Intelligent Costing- Further Refined
A joint collaboration between Joerg Millotat & Benjamin Wann.
Psychologists define "intelligence" as the ability to learn from experience, solve problems and adapt to new situations.
Artificial intelligence, simply explained, is the attempt to transfer human learning and thinking to computers and give them this intelligence. Instead of being programmed for every purpose, an AI can find answers and solve problems independently.
Applied to costing, this means that people involved in costing have the ability to understand costing as a constantly changing process that needs to be adapted to changing conditions and factors. This can solve the resulting problems by constantly expanding knowledge to adapt to new situations.
Artificial intelligence supports humans by predicting changes in good time, disentangling complex systems, and thus making interdependencies and relations visible. In this way, factors that influence outcomes are identified as they arise and take effect. AI creates transparency in cause and effect and makes process factors and parameters measurable and controllable.
As a result, Intelligent Costing informs about opportunities and risks by providing information about the costs and performance of the company's current business situation. Not only data of the own company is processed, but additionally data from the competitive environment and the overall economic situation. These measures enable the company to gauge its competitiveness to create timely measures to increase it.
The opposite of intelligent costing is an approximation- To use rough signals to determine overall product profitability by SKU, customer, and channel because a lack of subject matter expertise, technical limitations, and poor system design limits the current means. The profit results are unknown and unknowable due to these factors, yet organizations must make strategic decisions based on these approximations.
Basic organizational requirements for Intelligent Costing
The quality of cost accounting depends on the quality of the required data. To ensure this, the following accounting organization is required.
A complete chart of accounts for cost accounting.
Timely posting of documents utilizing electronic document entry.
Consistent organizational integration of all departments to the accounting and cost accounting system (usually ERP).
Orchestration of data collection and data entry of data relevant to cost accounting. In particular, the integration of a store floor data collection system for the purpose of timely transmission of production data to cost accounting.
A cost center accounting system that reflects the functions of the company. Collective cost centers are to be largely avoided.
If allocations are used, they should be 100% transparent and understood to represent the internal use of achievements cause-fairly on the cost center level. We believe the default should be cost traceability instead of grouping and allocating buckets of costs. There may be reasons not to do full traceability (salary confidentiality, for example), but the cost buckets should be small and similar in those cases.
In many organizations, overhead costs are substantial. Without understanding each cost line and how it supports the business, allocated costs can be overlooked to determine if they are truly needed.Cost allocation should be avoided wherever possible, instead defaulting to traceability and cause-effect % splits where 100% traceability cannot be determined. FP&A Software tools help to set these rules for each type of cost
System technical requirements for Intelligent Costing
Cost accounting is divided into different functions, such as:
Standard Costing
Absorption Costing
Planned cost accounting (German term literally translated)
Actual cost accounting (German term literally translated)
Product costing
Quotation costing
Final costing (German term literally translated)
Costing during product development (German term literally translated)
Life cycle costing
Economic efficiency calculation (German term literally translated)
Investment appraisal
Profitability Analysis
What if analysis, scenario planning
All functions must use the same database. This must therefore be uniform, consistent, and free of redundancies. To meet these requirements, data must be stored in a database management system such as, e.g., SQL Server, Oracle, or abab.
Further data processing in subsystems should occur without system breaks via interfaces. A robust database management system connected to the ERP is key. Organizations often get stuck in over-customizing their ERP- it is expensive and resource intensive- FP&A software should be used to collect and synthesize supplementary data instead of trying to feed it through the ERP first. The ERP should collect good data on production and costs going through the GL, not also try to be an analytical/reporting tool.
Drill-down functions allow data to be traced down to the document level. Excel should, therefore, only be used as a presentation tool. However, this purpose is better served by modern BI and reporting software such as Power BI or Tableau.
The most relevant systems of Intelligent Costing are Business Intelligence and Machine Learning (Deep Learning) software. With these systems, the causes of cost generation, their factors, and the interrelationships of success-reducing processes and their factors are identified and controllable in the company.
There is a significant opportunity to highlight how FP&A software can improve this functionality to make it easier, more approachable, and less technical than before. FP&A software fully integrates with Excel for analysis, and BI tools can still be used. Still, they should be plugged into the FP&A software to produce more meaningful insights instead of being connected to the ERP/Database, which is unprocessed data- not information.
The best way forward is to provide 2 alternatives, organizations can do all of what you have prescribed in this document using the ERP with customization/databases/BI tools, OR they can fastrack the process by investing in the right FP&A software- that way, we are presenting multiple options.
Product success on the market, costs, prices, and margins define the overall success and profitability of the company. AI software opens up the possibility of forecasting the development of market-relevant data. The dependencies of cost-determining processes can be analyzed and presented in context. In this way, opportunities and risks are identified in good time.
Measures can be developed effectively based on empirical data.
Examples are:
Timely recognition of the inefficiency of plants
Basis for reliable investment decisions
Recognition of success drivers in production processes
Identification of success drivers in the entire value chain
Timely avoidance of risks
Creation of transparency of factors in the awarding of contracts for customer projects
Exploitation of market prices
Reliable control of costs and margins in complex projects
Prerequisites for the qualification of employees for Intelligent Costing
The meaningful use of modern systems for Intelligent Costing requires a division of tasks between employees.
This involves departments outside of cost controlling.
Market analysis (marketing)
Sales (provision of customer information such as prices, quantities, and payment terms)
Production (all relevant production data)
Project management (project data)
IT managers (creation of interfaces, queries, data pools, data warehouses)
Within Controlling
Data Analyst
Data Scientist
Costing Manager
Management Accountant
A Senior Management Accountant or Director manages these functions.
Recommended Reading
Intelligent Costing- How the Software Landscape Must Look to Succeed