 |
| David Cuykendall Click for Index Page |
Financial information is not homogenous
One kind of financial information has a historical or results perspective. Its purpose is to record what has happened in the past, and it is used by people outside of the organization — bankers, creditors, fiduciaries, regulators, and taxing authorities.
A second kind of financial information has a right now time perspective. Internal workers making day-to-day, real time decisions about the organization’s business processes need this kind of information.
A third kind of financial information has a forward-looking predictive perspective.
Standard packaged financial management systems based on transaction processing are biased toward delivering historical or results only financial information and the analytical model they deliver relies on deductive inferences driven by extracting useful information from massive quantities of transactions-based data.
This is the compliance oriented, and operationally commoditized side of financial management. The information provided by these systems is incompatible with the side of financial management that is concerned with forging profit enhancing entrepreneurial efficiency.
Further, the second and third types of financial information may depend on heuristics that rely on knowledge about patterns in the environment and simple rules to derive actionable financial information that is used to help forge entrepreneurial efficiency.
The term ‘heuristic’ is of Greek origin, meaning ‘‘serving to find out or discover.’’ Heuristics need to be distinguished from analytic methods. Analytical decision making depends on finding answers through analyzing massive quantities of data, the handling of which is computationally expensive. In contrast, heuristic methods seek ‘good enough’ answers and depend on simple rules.
We often have to rely on heuristics not simply because of computational expense of the alternatives but also because of the task environment. For instance, chess has an optimal solution, but no computer or mind, can find this optimal sequence of moves, because the sequence is computationally intractable to discover and verify.
Many problems of interest are computationally intractable, and this is why engineers and artificial intelligence researchers often rely on heuristics to make computers smart.
The development of an understanding of why and when heuristics are more accurate than strategies that use historical transactions data and a lot of computation rests not in the heuristics alone, but in the match between the heuristics and the environment.
Simple heuristics are best understood from the perspective of pattern recognition, where there are many examples of how biased inductive inferences can predict more accurately than unbiased deductions, provided the ‘bias’ is based on a pattern that is ‘ecologically valid’; that is — if the pattern the heuristics rely upon accurately match the environment from which the pattern is drawn.
Many problems of interest are computationally intractable, and this is why engineers and artificial intelligence researchers often rely on heuristics to make computers smart.
The development of an understanding of why and when heuristics are more accurate than strategies that use historical transactions data and a lot of computation rests not in the heuristics alone, but in the match between the heuristics and the environment.
Simple heuristics are best understood from the perspective of pattern recognition, where there are many examples of how biased inductive inferences can predict more accurately than unbiased deductions, provided the ‘bias’ is based on a pattern that is ‘ecologically valid’; that is — if the pattern the heuristics rely upon accurately match the environment from which the pattern is drawn.
However, if the pattern in the environment is unknown — meaning it remains to be discovered —computationally expensive analytical deduction methods are necessary. In that context, the data contained in standard packaged financial management systems and the transaction processing sub-systems that feed them can be leveraged.
Example of transactions-based data analysis
An example of a decision support model that relies on extracting useful historical information from massive quantities of data stored in ledgers and transaction processing sub-systems is Intrawest, the largest ski operator in North America. Here, it is ‘ecologically valid’ pattern that remains to be discovered.
Interwest uses query tools and highly configured report writers to analyze massive quantities of historical transaction-based data captured from its ski resort guests to determine the value, revenue potential and loyalty of each customer so managers can make better decisions on how to target their marketing programs.
These query tools and report writers segment their customers into categories from “passionate experts” to “value minded family vacationers,” etc. The company then emails video clips that would appeal to each segment to encourage more visits to its resorts.
Example of heuristic analysis based on non-transactions-based data found outside of the general ledger system
Interwest uses query tools and highly configured report writers to analyze massive quantities of historical transaction-based data captured from its ski resort guests to determine the value, revenue potential and loyalty of each customer so managers can make better decisions on how to target their marketing programs.
These query tools and report writers segment their customers into categories from “passionate experts” to “value minded family vacationers,” etc. The company then emails video clips that would appeal to each segment to encourage more visits to its resorts.
Example of heuristic analysis based on non-transactions-based data found outside of the general ledger system
An example of a heuristic decision support activity that relies on accurate patterns drawn from the task environment is how the shipping concern Voyage estimates financial and technical voyage details.
Voyage's financial calculations based on simple heuristics include: ship/time costs (fuel, labor, capital), freight rates for various types of cargo and port expenses. Technical details include factors us as: ship cargo capacity, speed, port distances, fuel and water consumption and loading patterns (locations of cargo for different ports).
Voyage's financial calculations based on simple heuristics include: ship/time costs (fuel, labor, capital), freight rates for various types of cargo and port expenses. Technical details include factors us as: ship cargo capacity, speed, port distances, fuel and water consumption and loading patterns (locations of cargo for different ports).
Conclusion
One kind of financial information has a historical or results perspective. Its purpose is to record what has happened in the past. It is used by people outside of the organization — bankers, creditors, fiduciaries, regulators, and taxing authorities. A second kind of financial information has a right now time perspective; internal workers making day-to-day, real time decisions about the organization’s business processes. A third kind of information has a forward-looking predictive perspective. It is used for entrepreneurial and strategic purposes.
When a vendor sells you standard packaged financial management software, you are not told that to make your financial management truly effective, there remains something to be discovered, and that the discovery may require conceptual innovations that have little do to with the vendor’s product. That something may be, for example, simple heuristics (rules of thumb) that directly support your decision making and monitoring. These may depend on data from sources outside of your ledgers, data whose existence is easily overlooked.
Your vendor is unconcerned because he is ignorant in the matter and thinks only in terms of historical, transactions based data. This is data that is useful for deductive inferences to derive patterns from categorization, aggregation, and segmentation, or to trigger alarms. But this little exhausts the need for the second and third kind of financial data that help forge the operational and entrepreneurial efficiency of your enterprise.
