Value Engineering and Cost–Benefit Analysis for AI-enabled Digital Programs

Understanding

The power of Artificial Intelligence (AI), which shall enable multiple facets of business and digital ecosystem has been strongly anticipated for a while. Substantial research is being undertaken on the AI models, algorithms and associated technology. However, majority of AI experiments are yet scale and move beyond silos to enterprise-wide usage. Often, companies engage in AI proof-of concepts without having a viable way of assessing its business value. A holistic assessment of the company’s value chain and where AI can bring value (new revenue, cost savings, time savings through efficiency or process automation) is necessary for companies to adopt AI at scale and to realize business value of embedding AI in its value chain. The methodology of value engineering or cost-benefit analysis to assess the strategic and financial aspects at the onset of AI-enabled digital program, pursues this solution.

Medium to large AI-enabled digital programs require systematic assessment and estimates for business decisions given higher investment involved to scale. Cost-Benefit Analysis would facilitate AI-enabled digital to become part of strategic initiatives of CEO and to be included as line item in financial budget of CFO as well as to develop business justification for investors and C-level executives keeping a long-term vision. Systematic identification and estimates of benefits, costs and monetary implications along with value drivers and cost drivers show a holistic view of AI-enabled digital for sound business and investment decision-making.

Cost-Benefit Analysis is a systematic process for comparing benefits and costs of a project/program to determine if it is a sound investment (justification/feasibility) and to see how it compares (ranking/priority assignment)

An immediate outcome of the cost-benefit analysis is clear understanding of the benefits that can be obtained by value drivers of AI-enablement, both tangible monetary and intangible, and costs that will be incurred to achieve the benefits over a longer time horizon. Being in concrete monetary terms, benefits from value drivers and investment schedule calculated from cost drivers help to develop financial model and cashflows for 5 year or longer time horizon of the AI-enablement program. This is immensely useful as the financial model and cashflows form the basis of decision-making financial metrics for investments in AI-enablement program, which are familiar with C-level stakeholders and investors.

• Internal Rate of Return (IRR) - What is the estimated IRR for AI-enabled digital projects in comparison to internal company / industry benchmark?
• Payback Period - What is the estimated payback period from the AI-enabled digital?
• Net Present Value (NPV) - What is the net present value of the AI-enabled digital projects for 5-year period? Is NPV positive or negative?

Since cost – benefit analysis is a value chain based holistic assessment, it also considers risks associated with the value drivers. With a clearer picture of costs and investment schedule now available, it enables the corporate finance and FP&A team to do appropriate budget and capital allocation for different years of the AI-enabled digital program.

An additional benefit of performing value engineering or cost- benefit analysis at the onset of AI-enabled digital program is that, now there is a framework in place with business and financial metrics to perform annual or bi-annual monitoring of the program including implementation and impact risks. This helps executives to make business decisions or take necessary actions in case of deviations while the AI-enabled digital program is in place. Overall, with these tools, methods and metrics, AI-enabled digital becomes part of the senior management of the business or organization.

Approach & Methodology

       I. DISCOVERY:

The first stage starts with performing a due diligence and asking key questions about the company or organization whose leadership is looking to embrace AI-enablement throughout its ecosystem. This is the discovery phase which involves conducting workshops and interviews about overall objectives of AI-enablement with executive stakeholders involved and find the current state. This is a joint exercise between the firm and the business which intends to embark AI-enabled digital. The personals who are the involved in this phase are the key executive stakeholder along with representatives of different departments (finance, marketing, operations, technology) who would be benefitted by AI-enablement. Some example questions that would facilitate discussion and discovery in this phase are,

• What are the overall goals of the organization?
• Who are the key stakeholders and the executive sponsor for AI-enabled digital program?
• What is the current state of activities/ services?
• What are the key value drivers/activities associated with AI/Digital intervention (employee, operations, customer, technology)?
• How the key value drivers can be measured and converted into monetary terms (in dollar value)?
• Who can provide quantitative values/ data for key value drivers (finance/ marketing/ operations, market research)?

An outcome of the discovery workshops, due diligence and interviews is clear understanding of current state along with identification of company specific value drivers which would be benefitted by AI-enabled digital. The value drivers enlisted show tangible and intangible benefits along with KPIs. The tangible benefits are worked upon to convert to monetary values for individual value drivers based upon the internal company inputs as well as external market research data. The quantified tangible benefits in monetary terms form the basis of benefits modelling of cost benefit analysis for AI-enablement. The benefits model differentiates the value drivers into new revenue or cost savings and breaks them into categories – customer, employee, operations, technology. Data plays a key role in the value drivers whether identifying data-driven insights leading to process improvement or using data through pipeline for AI-driven models and automation. The benefits model further considers the high, low boundaries along with baseline for scenario planning for individual value drivers which comes from subject matter expertise.

     II.           FOLLOWUP:

After the value drivers for AI-enabled digital have been identified based on company’s objectives and long term goals, the next phase is ideating and solutioning the value drivers and come up with the cost drivers that would be required to implement them along with cost estimates and YoY investment schedule. It starts with,

• Ideate AI/Technology solution to achieve key value drivers – During this step, high level solutions for each of the value drivers of AI-enablement and overall solution architecture are ideated. This is high level solution architecture considering the analytical and technological requirements for implementing AI-enabling value drivers.
• Identify hardware, data, AI cloud infrastructure, labor/ human capital required for building and using the solution – Once the high-level solution architecture is in place, hardware, data, AI cloud infrastructure and labor requirements are identified to implement the solutions at scale. Both the components and their quantities are estimated.
• Estimate hardware, external data, cloud, labor costs (CapEx/ OpEx) for multiple years/lifecycle – In the next step, for the components that has been identified for AI-enablement, their costs are calculated for multiple years based on quantity and usage estimates.

The above mentioned 3 steps form the basis of cost drivers and cost estimates which is converted to YoY investment schedule required for implementing the value drivers at scale as part of AI-enabled digital. Data forms the major raw material in case of AI-enablement. If the company has data internally, the costs incur in collecting, storing, processing the data while external data costs are accounted for the value drivers that require third party data. Infrastructure, storage, compute, security and applications with subscriptions are the primary components of AI-cloud infrastructure costs. In terms of human capital costs, the substantial costs components are full time employees (FTEs), professional services, training or up-skilling and ongoing system administration costs along with costs involved if external contractors are hired. Machine learning models and algorithm development costs form a part of human capital costs which can be either internal or external analytics services based. Some of these costs can be subtantially improved by subscribing to AI-enabled products which are offered by product companies.

Once the monetary benefits for value drivers and the YoY investment schedule for cost drivers are in place, it becomes concrete to perform financial analysis on AI-enabled digital program. The value drivers, benefits modelling, cost drivers and investment schedule forms the building blocks of the financial cashflow model for AI-enabled digital program. Based on monetized benefits and YoY investment schedule, the cashflows for the different years of the AI-enablement program is obtained. The cashflows are fundamental for deriving the values of the 3 financial metrics (IRR, Payback period, NPV) as part of for assessing financial aspects of AI-enabled digital program.

Further, as the financial model with cost, cashflows, and boundaries of low and high scenarios is now in place at the onset of the program, this can now be used for ongoing monitoring and accounting as part of assessing AI-enabled digital program performance as well as implementation and impact risks.