Scoring models for assessing supplier reliability
The right choice of a business partner is one of the main tasks of any business. The quality of goods or services that the company receives, as well as its reputation, depend on the reliability of the supplier. To make the process transparent and convenient, scoring models for assessing reliability come to the rescue.
Traditional approach to supplier evaluation
In medium and large companies, the function of checking potential counterparties is performed by the information security service. The specialists of this department conduct a comprehensive analysis of each participant who has applied to participate in the procurement, assessing various aspects of its activities and potential risks of interaction. As a result, the security service provides a conclusion on the possibility of cooperation with this counterparty.
Despite the high efficiency of this approach, it has a significant drawback - checking one counterparty can take several days and requires significant resources.
Optimizing Supplier Evaluation with Scoring Models
Cost optimization is partially possible if only companies that have passed the filter of preliminary automated checks at the market analysis stage are allowed to participate in the procurement. This approach is based on the use of scoring models of assessment, which, despite their modern association with artificial intelligence, have a half-century history.
Scoring is a calculation of points. Points are awarded for positive signs and deducted for negative ones. For the scoring model, a necessary and sufficient set of data is selected for verification. Each type of data is assigned a certain weight coefficient, which allows ranking factors according to their degree of importance.
Key groups of features in assessing supplier reliability
As a rule, the following groups of features are distinguished to assess the reliability of a commercial company:
1. Supplier reputation. This group of features usually includes factors that directly or indirectly characterize the company's status, its solidity, fame and integrity:
- Length of time on the market
- Current and Historical Litigation Amounts
- Evaluation of capital and reserves
- Payment and tax discipline
- Credit rating
- Facts about license revocation
- Feedback from customers
Particular attention is paid to stop factors, upon detection of which the reliability assessment service issues a recommendation about undesirable cooperation with this company:
- Inclusion in the Register of Unscrupulous Suppliers
- Going through bankruptcy proceedings
- Suspension of operations on the account
- The presence of the manager in the register of disqualified persons
- Excess of the amount of litigation over revenue
Read more materials on this topic in Compass CIO
2. Data on the company's financial stability
According to the legislation of the Russian Federation, all companies are required to maintain accounting (financial) reports. A commercial secret regime cannot be established with respect to these reports. Thus, it is possible to control the financial health of the company based on the calculation and analysis of the following indicators:
- Scale of operations (revenue, assets)
- Revenue dynamics
- Business sustainability
- Income of future periods
- Interest load
- Profit and profitability
The lack of financial reporting will be a serious cause for concern.
3. Signs of fraud. Sometimes it happens that companies are created not to conduct real business, but to implement various types of fraudulent schemes, for example, to receive an advance from the customer and then liquidate it. Here's what you should pay attention to:
- Revenue and asset analysis
- Availability of fixed and working capital
- Average number of employees
- Checking against various registers (mass address, mass manager, director and founder, blacklists, tax arrears)
- Length of time on the market
- Signs of real activity (SRO, availability of contracts and licenses)
- Assessing the reliability of affiliated companies
Warning signs may include, for example, the absence of tax reporting and inclusion on so-called “black” lists.
4. Company experience. The best proof of the quality of the supplier's work is the number of contracts concluded with it, as well as a number of additional features that can be calculated based on the data on the supplier's contractual activities:
- Number and dynamics of completed contracts
- Dependence on one customer or contract
- Experience acting as a customer
Lack of information on completed contracts or inclusion in the register of unscrupulous suppliers are serious stop factors that characterize the possible lack of experience or negative experience in conducting contractual activities by the supplier.
Data quality is the foundation of an effective scoring model
The following characteristics of the quality of the initial data, on which the effectiveness of the scoring model depends, are most important for evaluating suppliers:
| Relevance | The data must reflect the actual state of the company at the time of assessment. |
| Availability | The data must be accessible for collection and be “legally clean”, excluding the possibility of prosecution for disclosure of state or commercial secrets |
| Reliability | Data must be obtained from official sources. |
| Necessity | The model should include the maximum number of the most important features for assessing reliability. |
| Adequacy | The model should not include insignificant features or features with a high degree of correlation with other features. |
| Clarity | The user will have more confidence in the model's evaluation results if he understands what parameters were used to evaluate the company. |
How Data Turns into Decisions: Visualization Tools
No matter how high-quality the model is, it is necessary to take all measures to convey this information to the right point of the business process in a form that is most convenient for making a management decision. A good service implies built-in data visualization tools presented in key assessment sections based on the analysis of the user path configuration.