Customer and supplier perspectives on contracting for GIS services

Product Quality
There are many criteria that must be considered when assessing the quality of converted data. Because of the nature of the work, a certain amount of subjectivity can come into play when describing the quality of conversion. Wherever possible, this subjectivity should be worked out of the quality evaluation process and replaced with defined, measurable standards. This helps to assure the buyer that they are getting the product they paid for and it is effective in communicating to the supplier those characteristics that are most important to the customer.

Buyer’s Perspective - Most importantly, in the context of product quality, the buyer needs to be assured that they are getting what they pay for. They need to know that the product will function effectively when used as part of the various applications that are going to provide the project payback. A fimctionality driven focus on product quality will result in some aspects of the data being more important than others. The system used by the customer to score the data should take this into account. The customer has a right to expect that the quality of their record system will not be degraded as a result of conversion, but will more than likely be enhanced as a result of the complete inspection the records will undergo as part of the conversion process.

From the buyer’s perspective, the contract also needs to address what happens if a deliverable does not meet the customers specifications. Returned work has the potential of being very disruptive to the flow of data within a conversion project. A process should be defined which reduces the potential for this kind of disruption. Supplier’s Perspective - Because of the nature of the work, the amount of human involvement in conversion, and the vast amount of detailed information contained in a utility’s record systems, the final product will never be perfect. With the proper processes, however, it will be very accurate, achieving scores of 99% or better. Requiring perfect data in certain areas from a conversion vendor is only feasible when automated tests can be employed to check these areas. If the customer has a series of automated tests they will be using to check the product, the supplier will want these tests as well so they can run them prior to delivery, thus ensuring their success at the customer’s site.

It is very important to the conversion supplier that clear, objective, measurable standards are established for evaluating the quality of the converted product. If the deliverables are not consistently scored, or if too much subjectivity is allowed into the evaluation process, then the vendor may, in effect, be shooting at a moving target. This makes it difficult to train staff on the project and to provide them with meaningful feedback. Particularly early on in the project, it is important to get thorough, timely feedback from the customer, so that problems can be worked out of the conversion process, and rules can be established to convey the customer’s expectations to the conversion staff.

Specific Contracting Mechanism - The conversion contract between MidAmerican Energy Corporation and Supplier addresses quality issues in a thorough, well-defined manner. The acceptance criteria are very complete and cover all pertinent aspects of product quality. The following items summarize the MidAmerican energy quality criteria:

The data must load successfully, or the shipment is returned to the vendor with no fiu-ther evaluation.
Spatial accuracy tolerances are defined and rules for evaluating and scoring the spatial accuracy of the product are detailed.
Attribute accuracy criteria and scoring mechanisms are defined. Different attributes are weighted differently, depending upon their ultimate importance to the successful fimctionality of the system.
Database relationship criteria and scoring are established.
Systematic errors (errors due to programming or to incorrect specification interpretation, which cannot be characterized as random) may result in rejection of the batch, at the discretion of the customer’s Project Manager.

Initially, all delivered data is completely checked. As confidence in the quality of the product on the part of the customer increases, the verification moves from 100°/0 to spot checking. This spot checking is carried out on 10% of the features that makeup a deliverable. The features to be verified are determined on a random basis. Each incremental delivery is scored in detail, with an individual score given to each feature type (conductor, transformer, fuse, etc.). This provides good feedback to the conversion supplier, helping to focus attention on specific areas where training or process modification may be required. The pass/fail determination for the incremental delivery as a whole is made based on the average of the individual feature scores, weighted for the relative number of each feature type within the delivery.

The conversion contract between MidAmerican and Supplier contains a relatively unique section that addresses incentives and penalties for product quality. The following is a portion of that section. Upon receipt of deliveries from the supplier, the customer will perform a quality audit as described (elsewhere in the contract). Customer will not reject supplier’s data as a result of substandard quality, but will initiate penalties in the form of credits to the customer. For each percentage point data falls below 98% accuracy, a credit will be due to the customer calculated at 4?X0of each delivery’s billable unit value. Additionally, for each percentage point data falls above the 98’-XO accuracy, a performance bonus will be due to the supplier calculated at 4°/0 of each delivery’s billable unit value.

If the quality of the product is not at specified levels, customer personnel perform corrections to the extent they feel they are necessary. This minimizes the potential impact that a batch rejection can have on the overall delivery schedule, and also gets the individual batch implemented as quickly as possible.

This procedure is not followed for the first couple of batch deliveries for a project or work area, where the specific feedback gained from the return of marked up plots to the supplier is most important. Even after the procedure is invoked, and even for batches that are of acceptable quality, there maybe value in providing specific feedback to the supplier. In these cases, the results are reviewed together and discussed.