Cost reduction: how to optimize with AI

While the profitability of your business depends in part on the results of negotiations with your suppliers, there’s one parameter that’s all too often overlooked, but which can pay off handsomely: the comparison made between invoices received and initial estimates. In fact, while validating a negotiated rate is a first satisfaction, you need to be able to turn this success into a means of profitability and cost reduction. If you are faced with managing a substantial purchasing budget, or responsible for negotiations for an entity made up of several profit centers (integrated or franchised), then read the following carefully, because you’ll be able to go from successful negotiation to cost reduction in less than three minutes.

Once you have validated the pricing terms of a quotation, once a partnership has been signed around the supply of services or requirements in exchange for a price deemed reasonable by both parties, the time has come to fulfil the reciprocal obligations. Your supplier will undertake to deliver or supply the service as quoted, and you will undertake to pay for it in accordance with the terms of the quotation. Once your supplier’s obligation has been fulfilled, the first thing to do is to check that the goods have been delivered or the service rendered, which generally precedes receipt of the invoice for the service.

Once a quotation has been validated, you can take advantage of the potential of AI to turn a document, whose formalism depends on your supplier, into a compilation of raw data ready for comparison. Then, when the invoice is received, data cross-referencing will consist of comparing the pricing lines mentioned on the invoice with the rates negotiated and validated on the initial quotation. It’s at this precise point in the process that AI can help you cut costs, because the idea is not just to compare the amount of an invoice with the prices of an estimated service, but to spot the difference in record time:

• Differences between the content of the service negotiated and the content of the service provided or delivered
• The origin of the discrepancies that explain the difference between the invoiced amount and the negotiated rate
• problematic billing lines
• Data that are not consistent between what was planned and what was delivered (e.g. price per kilo changed by a few centimes, transport costs, VAT error, etc.).

AI is better than a simple excel spreadsheet at extracting the marginal data from an invoice that is at the root of the discrepancy between a quotation and an invoice. AI does more than compare erroneous amounts; it identifies the precise lines of the invoice that contradict the conditions negotiated upstream. In less than three minutes, you can not only identify an invoicing error, but also justify it.

This ability of AI to conduct an investigation in record time isn’t just about identifying unscrupulous suppliers or inattentive purchasing managers. It’s a tool that benefits all parties involved in a purchasing process, and we explain why:

• If a price has been negotiated on the basis of a price per kilo, but the AI identifies a discrepancy between the quote and the invoice received, this may mean that either the supplier has not informed the customer of changes in the market price, or that the buyer has not validated the setting of a price independent of market fluctuations. In this case, AI is a tool that can be used to review future purchasing conditions.
• If a price has been negotiated at the level of a group of several franchised profit centers, each entity will be able to benefit from the expertise of AI, which will compile in record time all the invoices received by each profit center. Once the data has been extracted and a comparison made between the conditions negotiated by head office and the invoices received individually, the penalized profit centers will be identified in a matter of minutes. You can imagine the benefits, in terms of credibility and cost reduction, for a head office of being able to identify the cause of a financial loss and inform the members of its network in real time.