Quality:-
Informed Decision-Making
Users require access to all supporting information in a timely manner for decision support, controls and management activities. Sounds logical, but the challenges here are often difficult due to: i) required data being scattered across multiple applications, and ii) required details typically being a subset of an overall transaction. These include currency, operating segments, projects, locations, product groups etc.
To effectively facilitate decision-making, data must be transformed and presented in a manner that is clear and actionable. Time is indeed a critical factor in business, and addressing challenges early with the right operational team can expedite actions that lead to success while minimizing risks.
Process Responsibilities
Process owners are responsible for managing and overseeing business processes to ensure they align with organizational goals. In contrast, process users execute daily tasks within these processes. Importantly and behind the scenes, automation significantly enhances efficiency with simple actions performed by users further orchestrating complex automated processes that streamline operations and improve data accuracy.
Enhanced Operational Efficiency
Effective business process design is essential for maximizing efficiency. By validating data sets at the point of entry, organizations can eliminate time-consuming manual steps required to resolve issues later on. This proactive approach is particularly important for analysis codes that may not have been actively validated in the past by legacy systems. This is caused by important information being squeezed by users into any available application fields—a practical solution to a common constraint in earlier years.
During processing, data will be fully and extensively transformed automatically as mandated by the process owner, ensuring that it meets the necessary standards and requirements. Additionally, comprehensive audit trails will be automatically maintained, providing traceability and auditability throughout the data lifecycle.
Data Enrichment
Critically, data can also be enriched by the system to further improve underlying quality in a multitude of ways. For example, source ledger details can be added to transactions as they flow within complex consolidations, and tiered allocations can contain information on where data has been allocated, as well as the calculation method. There are endless possibilities that are relevant to your business and that will improve data transparency.
Reporting based bottlenecks, particularly those involving various types of reconciliations that require tedious data transformations through spreadsheets, can now be fully or partially automated as needed. This automation significantly reduces the time and effort traditionally spent on these processes. Time that can be used for other activities.
Spreadsheets in Core Processes – The Options
Two options when it comes to spreadsheet use within core processes; 1) organisations that want to retain their ease of use can do so, noting that modern process technologies add full underlying cell data integrity, plus further additional value by being able to leverage practical application specific shortcuts offered by the vendor. These optimisations go even further to improve user efficiencies during set-up and operational use, noting that many vendors have now incorporated AI to make them easier to use and more intuitive to users; 2) alternatively spreadsheets can be removed from core processes, and here vendors still ensure ease of use with comprehensive but simpler functionality.
Enriching data and improving transformation capabilities for reporting and reconciliations improves quality and reduces errors. This helps organizations identify and resolve operational issues, whilst maintaining overall data integrity. Domain resource time availability improves, noting that the ability to share information more easily and compliantly with other stakeholders goes to facilitate deeper collaborations across the board, a point that is expanded on below.
Let’s add an additional level of sophistication.
Proactive not Reactive Strategies for Decision Making
Smart processes also mean that organizations can shift their decision-making from being reactive to proactive on critical issues, without information overload to users. To understand this, let’s take a look at how modern processes work.
Process Design
Latest software technologies enable ultra-granular transformations to be executed, including data enrichments and validations. This then drives actionable contextual business flows together with supporting information that may include: transaction lists, reconciliations, segments, projects etc. As touched on above, information is produced to the materiality required @anywhere @anytime within a process.
From this point simulations can be undertaken: for example to find i) the optimum sales mix to maximise cash flow or ii) to explore other scenarios like maximising profit overall or by segment. Variances can be ranked as required by segment, business unit etc.
As processes are built specifically for purpose they contain no excess functionality, and therefore can be maintained and secured more easily as programs are smaller in size. Different data types, both structured and unstructured, that continue to emerge can be used as appropriate to support decision making.
Artificial Intelligence
Processes may or may not contain AI, but if used will see combinations of i) Top Down AI; ii) Bottom Up AI (LLMs); and iii) GenAI. In all cases where corporate data is accessed, designs should ensure that existing user access controls are leveraged to restrict any unauthorised access to sensitive data sets. Some user cases for reference are provided later.
Going forwards there is going to be a lot more focus placed on understanding how processes are performing, as well as their related efficiencies.
Process Metrics to Drive Improvements and ROI
Except for those processes under BPO or within internal shared service centres, there are rarely detailed operational or efficiency metrics in place. End of month reporting for many has traditionally been the backstop for decision making at post monthly meetings, with little focus placed on understanding the time consuming repetitive activities to create them. Furthermore, different functional areas have had far from a desirable functional overlap between them for collaborative purposes.
Changing Process Dynamics
Dynamics are now changing as i) latest process technologies incorporate / automate both front and back office processes bringing them closer together; ii) which in turn is increasing end to end business velocity; and iii) which is starting to significantly shrink distances between domain areas.
Going forwards, modern systems can help provide metrics that can be used for measuring improvements and ROI. For example, the frequency of invoicing, typical dollar values per transaction, numbers of transactions, % accuracy statistics where AI is used etc.
Proactive Metrics to Reduce Risk
These metrics can also be used to add further compliance controls. For example, in payment processes to mitigate against Business Process Compromise (BPC) the system can help to proactively identify anomalies. BPC is where a threat actor tries to divert your funds to their bank account or others under their control.
At the end of the day, working with metrics will help drive productivity, but history has shown that driving ongoing improvements will need focused and proactive management attention to make it effective. However, there are intangible benefits.
Building Trust and Confidence – Improvements to Work:Life Balance
Every day, businesses have to process massive amounts of data, and at some point information this has to be used for decision support, controls and management activities. With business velocity ever increasing, producers of information have to get it right and are always time constrained in some form during execution, causing them to constantly check themselves on accuracy.
Working with high quality timely data makes information current and builds deeper trust for process owners and users, as data becomes more readily understood, transparent, and auditable. This makes it easier to manage volumes of data, providing additional time for checks and balances, including those for compliance and governance.
Compliance and Governance:
Today, regulations continually evolve to be broader and deeper, but on the positive side compliance and governance controls can now be increasingly incorporated within processes, rather than as an adjunct to them. This might also include understanding data outliers that warrant deeper investigation.
Recent legislation has placed a greater emphasis on the proactive management of data by organizations. Companies are now required to carefully oversee the compute and storage locations of their data, including the separation of personal information and sensitive personal information that needs to be managed differently at the data field level. This responsibility extends to ensuring full compliance is maintained with data retention strategies and policies, both within national borders and across international boundaries.
Risk
Failure to adhere to these regulations can result in significant legal risks and penalties. Organizations must therefore be vigilant in their data management practices to avoid potential repercussions, which can include fines, lawsuits, operating licence withdrawal, and damage to their reputation. As the landscape of data privacy continues to evolve, it is crucial for businesses to stay informed and compliant with the latest legal requirements to safeguard their operations and customer trust.
GDPR
Privacy legislation differs by country, but many corporates have standardised on GDPR and then managed differences over and above this standard on a country by country basis. APIs to third party web services comes with the responsibility of ensuring that compute, storage locations and policies are fully understood and consistent with legislation.
AI Legislation
Another more recent example pertains to AI legislation. As with all new technologies, this legislation typically lags current realities. It is fast catching up and is particularly relevant when it can impact personal outcomes, for example if used in recruitment.
All of the above contribute to making business systems more usable and efficient at a detailed level, but when combined to drive improvements across applications and ecosystems start to make a big difference.
Integration Across Applications and Ecosystems:
Digital enablement is about being able to drive both qualitative and quantitative processes in real time (or close to it) across multiple applications and ecosystems, using the granular process designs outlined above. By doing so, processes that were once run sequentially, in parallel, or more manually are increasingly becoming automated more fully. This in turn is leading to considerable time being saved by different types of stakeholders, including customers and suppliers.
Increased compute power, granular process designs, plus the ability to deploy processes across multiple applications and ecosystems is providing an environment to drive meaningful continuous improvement in more meaningful ways.
Continuous Improvement with Meaningful Change
Continually iterating systems to reduce friction is something that has played out over the years, with vendors and customers driving improvements. However, many users have argued that some vendor driven application releases have been due to vendors forcing version or platform “upgrades” on to them, in essence for their own financial benefit. This has been particularly noticeable they say, where these iterations and improvements might have been very small or non-existent in comparison to the implementation work by customers to deploy them. Other cases cited and currently playing out have been where a vendor forks on-premise vs cloud versions of the same software. This in practical terms means that new functionalities are available in the cloud version only.
Multi-Level Step Change
Nevertheless, there is no doubt that we are currently within a period containing two levels of major step change: first came smart processes with the ability to design ultra-granular qualitative and quantitative processes (ie all processes of any type) that can generate actionable contextual alerts and workflows @anytime @anywhere across applications and ecosystems, together with supporting information to required materiality. Next, came AI technologies that go even further to complement these processes to make them smarter.
Traditional strategies of upgrading applications have as a result been challenged for some time. Business systems have become a combination of Smart Processes + Applications (System of Record), plus no code / low code functionality has helped users be more in control of making changes. Applications can to some extent be written by AI, but early days.
Changing Modus Operandi
The previous reluctance of not upgrading core systems for fear of introducing inadvertent errors has been to some degree replaced by meaningful continuous improvement strategies. In essence, faster compute capabilities, coupled with the ability for smart process design has for the first time allowed corporations to more easily drive meaningful value creation with reduced disruption.
Furthermore latest technologies can see data patterns that are not typically perceptible to users, plus have the ability under certain conditions to learn from experiences. For example, agentic AI – a smarter hybrid combination of the AI types described above, is where multiple algorithmic models can work together and results can improve over time.
With the right technologies in place, moves can be made to drive improvements. However, before starting it is important to focus on operational barriers that will have to be overcome.
Overcoming Operational Barriers to Move Forwards
Here are some steps to move forwards:-
Trusted Partner: select a partner that has developed optimised process technologies (domain and functional area), and has a proven track record with multiple customers.
Change Management: think this through carefully.This is hard area to get right at the best of times, and is an area that causes many digital enablement projects to not deliver against expectations.
Project Teams: build project teams that have representation from all impacted areas.
Software Integrations: spend quality time assessing how integrations will be achieved, recognising that this is key area that derails many projects. Simply put, this is down to software being like an iceberg with many dependences to your success being not immediately visible to operational users ie they are below the surface.
Staff: proactively communicate your intentions associated with change, and address the elephant in the room regarding foreseeable future role changes.
Breadth and Depth of Change: There can be multiple significant changes taking place at the same time that include i) increased business velocity; due to ii) more tightly integrated front / back office systems (including ecosystems like Open Banking); and iii) which leads to shrinking demarcation lines between and increasing number of functional areas: operations, compliance, HR, Treasury etc.
Individually these can each be major challenges in their own right. It is important to recognise that a greater depth and breadth of planned change will significantly magnify your challenges, but conversely it also has the potential to drive deeper value creation by fostering broader levels of communication.
Therefore, one has to decide on how to tackle these changes. Some have created independent digital teams to work with functional domain areas to overcome; i) the inevitable initial digital skills gap, and ii) the inertia associated with managing change over and above normal daily activities. Others have taken a more strategic approach to drive change across domain functional areas. Recognise that as well as challenges, that increased collaboration across functional domain areas will benefit longer term co-operation to drive deeper value.
ROI: Justifying and measuring ongoing success of a project is not always easy to achieve. The key issue is metrics or lack of them as described above. Traditionally many organisations have relied on month end reporting to drive activity, so overall progress has been typically measured in the number of days to close the books, despite the fact that many of the current processes are drowning in inefficiencies. Smart processes today support the building of KPIs and OKRs, which can form the basis of measuring improvements across multiple years and can provide a basis for ROI discussions.
Prioritisation and Focus: The big picture is that the organisation needs a vision of how it will develop its processes over the short, medium and longer term. One can quickly see that any ad hoc process development might have too narrower or too larger scope to deliver results. Having said that there are clearly many improvements that can be made within functional areas that will not impact other areas. Prioritisation and focus are therefore important.
Ongoing Improvements: Metrics are going to be more important going forwards eg document recognition, as it is likely that these can be improved, recognising that you will first need to have the relevant metrics in place. Constantly review and improve your quality metrics and do not lose track of maintaining trusted systems of record. When deploying AI, focus on i) implementation and infra structure costs; ii) ensuring no bias in HR related deployments; iii) and ensuring that you have in place the data quality that will let you drive home your objectives.
Examples
A few examples follow as to where some have proactively focused to drive value:
T&E: Assuming that you have reached the point of ensuring that all travel is within policy and pre-authorised, then with higher data quality, and depending on your volumes to make it viable, you can leverage data to get higher discounts with airlines and hotels as negotiations can be backed by meaningful statistics and commitments. Additionally, on an alternate note, quality data allows you to identify more detailed outliers for further investigation.
Procurement: Automation of procurement to pay processes using OCR / AI for document recognition, combined with top down AI to ensure that available discounts are leveraged. This approach is used as available discounts might be under corporate umbrella agreements with specific overriding discount information not shown on an individual invoice.
Consolidation: Automations today allow all aspects of complex consolidations to be automated, including aggregations, segmental analysis, reconciliations, with ranked variances plus actionable, contextual reports supported to the materiality required by users. Going further, latest technologies can be used for simulations to explore and probe many areas, such as product combinations that generate best cashflow, or profitability etc.
Compliance Controls: Additional controls can be put into place. For example, as illustrated above for the payments process, additional steps can be added to mitigate against Business Process Compromise, or in final reporting pack submissions to ensure that nothing is sent to HQ without a final review and sign off, or in capex that assets and asset improvements are completely insured according to company policy.
Conclusion
In conclusion, prioritizing data quality is essential for organizations aiming to enhance their decision-making processes. By addressing the blind spots related to data quality and leveraging data enrichment strategies, organizations can transform their decision-making from reactive to proactive, allowing them to navigate challenges and seize opportunities more effectively.
Moreover, fostering a culture that values data quality not only improves operational efficiency but also builds trust among stakeholders. By investing in robust data management practices, companies can unlock the full potential of their data and in future leverage AI capabilities, ultimately leading to better outcomes and sustained competitive advantage.
