Enhancing Financial Efficiency and Resilience

Areas to Consider

Amid ongoing global challenges, such as geopolitical tensions and economic fragmentation, the urgency to enhance financial efficiency, business strategy, and resilience has never been greater. The current economic landscape is marked by a renewed cycle of tariffs and retaliatory measures, persistent inflationary pressures, and ongoing supply chain disruptions, all contributing to a climate of uncertainty, whilst also presenting opportunities. Furthermore, governments are simultaneously adjusting their policies to reduce deficits, also presenting further opportunities and threats for go-to-market strategies.

As businesses navigate these turbulent waters, operating margins will inevitably face added pressures. Therefore, it is essential to adopt proactive strategies that enhance financial stability and adaptability, positioning you to better withstand economic shocks. Three proactive strategies are explored: 1) Eliminating Transactional Friction: focusing on two discrete, but connected areas – Process Design and Reporting; 2) Proactive cf Reactive Processes – building an actionable contextual data driven organisation, while understanding where Artificial Intelligence (AI) adds trusted value; and 3) Change Management – ensuring that projects deliver intended results, by avoiding why many can fail in today’s operational environments.

1. Eliminating Transactional Friction

Lack of time within functional domain areas is a persistent challenge that organizations have faced for years. To address this issue, companies continuously seek to find productivity improvements that enhance business focus and maximize margins.

Whilst incremental improvements are made, operating environments continue to be volatile and dynamic. Furthermore, there are always more regulations to address, both within operating entities and across them, including for example cross border data transfers. In aggregate, these time based requirements quickly absorb any newly freed time made available to experienced staff.

As corporates inevitably continue to look for additional functional domain time, you should look to execute the following two macro initiatives by; i) critically evaluating how your processes actually operate from start to finish; and ii) focusing on sorting the inefficiencies associated with all aspects of reporting, including the resources required to execute them, identifying any typical information gaps that are hindering decision making, and evaluating whether reports are actually being used by stakeholders.

A key objective and deliverable, should be to establish more detailed operating metrics (examples below) that will help with future ROI justifications and provide a basis for subsequent year on year improvements, including those that incorporate AI.

Looking closer at Process Design and Reporting:-

1.1 Process Design

The root causes of existing friction are typically down to two things; i) timeliness of usable data received by stakeholders; and ii) data quality, both depth and breadth. On the surface, both of these would seem easy to address, so why do operational challenges still persist in abundance, and how can they be solved.

Fundamentally as a starting point, modern day technologies are now able to resolve the aforementioned two root causes of friction. This comes from today’s powerful processors being able to drive additional compute capability but, cutting to the chaste, is more about software being re-architected to leverage the power and flexibility offered by today’s processors, combined with being able to fully design processes at an ultra-granular level with all required transformations to achieve your goals. The combination of these is extremely powerful.

Modern Process Capabilities

Modern process software, as a result, is therefore able to achieve something that was not possible before. Providing the capability for organisations to define ultra-granular processes from; i) data collection; ii) thru all required data transformations to reach intended outputs; and iii) to generate actionable contextual alerts and workflows, together with supporting documentation to the level of materiality required by the process owner @anytime @anywhere + Simulations +AI + APIs.

Regarding AI / APIs: i) AI is optional, and is not necessary to drive timely rich data sets, noting that AI types include: GenAI, Bottom Up AI, and Top Down AI and ii) APIs can be used to work cross application and cross ecosystem. All “actions” can be achieved on a repeatable and auditable basis. Both of these items are explored below in more detail, including areas where their needs to be additional scrutiny for purposes of accuracy and compliance.

Data Transformation, Key Point of Friction

There is a subtlety to emphasise and appreciate here. Both transactional friction and resultant time delays often stem from the necessity of having to execute, manually or semi manually, numerous data transformational steps.

Furthermore, adding to the overall complexity, there are multiple connected and independent processes in existence at any one time, with each one requiring the collection and transformation of relevant data to support a number of important operational activities, including i) decision-making; ii) ongoing controls; and iii) facilitation of management activities.

This challenge, is further exacerbated by the fact that relevant data always resides across various sub-applications, meaning that it has to come together with transformation and reconciliation before data quality sees significant improvement.

Historically, required data transformations have resulted in the creation and reliance on multiple standalone spreadsheets, sometimes owned by different people, which additionally complicates matters from a time and execution perspective. These ad hoc standalone spreadsheets are not only difficult to maintain, but can create challenges during staff transitions, as new staff have to learn the intricacies of their design.

Unexplored Value Creation

Taking this further, the capabilities of modern day processes are not only solving many new aspects of cross application reporting, but increasingly include cross ecosystem reporting, as smart processes seamlessly connect both within and externally to different stakeholders.

In other words, it is at the internal and external intersections of applications and ecosystems where additional currently untapped added value can be leveraged for value creation. It is, to a large extent, an unexplored area for many organisations; noting that your competition could equally encroach into some mutually adjacent areas of your business or beyond.

Embedded Banking

For example, incorporating Embedded Banking is an area where cross ecosystem value can add significant value. This can be done in sales processes to fasten the pace of settlement for cashflow improvements, and also in purchases to smooth payment related processes for all stakeholders. This is a significant improvement for overseas payments, where modern FinTech financial service providers can help you add end to end transactional transparency and immediate clarity of execution cost. This resolves multiple pain points for those who extensively send funds overseas.

With particular regard to large supplier / payroll type payments, more comprehensive controls can be added to mitigate against Business Process Compromise (BPC), which is where your funds are hijacked and diverted to an account of a threat actor.

Embedded banking functionality can also enhance your treasury functions; for example ensuring that funds are available for payroll, and ensuring that any excessive cash is rolled up for purposes of offsetting to reduce interest payments, if that is made an available option by your bank.

Client Shipping

Another example is the comprehensive support for client shipping processes, which includes the capability to automatically process and manage orders. Cargo tracking systems can continuously monitor the real-time location of goods and provide transparent updates on shipment status. This real-time visibility is crucial for enhancing customer satisfaction and managing expectations effectively.

Additionally, automating customs declarations minimizes manual errors and accelerates the clearance process, thereby enhancing overall efficiency in logistics operations. This automation not only streamlines the process but also ensures compliance with regulatory requirements.

Process Ownership

Ownership and responsibilities of modern day processes are typically tiered to include process owners, and process users (day to day usage). Contrasted with older style applications, where available functionality was designed by vendors and updated perhaps annually, new processes are highly agile, specifically built for purpose, small in size, easier to protect as they have a lower footprint, and are faster to execute. Multiple process owners might exist for redundancy and mission criticality of operations (lead and supporting resources being deployed as required), again pointing to the fact that the functionality within them is highly functional and relevant.

1.2 Reporting

Once data is in the system it is logical that any report design will harness that information to generate meaningful results. Sounds obvious, but many organisations have a combination of legacy and moderns systems, and as referenced above getting that information into a usable format can be time consuming and tricky. This is especially the case, when data has been squeezed into multiple applications, using ad hoc methods. For example, some critical supporting ancillary data might not be fully validated and might exist in a non-indexed form.

Spreadsheets

Companies have different policies when it comes to using spreadsheets in core processes. Some, do not want to see them in any shape or form for fear of introducing ad hoc or design errors, whilst others do not want to lose the appeal of their broad based functionality and ease of use. Both approaches are possible.

Where spreadsheets are required, think of them as incorporating all the best aspects of spreadsheet design, without the weaknesses of being standalone. Modern day process designs can achieve this. You retain their ease of use and execute in a way that maintains full data integrity, as well as required access controls for compliance.

Critically, they also have comprehensive integrations to more advanced functionality that can be leveraged from your existing core financial systems to save you execution time. Going forwards, AI will further help here and guide you to achieve advanced reporting characteristics more easily than ever before.

Where spreadsheets are not required at all, the user interface within a process becomes self-contained with no independent access.

Reconciliations

These can be extremely time consuming, but are now of course much easier with repeatable automations, as matching is automatic. Reconciliations work well, as the process itself is ultra-granularly defined, and results can easily be reproduced on a regular and auditable basis eg credit card charge matching.

As another more in depth example, consider your current processes for agreeing inter-company balance sheet reconciliations with counterparties. This can now be facilitated by toggling between currencies, removing an area where currency values might hinder the reconciliation exercise. Furthermore, AI can extend this functionality to provide AI generated cashflow insights from bank statements.

Consolidations

As a testament to the power of these modern ultra-granular process technologies, complex financial consolidations can be automated completely or partially to the extent required, to produce consolidated reports, notes, and segment based financial statements.

Missing reporting packs can be chased / escalated, freeing up time to ensure deadlines are met with confidence. Virtual Eyeballs, as explained below in more detail, can provide contextual actionable insights with supporting documents to the materiality required by the user.

Where appropriate, it can also support a) changing M&A date changes, as well as changing corporate structures; b) controlled eye-ball sign off reviews by certain experienced personnel; and c) 3,2,1 based acquisition of data, where a complete process is deployed and controlled by HQ to the subsidiary. This is useful where expertise for an area is centralised or regionalised. The reverse 1,2,3 is where you might send reporting packs for entities to complete. They then complete them from their automated processes.

Data enrichments are an extremely powerful addition here to help with traceability. For example: i) the “source ledger” can be added to a transaction as an enrichment; and ii) in tiered allocations, additional transaction notes can be added to show the “user” how something has been calculated / apportioned for a specific transaction. This makes it easier to track back later on, particularly when trying to understand key underlying details.

Capex Budgeting

Here, copious notes can be added for justifications and operational completeness. End to end process capabilities, enable you to ensure that assets are completely and accurately amended; for example that new material assets or additions to existing assets are included in insurance processes and are included at the appropriately insured value; another example might be to execute special cybersecurity or privacy tasks for an important asset; another might be to identify building modifications to house a resized asset.

Leasing

Leasing capabilities can be a powerful addition where needed by an organisation ie properties, retail outlets, and cars / aircraft etc. This can provide a complete view of leasing types used / calculations / amortisations, and ongoing support for complex management decisions in accordance with the detailed provision options of each agreement.

Quantitative and Qualitative Processes

It should be emphasised, that above process designs can be for both quantitative and qualitative processes. Both can be designed in the same way. Low code / no code (LCNC) functionality helps reduce costs, but it is always worth providing context, in that ease of use and depth of functionality are diametrically opposed objectives and work to a point. As time passes, more advanced and sophisticated functionality will become available, for example with GenAI as explored below.

Here are some more examples of they might co-exist, but clearly all processes described above and below can have an element of both:-

Supplier Management

Highly relevant today under volatile economic conditions, supplier management is a good example of where a large number of supporting details have to be maintained:- compliance audits; evaluating various KPI / OKR performance results over time; surveys; delayed shipments; returns; and more recently the multiyear management of key indicators for sustainability reporting. Electronic signature technologies and document management systems facilitate trusted supporting processes.

Employee Self Service (ESS)   

Touching everyone in the organisation, ESS systems continue to become more sophisticated, improving work / life balance for staff by enabling tasks to completed @anytime @anywhere, and acting as a powerful conduit to share information proactively throughout the organisation or via specific groupings.

ESS can be used to communicate details about how the company is tackling initiatives surrounding sustainability, invite feedback communications on various topics (surveys; visiting colleagues; inter-departmental events etc), and also acts as a managed gateway to continually educate staff on cybersecurity and privacy best practice through regularly managed dissemination of information. The last point is another example of untapped value that exists between domain areas. Here it is HR and IT/Compliance, noting the potential to leverage existing ESS infra-structure.

All of the above examples go to provide relevant functional teams with ongoing confidence in their reporting pack submissions, with the last point’s subtlety introducing another important area for discussion. That is, how to make reporting contextual and actionable. This would make decision making more accurate, timely and precise.

2. Proactive cf Reactive Processes -Actionable / Contextual

The creation of reports has vastly improved, but at the same time businesses should make changes to ensure the usability of data, without introducing employee overload. In essence, an important transition to become a data driven company.

Reporting overload, currently exists in abundance. Reports are often cumbersome and not easily usable, simply due to their length and complexity. Frustratingly, more often than not, report creators do not know how much time is spent by others in engaging and scrutinising their deliverables in detail, but many have a hunch that some are not used much. This is despite that fact that producing some of these reports is highly labour intensive on a regular basis. Time could be saved.

Modern process technologies can help here by working for you and not vice versa. They make reports more valuable to both the process owners and process users.

Think of them at a conceptual level as being “always on”, in the background, approved + authorised Virtual Employees, with each capable of having Virtual Eyeballs and Virtual Fingers within a highly controlled environment, noting that these benefits can be achieved with or without Artificial Intelligence.

• Virtual Assistants: improve end user productivity @anywhere @anytime within the process. They can ensure that high quality information is presented to the right person(s) in a timely manner for proactive decision making. The key here is presenting data to the required materiality level, so as to streamline processing.
• Virtual Eyeballs: allow for operational variances of different types to be ranked on the fly to drive specifically aimed, actionable contextual workflows with supporting content to the materiality required. This can be for numbers in aggregate, by entity, by operating segment etc.
• Virtual Fingers: gives process owners an on-screen only message that are not printed on reports as a pointer to potential areas of trouble.

The key benefit to organisations is that decisions can move away from being triggered by month end reporting processes to be @anytime @anywhere. For example, in HR you can use this to provide insights of staff who might be overbilling but who have no planned leave. How you reach the result can be simple or complex (ie using AI) depending on your needs.

Artificial Intelligence (AI)

AI can be deployed and leveraged in the above processes to add additional value, which means that user access, data compute and storage locations, and required accuracy of output, will be included as important considerations.

However, also note that GenAI deployments could be used on a more ad hoc basis, so you will need to ensure that compliance characteristics are built in and users educated about which GenAI they can use and the implications to privacy should public versions be used, as information is entering the public domain.

Each AI type described below can be added to be within an end to end process.

Types of AI:-

• Generative AI: often abbreviated as GenAI, is a subset of artificial intelligence that focuses on creating new content across various media, including text, image, audio, and video types by utilising LLM’s (Large Language Models). This technology utilizes “generative” models to learn patterns from existing data and produce new outputs based on user prompts or speech inputs. It can provide immediate results, with an easy example being the extension of ESS processes, so that employees can use natural language to get relevant information contained within their employee handbook. Circular process management enables the fine tuning of source documents, so as to ensure clarity, so processes become more useful to staff
• Bottom Up AI: models start with raw data and learn patterns autonomously, making decisions without predefined rules. This approach, fundamental to machine learning and neural networks, relies on training data (LLM’s) to learn from examples rather than following explicit instructions. A good, easy to reference example, would be document recognition for processing incoming invoices, so these would be executing, as required, within a specific process
• Top Down AI: This is used by those who prefer applying previously acquired knowledge to focused problem solving. Top-down AI = ‘Data-efficient AI’, because operationally it needs less data. The top-down approach involves starting with a high-level understanding of the problem and then breaking it down into smaller, manageable components. In essence letting the system make decisions based as if it were human. These would be incorporated, as required, within a process. An example might be processing multiple leave applications fairly, based on multiple criteria ie seniority, other staff leave being requested for the same dates, project status, staffing levels etc. Conversely, above we discussed identifying employees who might burn out and that simple or complex determinations might be made eg No AI / AI
• Hybrid AI: A combination of the above. For example GenAI for the employee handbook, infused with specific corporate policies: an example might be maternity leave guidance over and above statutory allowances etc. Another would be Procurement: using vendor packaged Bottom Up AI for document recognition (no need to build all of your own code), but top down AI where there are complex overlapping / prioritised overriding tiers of supplier discounts available
• Simulations: As technologies have improved, their capabilities are providing additional capabilities. One such capability is the compute power to undertake simulations, as one searches for ways to diversify or consolidate business operations. This can help in your quest to identify what product mix will generate the best profitability or the mix that will yield optimum cash-flow. However, it is also an area that is under-utilised in many businesses due to understanding what can be achieved with this technology, as well as how to execute. This would typically use probabilistic and deterministic algorithms.

Deterministic and Probabilistic Algorithms

Earlier, it was stated that specific reporting results can be achieved with or without AI but the devil is in the detail. As a starting point, the word AI is a much overused word by vendors. This is because it can directly impact vendor company valuations by mentioning it, and also because only now are real macro definitions of AI emerging. Even then there are ambiguities, reflecting that it is a broad subject area.

When it comes to the deployment of AI, algorithms are deployed to achieve desired results. There are essentially two types of algorithm within each AI category: probabilistic (scalable with less false positives) and deterministic (not scalable and increased false positives).

The former is based on probability and the second is more absolute and factual (but obviously probability could be 100%). In different scenarios you want both and they can co-exist, noting that reporting in finance has to be specific AND critically that user access to data has to be controlled.

3. Change Management

Whilst the above might sound logical, achieving it on a meaningful basis can be a challenge. Let’s explore why projects do not always deliver their intended results, so that steps can be taken to avoid failure.

• Technological Mindset: Part of any inertia in moving forwards, is understanding how technologies have changed and how they might be used to release deeper value by removing operational bottlenecks and providing those harder to detect useful insights that are not so obvious. The move from static month end reporting to proactive @anytime reporting is a good example. How AI or embedded banking can be leveraged by companies, are examples of others.
• Project Team Makeup: In many cases, project teams do not reflect those areas impacted by change, and therefore important detailed scoping requirements are not fully understood at the outset. Additionally, companies can have functional or geographic operating structures or combination thereof, so they should also assess how improvements might be leveraged and sponsored by senior management in the wider organisation to maximise investments.
• Planning and Scoping: Building on from the last point, a critical deeper question emerges regarding scoping; the wider the scope, the harder the change management becomes, BUT conversely, the wider the scope, the greater the opportunity to drive value creation. From this a whole host of questions arises about senior management support in these types of initiatives. Critically one should be very aware that, i) business velocity is going to increase as frictional barriers are removed, and ii) that gaps within and between domain functional areas, whether internally or externally focused, are reducing. These, both individually and combined, provide additional stresses to your current modus operandi and opportunities to redeploy staff, but on a positive note opportunities to drive deeper value creation, and eliminate duplication, will emerge as domains get closer.
• Benchmarking: Many organisations require detailed justification processes, including ROI calculations, but this can be a challenge at the start. Consider, for example, the move to proactive reporting from reactive reporting. In most cases, benchmarks for a specific process do not exist. There are, for example, no established multi-year KPIs or OKRs (Objectives and Key Results) in existence that reference how a process works: for example i) how long it takes to achieve an end to end process, ii) the FTEs (full time employees) to achieve it, and iii) the # of transactions by type and in aggregate. Having these in place supports multiple year on year improvements.
• Systems Integrations: A combination of legacy and modern systems creates challenges when trying to achieve access to required data sets. This is often under-estimated in initial planning phases, as additional complications can arise in multiple areas.
• Data Management / Technical Considerations
  • Aspects of compliance have to be continually thought about in process design, with the fact that many requirements can be built ground up within them.
  • Management of security and privacy, including cross border data transfers will be important considerations. This includes the interconnectivity between internal applications and external ecosystems.
  • Deployment Options. Processes and Applications can be on-premise, cloud or hybrid. Today this includes the leverage of Edge Computing, where process control can be more granular right down to i) leveraging the cloud for compute purposes, but ii) maintaining storage on-premise. This results in high availability systems, real time AI processing, reduced latency and enhanced data privacy.
  • Algorithms:-
    • Own Company Design: This is going to be a new competency and whilst we touched on deterministic and probabilistic formulas above, companies must recognise that for your own in-house initiatives, where you cannot leverage established AI models, that you will need to identify and build algorithms that utilise variables that actually directly impact your intended result, as opposed to those that move as consequence of driving forces. It will likely take trial and error noting that compute, storage requirements, and running costs will change based on the number of variables etc. Worthy of mention, when it comes to process design for compliance purposes, is to consider how to use algorithms to leverage data outliers to identify areas of concern.
    • Industry Heavyweights: Algorithms can leveraged from industry heavyweights, rather than programming them from scratch eg document recognition, OpenSeek, ChatGPT, ERNIE, Manus etc. Access control, compute and storage compliance controls are critical.

Conclusion:
Enhancing financial efficiency and resilience in the face of economic turbulence is imperative for businesses aiming to thrive amidst uncertainty. The current economic climate, characterized by geopolitical tensions, inflation, and supply chain disruptions, necessitates a proactive approach to financial management. By engineering out friction in processes and leveraging modern technologies, organizations can streamline operations, improve data quality, and enhance reporting capabilities. This not only facilitates timely decision-making but also fosters a culture of adaptability and responsiveness to market changes.

Moreover, embracing innovative solutions such as embedded banking and artificial intelligence can significantly bolster financial operations, ensuring that businesses remain agile and competitive. As organizations navigate these challenges, a focus on change management, effective project scoping, and robust data management practices will be crucial in driving sustainable growth and resilience. Ultimately, the ability to integrate these strategies will empower businesses to not only withstand economic shocks but also capitalize on emerging opportunities in a rapidly evolving landscape.