Go to the homepage

FAQs

Doesn’t this require deep knowledge of forecasting and statistics?

It isn’t rocket science. In the same way that you don’t need to be an engineer to drive a car, we believe that forecasting is too important to be left to the boffins. It is a collaborative exercise that provides the foundation for business planning and which can generate… or destroy…a huge amount of value.

We have built ForecastQT so that forecast performance is transparent and insights can be shared widely, while also providing your experts with a range of powerful analytical tools.

How can you benchmark performance in my business, which has the added complexity of many very different products, markets and geographies?

ForecastQT's value added calculation allows for differences in the volatility of demand between products, which is the primary driver of forecastability. In this way forecasts across different geographies and markets can be objectively compared, using a metric that also measures the scale of the contribution that any particular forecast process makes to the business.

How do I know whether this will deliver value to my business?

Typically, the forecasting of up to 50% of SKU’s will be destroying value; ForecastQT helps you identify and eliminate this. As a result, you can be very confident that ForecastQT will pay for itself many times over.

How does measuring things differently help me add value?

ForecastQT does not measure abstract concepts; it focuses on the value that forecasting adds to a business. We also recognise that measurement is not an end in itself; ForecastQT includes a sophisticated set of tracking and analytical tools to help identify and eliminate the root cause of problems as soon as they become evident.

How does poor forecasting add 4% to the cost of production? This seems too high.

Poor forecasting costs money in a range of ways, many of which are hidden or cannot be easily traced back to forecasting using conventional costing metrics.

For instance, up to 50% of stock and therefore the associated financing and warehousing costs could be the result of over-forecasting or excessive error, which is also a major driver of stock obsolescence. Under-forecasting, on the other hand, will manifest itself in lost sales and expediting costs (last minute changes to production schedules or additional transport costs).

ForecastQT’s costing model calculates these at item-level, using the minimum of assumptions and enabling full transparency and drill down functionality.

I don’t believe that 50% of my forecasts destroy value. Where does this come from?

The failure of a significant proportion of forecasts to add value is a consistent feature of recent research [downloadable here], collating data from many different companies operating in many different markets. This finding very often comes as an unpleasant shock to managers used to seeing performance measured in an abstract manner.

Unsurprisingly, software vendors and those that have placed faith in their products ‘solving’ the problems of forecasting often find this result difficult to come to terms with. The good news is that, once it is recognised, with the right tools it is easy to remedy.

I have just invested in inventory optimisation software; why do I need another tool?

Inventory management software helps optimise inventory levels for a given a level of forecast quality – it cannot help you improve that quality. ForecastQT and inventory optimisation are complementary applications; the better the forecast, the better the inventory optimisation will work. ForecastQT should accelerate your time-to-value from your earlier investment.

If I'm already measuring MAPE, and forecast bias, why do I need another measurement tool?

The problem with most traditional forecast measures is that they take no account of the ease or difficulty of forecasting – what we call ‘forecastability’. So, for example, an average error of 10% might represent excellent performance if it relates to a product which is very volatile, but if it relates to a stable, easy-to-forecast product, it could be crippling – worse than no forecast at all. Also, in the supply chain, what matters is the quality of forecasts at the lowest level – since these drive replenishment. Because they are aggregated, high-level measures such as forecast bias, usually fail to expose problems at a low level.

Isn’t ForecastQT a ‘nice to have’ rather than a necessity?

There is plentiful evidence that most forecast processes, even those produced by acknowledged experts, struggle to add value much of the time and that most judgemental interventions in the process actually make things worse rather than better. Without using a Forecast Performance Management tool (like ForecastQT) operationally, there is no way to know whether any part of the process is delivering business benefit – and if not, where it is breaking down.

Isn’t this capability available in forecasting packages?

ForecastQT is the first and only application with the functionality to measure and manage forecasting performance operationally, in a business-relevant manner, focusing on whether, where and how forecasting is adding value.

We already have a forecast optimisation capability; how can ForecastQT improve on that?

Forecasting optimisation capabilities are usually based on identifying the algorithm that fitted best to history, despite this being well known as an unreliable guide to future performance. This is because in this process, any recurring signal in the data (which can be forecast) could be heavily masked by noise (which cannot), which skews the result.

The only way to assess whether a method is forecasting adequately is (preferably immediately) after the event, using a metric that takes account of forecastability, like that provided by ForecastQT.

What makes CatchBull different?

We don’t produce forecasts, and ForecastQT is not an afterthought bolted onto another software application. This means that we are free to adopt a cold-eyed, pragmatic approach to forecast quality; we are not promoting a proprietary forecasting tool or method and we aren’t trying to hide any embarrassment when these fail to deliver on their promises.

Finding ways to improve forecasts is all we do and we have made a number of major contributions to the science, which are all embedded in ForecastQT.