Traditional FP&A systems are typically focused on one aspect of the management process. For example, setting...
It’s a shame that many companies do not put the full capabilities of IFP&A into use. Some seem content to use modern systems to replicate the same old dashboards and fancy reports they have always produced, while others fail to recognise that planning and analysis in today’s fast, complex world requires a different approach from that of 10 years ago.
Fortunately, an increasing number of organisations have come to embrace what IFP&A can offer. These people ‘step outside the box' and create solutions that are both innovative and that help management to improve organisational performance.
Below are a few examples that incorporate AI techniques to guide managers along the best course of action, which they would not have been able to do using traditional approaches.
Increasing the accuracy of forecasts
This company produced goods that were sold to large manufacturers as well as retail outlets. The products were typically seasonal and whose lead time for sales was a few months.
The problem they had was in collecting accurate forecasts for the different product streams in order to maximise the efficiency of production. Like most organisations, sales teams would always forecast the end of year budget figures, as no one wanted to be seen to fail or have targets increased during the year if things went too well.
The FP&A managers’ solution was a combination of trend and correlation information that accessed 2 years of historic data for each product revenue/geography stream. As each sales manager entered their forecast for the rest of the year, the system would instantly give them a probability rating of how likely that was, along with a suggested revised target. The managers could then still submit their original forecast, or a revised one, but with an explanation as to why they thought the figures submitted were the ‘real’ ones. The system was a huge success and resulted in considered forecasts being submitted by which senior managers could set production or develop actions that took advantage or mitigate the business environment they were in.
This company offered services to consumers that were typically bought a few times each year. The market was very price sensitive and so it was important to ensure that the price of the services they offered were the best value at any time of the day. Due to the Internet, the prices of competitors could be monitored. They also had a good idea using market research, the kinds of sales levels they could expect.
The IFP&A solution took in data at set times of the day by service and region, which included both their own sales and the prices of competitors. If sales went below a set threshold during the day in any location and that price wise they were more expensive, the system automatically recalculated the optimal price for that location and send it to the local manager concerned. If accepted by the manager the system would update the online prices for new customers - something that could happen several times a day. This approach could never have been achieved manually. Additionally, the historic trends it generated helped management devise pricing strategies that resulted in more reliable profit forecasts.
This example allowed management take a different approach to planning. The organisation is capital intensive but whose equipment could be swapped from one type of production to another but at a cost and a delay before output could be restored. They had several production facilities throughout the world but to save on transport costs they would try and produce products near to customers.
The system would monitor forecasts and compare this with stocks already held in the different production facilities. If one product line was in danger of running out then the model would simulate whether it was cheaper to change the production schedule with its inherent costs and delays, or ship in products from other locations. Alternatively, customers could be offered an incentive to delay deliveries.
Whatever was modelled as being the most profitable for the company was then initiated.
What made this example different is that the system wasn’t so much about setting local targets but treating the whole company and its results as being the main objective. By not ‘penalising’ a local unit when not producing, the company was able to plan and ensure everyone felt they were acting as a single entity that together served the overall organisation’s interest.
The article was first published in the D!gitalist Magazine