Sales forecast

As well as generating forecast information, forecast requirement budgetary numbers can be generated (different from Sales budgets) – that can be used to measure the accuracy of what the buying department thought would be needed compared to actual historic sales.

A facility is provided to define annual forecast targets and allow manipulation of buying and selling prices and applicable factors, both for existing items and for unapproved new items. Unapproved new items are recorded on the item masterfile, which decodes on the database management module.

Budgets can be calculated in quantities and values. Forecasts can be based on existing budgets or forecasts or sales history. If the definition is based on a budget or a forecast, this source needs to be identified to the system via the forecast source ID.

There is a facility to forecast at branch level or company level or in total for all Branches/Companies.

Previous items historical sales data can be retrieved to take into account supersession history.

To facilitate the manipulation of generated budget forecast quantities, the forecasted quantity can be copied to another forecast ID.

Marketing factors in the budgeting process are independent of the (Price Simulation Module) PSM pricing system, except for costing factors and exchange rates. The pricing system on which the generated forecasted values are based can be viewed within the forecast maintenance program.

A minimum margin can be defined by item hierarchy. Multiple minimum margins are allowed for each Level, with the help of effective and expiry dates. This facility resides on the Database Management module.

Once a forecast has been calculated it is broken down over a twelve month period, and seasonally adjusted. The seasonal patterns are defined by Forecast Profile codes. The forecast quantities are automatically adjusted when the forecast is generated for a period less than one financial year.

On approval, a forecast can be transferred to the sales history file. This is done via a transfer program, requiring specification of the budget’s financial year, and the Bond/free/composite code. Once a transfer has taken place, maintenance to the forecast is no longer allowed.

The Forecasted data can then be used in Purchase Planning where it is the Demand figure used in the calculation to determine the Suggested Order Quantity.

Management reports can be created in the normal financial year and, if required, in the Head Office financial year. For example: The Head Office financial year is from April until March, the normal financial year is from January to December. A financial forecast report can then be printed where the first three months of the forecast are used as the last three months of Head Office’s financial year.

DRP provides facility to define and adjust expected annual demand either keyed manually or retrieved from the sales history files. The adjusted annual demand can be broken down over a period of months, allowing precise control over a timing peak period. This results in a closer, more accurate purchase order, thus eliminating overstocking at other times of the year.

A Forecast Definition is created for the various ways in which a Forecast can be calculated. The results of a Forecast Generation can be manipulated and then used in Purchase Planning to determine what orders need to be placed on suppliers to keep an adequate stock level. 

Forecasting can be done nationally or by branch. If at a Branch level, a further definition is required to indicate which sales figures are to be used. This is indicated in the Sales Analysis Level field. Forecasting can be done in Bond stock or free stock or all stock. If a forecast on bond stock is selected, a separate price code, currency and type of sale can be nominated. Forecasting can be done by Publication/Release date

The definition determines the number of months of historical data to be selected (be it sales, budget or existing forecast), and then how many of those selected months to use, as well as the number of months for which to generate a forecast.

The forecasting calculation methods are:

• A direct transfer of monthly sales figures from a previous financial to the corresponding months in the current year
• Total average sales of a particular financial year can be broken down over a period of months

To these figures a Growth Rate or Seasonal Trend Profile can be applied. Generated forecasts are fully maintainable and can be maintained by item hierarchy. Changes can be cancelled prior to an update taking place.

Focus forecasting

The concept of forecasting is to setup multiple forecasting parameters. These parameters are retrospectively applied to previous periods. The resulting definition that is closet to actual is then used to generate the forecast.

Forecast definition maintenance allows multiple forecast definitions for a single forecast.

The definition also includes forecast generation rule. The generation rule provides the flexibility to selectively include and/or exclude certain conditions into which selected data is to be used for forecasting. This allows exclusion of special promotion sales, giveaways and extra ordering peak sales otherwise provide inaccurate sales data in forecasting future sales. Conversely, forecasting of non streamline sales is possible by selecting only the non normal sales in generating the forecast.

Sales history budget summary

The budget can be generated in various ways as follows:

• By direct entry of budget figures

Budget is setup using budget maintenance. Enter the quantity, amount and gross profit budget by total or month or just by total. If the total is only entered, the system allocates the monthly figures based on a profile code. The level must be by total customer by specific item.

• By transferring existing PSI to forecast

Transfer the PSI forecast database to the forecast file using program Transfer PSI to Forecast. The forecast ID needs to be setup. Copy the generated budget data from forecast file to the historical sales database file as budget using program Budget/Forecast Transfer to History.

• By transferring historical sales data to forecast data and finally copying forecast to a budget

Budget is setup using forecast definition maintenance. Using the forecast ID created, generate the forecast and the budget data into forecast file using forecast generation program. Transfer the generated budget data from forecast file to the historical sales database file as budget using program Budget Transfer to History. All data will be cleared after transferring budgets to sales history.

Adjusting for external factors

The base forecast can be adjusted to take into account other factors that are independent of previous sales, such as a promotion of a particular item is expected to generate more sales – thus they will adjust the forecast accordingly i.e. they will order more than the system may predict due to external factors that can now be applied external to the system.

The most powerful feature of DRP is the PSI (Purchase Sales & Inventory) panel – which allows the user to test multiple what if scenarios.

They are able to take the system forecasted order quantity and test what the outcome will be if the demand were to be higher or lower than expected. What if we exclude returns from net demand – what if we exclude FOC sales from demand figures? What if we add on an addition 1000 that I know we are going to sell to a buying group in 2 months time.

What any distribution business wants to achieve is the optimum inventory levels – they do not want to have stock in the warehouse that they can not sell – this is the biggest hidden cost to distribution businesses. They are able to look at various scenarios and determine what the effect will be on the business of each before going ahead with the order.

Management meetings

Managers are able to construct various scenarios to present at management meeting in order to discuss the optimum quantity to order. The PSI panel is used as an on-line management information/analysis tool to facilitate the buying decision. Once a suggested order is finalised a purchase order can be created directly from DRP.

PSI is particularly useful for those managers who are responsible for multiple products as they are able to construct pre-defined scenarios to present at meeting – and step steadily through the group of items that they are responsible for. All information relevant to the buying decision can be presented on these tailored panels.

The main benefits of using DRP is that publishers will be able to more accurately forecast demand and reduce costs in holding excess inventory or minimise lost revenue from out of stock situations = Optimising inventory levels, bearing in mind lead times.

For new products

The forecast can be based on a similar items sold previously – e.g. a new Harry Potter book can use sales history from the previous Harry Potter book, from Pub Date – not the last 12 months sales of a previous item. Alternatively a family profile can be used – this type of item tends to sell in the following pattern, which can be applied to base forecast sales for this new item.

Background on publishing

The publishing industry traditionally has not embraced the concept of forecasting and inventory control. This might be due to the fact that the industry has a ‘sell or return policy’ – where if booksellers buy items from the publisher and they don’t sell them they are able to return them to the publisher and regain their cots (with certain restrictions) – this leaves the publisher with a lot of books they have to dispose of – lots of wasted money. Another factor is the low print cost of books meaning that excess inventory does not come at such a high cost for publishers, as it may do for other types of distribution business. As the market becomes more competitive and larger buying groups and retailers put more pressure on margin, these types of efficiencies are becoming more relevant.

Returns can reach as high as 40% for certain items and can amount to 15% of total sales. Undoubtedly this is a large cost to publishers and returns management is one of the major costs to publishers.

Iptor’s point of view is that returns rates can be reduced if publishers and book distributors use forecasting tools to better evaluate demand for certain items. Accurate forecasting and controls will optimise inventory levels both centrally and at the retail outlet meaning returns rates and inventory wastage will be reduced, forcing a huge cost and inefficiency out of the business.

PSI (Purchase Sales & Inventory)

Forecast Rules are used to drive forecasts based on sales history – this can be a global rule, to be over-ridden by exceptions – such as hierarchy based or even item level. Forecast Rules can be applied by various levels – some may use an average sales over the last 3 months. Some may use focus forecasting to choose the best available method for future forecasts. New items can have demand copied from previous items or from Item Families – like products.

Profiles can be applied – uplift and seasonal. The forecast then gets transferred to PSI – where it is analyzed against inventory levels. These can all just be processed through to suggested orders, can be manually changed, ‘marked’ OK, or have an individual order placed for them as each are reviewed. The items will come through in a system driven sequence for review but commonly an item hierarchy or classification selection will be made = or specific items picked up.

User access determines what type of access individuals will get – can they view or maintain certain data.

In summary, the PSI function enables an online report and maintenance option to users as they review the system suggestions for purchasing.

Adoptions

Hierarchy of the academic database – down to lecturer level – Samples and adoptions are determined at this level. Main customer account can be determined at any of these levels. Alternatively, customer accounts (bookstores) can also be linked to the institution.

Using these links, sales analysis can be performed, so that Sales Versus adoptions can be generated.

Sales are tracked against the customer accounts but these accounts are tracked against the adoptions. The adoptions are always detailed at the lecturer level where competitor info can also be tracked.