INTRODUCTION
A system is perfect, not
when there is nothing more to add,
but when there is nothing more to delete
Saint Exupéry
WALTER’S is a programme whose designed primarily to assist professionals in
stock and supply chain management in making their decisions. WALTER’S is
equally at home in the retail distribution and the manufacturing environments.
This document aims to describe the different modules which make up WALTER’S,
and to assist future users to obtain as clear a view as possible of the help
that the programme will bring to their daily tasks.
The major objectives of WALTER’S are to ensure that the desired quality of
service is achieved, and to reduce significantly the rate of stock-out. To
achieve these aims, the only way is to anticipate the future demand.
WALTER’S undertakes this, by using different modeling techniques of potential
demand. Knowing, as we all do, that purely mathematical and statistical forecasts cannot take into account random demand changes or commercial promotional programmes, WALTER’S includes a number of mechanisms for correcting the short term, by following the evolution of sales and the order book, and then correcting the forecasts.
By allowing the users to foresee supply incidents, and to take decisions as late as possible, WALTER’S guarantees first-rate management of the first fill rate, a significant improvement in supplier and customer relations, and a notable change for the better in stock levels and stock turn.
WALTER’S is very simple to install: to convince yourself of this, just send us a diskette containing at least the Items and Historical Data described at the end of this document. We will diagnose, free of charge, the state of your current logistics, and will show you how WALTER’S will help you improve performance in your business, with your products, and your constraints.
WALTER’S: FORECAST TO WORK
Objectives
WALTER’S is designed to ensure Service Quality at every level of the distribution and production chain.
The right product at the right time in the right place: not a new concept, but nonetheless the leitmotiv that has been offered for years as the main sales argument of all methodologies and computer programmes aimed at stock and supply chain management.
The right product at the right time in the right place: today, it is the basis of Just in Time (JIT). WALTER’S interprets this, in simple pragmatic terms, as Just Enough, When Needed.
To meet this objective,WALTER’S must therefore:
Ensure product availability
Minimise Work in Progress and stocks
WALTER’S: a State of Mind
One of the key factors of success through the Quality of Service, is the creation of relationships between suppliers and customers which are based on mutual trust and confidence.
The first and unavoidable step in achieving this is the development of credibility between the partners. WALTER’S helps you to obtain this, by applying some simple, basic rules:
Clarity to the customer on the constraints the company operates under.
Acceptance that internal constraints exist.
Deal with suppliers as we would wish customers would deal with us.
WALTER’S: A tool for demand-driven flow
WALTER’S is designed to anticipate needs and to react to demand. It achieves this by using data collected as far down the supply and production chain as possible, and to structure these data in order to work their impact back up through the supply chain.
To do this, WALTER’S has a number of forecasting tools, sales modelling tools, and self-adaptation to the seasonal or periodic variances in demand and the Client- Supplier collaborative effort
Quality of Service.
While continually striving to improve the control of stocks, a company must also seek to improve the level of Customer Service.
The objective is clear: Customer Service at 100%
It is a fact that, for a company with a significant number of transactions, a service rate below 96 to 97% results in low customer credibility.
Long-term customer satisfaction is obtained when the service rate is 99% or better.
No small task, just to remain competitive!
The stock management system must therefore allow the company to take into account the dynamic nature of business and of order flows, and assist in forecasting such changes.
KEY FACTORS FOR QUALITY OF SERVICE
MATERIAL MEANS ENVIRONEMENT MEASURES
Availability Forecast Credibility Customer lead time
Storing Needs plannings Confidence Supplier lead time
Obsolescence Follow demand changes Forrester effect Stock-out risk
Aged Follow deliveries Real Forecast
Unusable Inventory
-------------------------------------------------------------------------------------------------------------------
==> FIRST FILL RATE
Many factors affect the Service Rate.
To achieve and go beyond 99%, all the factors that have a significant impact on it must be taken into account.
In particular, it is vital to monitor and control all the “vicious/virtuous circles” that can develop, with catastrophic or beneficial results. One example is customer credibility. A drop in the First Fill Rate causes a drop in credibility with customers, which creates a drop in accuracy of forecasts, which creates a drop in credibility, which creates a reduction in the Service Rate, which….
The mechanisms of “Flexible Planning” , in tandem with those of “Supplier/customer co-operation” allow us to take into account the overall dimensions of the problem to be solved.
Luckily, it is these same factors that influence the control and the reduction of inventory.
Useful inventory
Useful inventory is that which can be sold, delivered or used between two supplier deliveries.
Their volume is a function of the degree of uncertainty surrounding demand, lead time, supplier capacity, transport economics, cost of storage, cost per batch, and finally, the sales frequency.
“Tight” management should lead to actions to reduce inventory to the strict minimum required to meet quantifiable constraints (batches, delivery cycles….), increased by an amount corresponding to the uncertainties (customer needs, lead times,….), and which are calculated in accordance with the Quality of Service objectives set for customer satisfaction.
“Flexible planning” entails the continuous monitoring and modification of the forward purchasing requirements, in line with the evolution of customer needs, in order to control inventory and to keep it as low as realistically practical.
The continuous modification of the management parameters of each item in relation to the progress achieved is the guarantee of long term and durable results.
In the context of Flow Management, the inventory is a consequence: it is the difference between that which has been purchased and that which has been delivered or sold.
In relation to the defined Quality of Service level, the firm forward plan and supplier flexibility, WALTER’S automatically creates a dynamic “safety stock”, which is changed every time the Master Plan is rerun.
WALTER’S and Suppliers
WALTER’S is totally oriented towards customer service. The Quality of Service in a JIT environment cannot be achieved by purely localised actions: it depends also on suppliers and the quality of the relationships established with them.
WALTER’S takes into account the supplier constraints, which are principally: Minimum re-order quantities, manufacturing batch size, downtimes, and manufacturing lead time.
Setting realistic parameters for these constraints will WALTER’S users to create orders (supplier orders and/or manufacturing requirements) that are generally acceptable to the supplier. This will reduce blow-by-blow actions and progress chasing.
From being a source of emergency orders, WALTER’S users will become a source of realistic and credible forecasts. This modifies the balance of power with the supplier, and leads to the development of a relationship based on positive and objective criteria, rather than a trial of strength.
WALTER’S: Forecast to Work
At the lowest level of the data network managed by WALTER’S, the demand requirement is generally unknown, because of the local market size and potential, and the difficulty in forecasting customer behaviour.
Nonetheless, the demand requirement must be filled in line with what the same unpredictable customer requires. In an industrial environment, the available lead time is continually being reduced, and in the retail environment, is already at zero: product not available? Lost sale.
More and more, plants manufacturing products are aligning themselves on the criterion of demand satisfaction. Nonetheless, a certain lead time is inevitable between the decision to produce, and the availability of a product at the point of sale.
These same plants are also subject to their own component and materials supplier lead times and delivery times.
All the actors in the production and distribution chain are therefore required to anticipate demand. The degree of anticipation varies: one day between Point-of-Sale and the distribution platform, 3 months between a plant and its suppliers, 6 months or more for high value capital goods, some raw materials, imports…
To anticipate demand, we forecast. WALTER’S has a multi-model forecasting module which will automatically select the model or algorithm which is best adapted to the data supplied.
By definition, a forecast is inaccurate. The degree of inaccuracy increases in relation to the randomness of product behaviour. WALTER’S manages this inaccuracy by using the standard deviation between Reality and Forecast, which forms the very basis of the notion of “Risk Management”.
The risk managed by WALTER’S increases in direct proportion to the level of Quality of Service required. Furthermore, it increases with the level of forward forecast required: the further out one goes, the higher the risk. Finally, the risk is reduced by the ability of the suppliers to take on a portion of the risk themselves. This notion of supplier participation is built in to WALTER’S by the definition of a “Supplier Profile”, which quantifies the ability of a supplier to accept a modification of his order book dependent on forward visibility provided.
WALTER’S creates its medium and long-term forecasts on a monthly basis, and includes a means of analysis of daily movements which allow short-term correction of the forecasts, and the detection of “Exceptional Orders”.
Risk management and exceptional order management are the two fundamental tools which provide WALTER’S with the means to achieve its quality level in inventory control and product availability.
These tools are used throughout the supply chain from manufacture to product delivery. The Requirement Plans are modified at the start by whether or not the demand is independent or dependent.
Forecast to Work
WALTER’S:
Analyses product movements in terms of:
Value
Margin
Volume sold
Sales frequency
Customer numbers
Responds to queries:
What products should be stocked?
Where?
How many?
When should they be ordered?
Continuously revises the management parameters:
Safety stocks
Forecasting models
Gives tangible results:
Inventory reduction
Improved Customer Service rate
Reduction in administration costs
Fits in to all forms of organisation:
Single user
Multiple work-stations
Networks
Is in the Windows © environment:
Reliable
Transportable
Long life
Network compatible
Compatible with Db’s (Oracle, Ingres, dBase, Infromix, AS400…)
Easy communications
WALTER’S:
Forecasts 18 months rolling:
Trends
Seasonality
Life cycles
Random factors
Manages risks:
Over-stocking
Stock-outs
Seasons
Obsolescence
Factors in supplier constraints:
Minimum re-order quantities
Delivery batch sizes
Supplier flexibility
Shutdown periods
And global constraints:
Complete containers
Freight paid….
Plans:
Day by day
Week by week
Month by month
With finite and infinite capacity
Helps the Points of Sale:
Activity analysis
Forecasts
Forward planning
Performance indicators
Supplier negotiations
Informs centralised purchasing functions
Real movement by item
Planning of POS needs
Regulates central inventory flows.
Demand “pull”, not supply “push”
Avoids peaks and troughs
Realistic forward planning
PRESENTATION OF THE DIFFERENT MODULES
Calculation of forecasts
Customers to deliver upon receipt of order, but plants and suppliers requiring several days, weeks or months lead-time: a normal state of affairs in the manufacturing and distribution management environment.
To cover the difference between customer needs and supplier constraints, everybody in the manufacturing and distribution chain has to anticipate requirements. Whether it be explicit, through the use of methodologies, or implicit by the creation of safety stocks or emergency stocks, operations need forecasts.
Many methods of forecasting exist, from the simplest (e.g. rolling 3,6,12 months) to the most complex (e.g. algorithmic models like ARMA).Between these two extremes, many mathematical and statistical models exist to provide indicators of the future activity of the business.
The choice of a forecasting model is a delicate one: each type of model is more or less applicable, dependent on the type of data to be handled. Whether the choice of forecasting method be one of rolling month(s), smoothing, weighted, regression analysis, correlative analysis, ARMA or ARIMA, the “accuracy” and the viability of the numbers that come out are totally dependent on the correct choice of model. Mathematical or statistical tests allow us to qualify these models.
In virtually all types of industrial and retailing companies, we find products which have totally different behaviour patterns, in terms of frequency and volume of movements, or trends, or seasonality, or life cycle.
Detailed analysis of product behaviour, and a lot of experience in using the different types of model allow us to make a sensible selection. In practice, however, the time required to analyse each product or product family and select the most appropriate model far exceeds the time normally available.
In making its forecasts, WALTER’S uses a strategy based on the optimisation of the selection of forecasting model. By analysing the data both in terms of frequency and amplitude, WALTER’S makes an initial choice of the types of model that are useable.
WALTER’S then tests the relevance of the models initially selected by analysing the variances between the model and the real data to be processed. It then adapts the models’ parameters, in order to select the model that, on historical data, would have given the least variance between Forecast and Real.
This process of qualifying the model to be used guarantees not only the correlation between the model used and the data to be handled, but also allows the association of a standard variance Forecast/Real to the forecasts period by period, which is of major importance in risk management.
WALTER’S handles entirely for the user the generation of forecasts on the available historical data. This data, which the user may have reworked to generate “virtual historical data” in order to correct data known to be inaccurate (e.g. stock-out leading to loss of turnover), or to fill out incomplete data, or to generate historical data for products in the ramp-up or launch phase of their life-cycle.
To achieve this, WALTER’S has a number of built-in functions that assist in the formatting of data, by analogy with other products or product families.
The forecast calculation is done in real time by WALTER’S.
WALTER’S builds its forecast on a monthly basis, then breaks this down on a weekly and daily basis – always with an 18 month forward view.
Each time the Master Plan is run, the forecasts are adjusted on a daily basis, and corrected to follow the short-term trend-line established: it follows therefore that WALTER’S is particularly reactive to changes in the product environment.
The algorithms used by WALTER’S in its calculations are especially effective: an average of 500 items per minute are analysed and processed.
For products that can be substituted by others, WALTER’S offers the user the choice of generic items selection.
The generic item is a fictitious product that groups together similar items: e.g. the generic item “80cm TV” will group all brands of 80cm screen TV’s.
The sales are collected item by item and totalled to give a number for the generic item. The forecasts are then made at the level of this generic item, then broken down by real item, in accordance with a code or rule defined by the user company.
Use of the FORECASTING and SALES BEHAVIOUR modules of WALTER’S covers almost every imaginable practical instance where anticipation of the need is required.
WALTER’S keeps a permanent record of real sales, which allows the user to switch without risk or trouble from a forecast mode to a behavioral mode, and vice versa.
In every case, because WALTER’S manages both the forecast need, and the standard variance between Real and Forecast, the programme guarantees risk management.
At the most detailed level, WALTER’S Calculates forecasts for:
Customer requirements
Point of Sale
Item
Supplier
on a daily, weekly, or monthly basis in its standard format, 3 years out. Each item managed may have its own method of management. Moving from one method to another can be done at any time.
The calculation of the forecasts always takes into account the specific activity characteristics of each individual product, and automatically adapts itself to take into account any variance in product movement, through the use of a short-term correction formula.
The main models managed by WALTER’S take into account:
Trends: long, medium and short term
Seasonal or cyclical phenomena
Erratic item behaviour (sequential phenomena)
Product sales behaviour (new products)
When managing a network, the forecasts are calculated form the data supplied by the point of the supply chain closest to the customer: the Point of Sale (POS) They may also be calculated using shipment data (supplied by the distribution platforms). A corrective mechanism is used during the calculation of the Plans, in order to ensure that possible distortions provoked by the normal, real-time functioning of the network are taken into account: e.g. batch ordering by the POS or Platforms.
The forecasts can be cumulated, or split down to whatever level the user defines.
Managing the constraints
While the principal objective of WALTER’S is to ensure product availability throughout the manufacturing and distribution supply chain, the same base data of needs could also used to measure and evaluate the economic impact of the supply constraints that suppliers impose upon their customers.
WALTER’S has therefore developed the SUPPLIER NEGOCIATION module, which completes the panoply of tools that makes the programme the complete method of purchase and supply management for industrial, wholesale and retail organisations.
By virtue of its methods of calculating forecasts, modelling sales behaviour, and taking into account the impact of sales promotional programmes, WALTER’S allows the user to both build, review and revise the product requirements of a company. Not only that, but the programme builds a very realistic picture, by taking into account the inherent uncertainties of forecasts, and order and delivery lead time.
WALTER’S supplies the user with a total view of his needs, item by item, over a one year time frame. While this is perfectly adequate for most negotiations, it is possible to go further, and use the definition of the needs to measure the economic impact of the supplier constraints.
The major constraints are usually batch size, and the lead time between order and availability.
The results are therefore felt in terms of excess stocks, which, as we all know, cost approximately 20% to 30% annually of their initial purchase price.
As WALTER’S knows the correct cost of ownership of the stock, the economic cost of the excess created by the need for safety stocks tied to lead times can be calculated, separate from the purchase conditions such as minimum order quantity, packaging , etc.
This data can be assembled at the lowest level (i.e., one product code) or at a consolidated level (e.g. for a range of product) tied to delivery conditions, such as a full container, or complete truckload. The user can thus rapidly evaluate whether the pricing offered by the supplier does actually compensate the logistical on-costs which result from batch buying.
Supplier performance is also measured by the calculation of service rate. We all know from experience that one point extra on the service rate is worth between 0.3% and 0.7% of turnover.
WALTER’S calculates the needs in order to ensure a service rate of close to 100%. Failure to meet delivery schedules by a supplier automatically impacts negatively on the service rate, and the cost of this is calculated by WALTER’S. The monthly indicators by supplier are an excellent source of information on past supplier service quality.
WALTER’S focuses entirely on meeting the real quantities required to run the business.
The financial impact of any supplier constraint that requires the company to increase its purchases relative to this requirement should be quantified and available for use during negotiations. Often, committing only to the definite or probable short-term need will cost the company less than taking up the offer of discounts for additional quantities whose sale is far from certain.
The calculations made by WALTER’S enables the user to develop a negotiating position that takes into account the supplier constraints: if, for example, the constraint is to order by complete truckload, then WALTER’S will calculate the optimum product mix to reduce the cost of excess stocks.
Once negotiations have been successfully completed, WALTER’S then stores the defined parameters, which will be used by the operational staff placing orders as and when the need requires.
Because WALTER’S uses reliable, medium-term (12-18 months) needs as the basis for its proposals, right from the start, there is coherence between the conditions negotiated and the execution of the contract on a daily basis, which is a major factor in ensuring a company’s credibility with its suppliers.
Master Plan
Definition
Simply put, “Flexible Planning” can be defined as ‘forecasting as early as possible in order to act as late as possible’.
It is based on the use of a series of Plans whose time frames and time periods are adapted to the events to be managed or controlled.
A plan with an 18 month time frame, updated monthly (time period), is a useful tool for General Management, and as a negotiation tool with
suppliers.
A weekly Plan with a 5 month time frame will help in decision-making for supply
and/or production.
A daily Plan with a 2-week time frame enables effective transport planning.
All of these Plans can be consolidated (by product family, by management
category, etc.) and financially evaluated. They then become the basis for the
Real/Forecast indicators.
The key quality required of these Plans is realism.
Secondly, the Plans must all be coherent, and must therefore be built from the same starting point.
WALTER’S uses throughout the same methodology for planning, whether it be based on data concerning points of sale, distribution
platforms, manufacturing plants, or product lines.
Calculation
Each item managed can have its own method of calculation. It is possible to switch between methods of calculation at any time.
The calculation takes into account all of the constraints tied to items and
suppliers, and notably:
lead times
batch sizes
required service rate.
The safety stocks are automatically adjusted to take these factors into account.
Built in to WALTER’S are complex data filtering and sizing principles (Intelligent Lot
Sizing: ILS)
Validation
The results obtained from calculating the Plans are examined by the User in order to validate
them.
This validation takes into account possible global constraints tied to batch size, in order to minimise the impact of excess stock generated by these
constraints;
This validation by the User then allows the proposals made by WALTER’S to be transferred to the main I.T. system.
The User can modify all or part of the proposals (Quantity, value, date).
Bill of material
Based on the Production Management bill of material,, WALTER’S also gives you the option of managing “Dependent
Demand”.
In this scenario, the Forecasts and Real/Forecast spreads are no longer calculated on sales history or component call-off for production, but by splitting back down the Forecasts and
Real/Forecast spreads of the finished product to the component level defined in the bill of
material.
Finished products are managed on the basis of “Independent Demand”, while
sub-assemblies and components are managed on the basis of “Dependent
Demand”.
Using the Forecasts and Real/Forecast spreads generated by following the product bill of material avoids the traditional – and costly – piling up of safety stocks at each intermediate level that occurs when standard M.R.P. methods are
used.
The bill of material can be entered directly into WALTER’S, and the quantities tied to each link are given a loss
factor, to take account of production losses.
The bill of material used in WALTER’S also takes account, in the calculation of the
“Dependent/Independent Demand”, of items that can both be used in the manufacture of a finished
product, and that can also be sold or shipped independently of the finished product
(e.g. after-sales parts supply).
Various functions (copy, modify, cancel….) are available to the User. Imploding the bill of material enables component use to be
visualised.
Cost and Price Management
Cost evaluation
WALTER’S uses the data supplied
by the Master Plans, to calculate the
associated costs, and defines them as:
due to lead times
due to batch size
due to safety stock
due to stock-outs.
This module will usually be adapted specifically to the User company, at no
on-cost, according to the information available and the ground rules for evaluation existing in the client
company.
Simulations
The screen “Stock and Purchase Strategy Evaluation” gives an initial approach to the evaluation of performance, dependent upon the level of constraints identified
(lead time, minimum reorder quantity, packaging, service rate…)
But this only gives an indicator at one level.
To help the User evaluate strategies on a more global level, WALTER’S can be used to simulate various
hypotheses, and to evaluate the costs tied to different network structures.
For a group of products, the need is to define each step in the products’ progress through the network. These steps may be real
(e.g. physical stocking) or virtual (e.g. transport, handling).
The constraints associated with each step must be described: Service Rate for the security stock level at each
step, lead time between steps, batch sizes….and values associated with each.
The simulation must also take place over a time frame of sufficient duration, and will evaluate the costs created at each step in the chain.
The structures and steps analysed, and the results obtained, are stored in a data base.
These advanced options are made available to the User after analysis of the data available and the specific needs of the User.
The relevance of the results is directly dependent upon the use of correct parameters in the definition of the Master Plans, and particularly of the historical data
employed.
Price Management
Prices are managed and take into account by:
thresholds (price list in columns)
valid from...to…
supplier
unit (Kg., one, 6-pack etc)
packaging (box of 1,144, 12x12 etc)
Managing the
Parameters, and setting up the Data
The main functions undertaken by this module are:
Parameters by batch:
Sets constraints of lead time, batch size, service rate etc. for a range of items selected by the user.
Historical data parameters:
Creation, modification, correction of historical data: the major options are:
set to zero
reset to initial values
smooth
filter
increase
reduce
copy to
copy from
force seasonality
Management of links between items
product substitution
kits
Sales promotion parameters
start date
end date
impact on forecast
volume impact
Calendars
sites managed (shutdowns, inventory….)
suppliers (shutdowns, non-shipment periods)
User forecasts
Activity Analysis
Not all items have the same sales pattern; some will be fast moving, with a high level of stock rotation,
while others will be of much lower and irregular volume, but must form part of the product offering and therefore be available on
demand.
Numerous criteria must be taken into account in the choice of items to be
managed, and the physical locationof the stocks.
The multi-criteria “Activity Analysis” is of great help in determining these
choices. They are simple to understand, and alloçw the User to categorise products into several
families, defined by their economic contribution (turnover and margin) and by market
activity.
Using these global analyses, the User can set the specific management parameters for each article.
WALTER’S continuously analyses product activity in terms of:
Turnover; Margin; Quantity; Frequency; and Number of Customers.
Multiple Master Plans
Master Plans can be calculated for a very large number of products, many
suppliers, and many sites.
The simple fact of calculating the Master Plans gives the user the volume for a group of products and for the main supplier.
The Plans for Offers allows the management of purchased or manufactured products with several suppliers and several sites.
The consolidated Plans allow the User to obtain a synchronised plan across a distribution network or production
process: Point of Sale, Distribution Platform; Central Stock, Production source……
Performance analysis
Performance analyses can be carried out at the User’s discretion, checking:
Delivery quality (lead times)
Financial status
Order reprogramming
Late orders
Management of sales promotions
Generally speaking, forecasts are based on past sales history of the products
concerned. Now and again, companies decide to run promotional campaigns: the WALTER’S module called SALES CAMPAIGNS has been developed to help in the management of this type of
operation, and to manage the sudden potential increases and subsequent decreases in volume that this
engenders.
Sales campaigns create a “hiccup” in the product life cycle. To begin with, their objective is to increase sales. But they can also “cannibalise” the sales of other
products, whose volume is therefore negatively affected. They can also merely pull forward sales, rather than generate an overall
increase. Finally, they can also have the effect of improving customer loyalty and therefore also impact on sales.
A campaign usually carries with it a degree of risk, which must be minimised. The auto-adaptive capabilities of WALTER’S is of significant assistance throughout the duration of the
campaign.
A sales campaign can be simply described by a start date, and end date, and an estimated volume
increase.
From this starting point, WALTER’S will ascribe to each concerned product an additional sales profile
(ramp-up, full flow, phase-out), by means very similar to those used for managing normal sales
behaviour. WALTER’S will manage this sales profile throughout the campaign
duration.
As and when sales or orders are registered, WALTER’S compares them to both the
forecast, and to the sales profile that has been attributed. The volume exceeding the “normal” forecast are filtered out, so that they can be eliminated from the sales history that will form the basis of future
forecasts: but, at the same time, the “excess” volume is saved in the history of the
campaign, and the additional sales profile is modified to take account of the real numbers
registered.
These means ensure: that the historical data remains “uncorrupted” and relevant for normal business
forecasts; that the real effects of the campaign are taken into account during the campaign and versus the initial campaign
forecast; that the real effects of the campaign on volumes, turnover and margin are measurable at the end of the
campaign; and that those results are available for the planning of future
campaigns.
These means of correcting profiles and measuring impacts are feasible only because WALTER’S works from a reliable base and standard deviation between
Real/Forecast figures. This allows the correction of product sales profiles and to identify the impact of the sales
operation.
This data entry screen is used to supply the data to the programme for sales
campaigns.
The objective is to estimate the impact on sales before, during and after a sales
campaign, for each product concerned.
The estimation of the increase or decrease in sales can be made in percentage impact
terms, or in absolute sales volumes.
The impact can be positive of negative (due to cannibalising), and the WALTER’S data base will allow the user to attribute an impact by analogy with previous
campaigns.
CONCLUSION
WALTER’S: for Demand Driven Management
WALTER’S : a demand driven planning tool
The concept is to acquire the data for planning as far down the demand chain as possible, in order to
work it back up the supply chain as far as possible.
Demand data is thus safe from distortions due to the management methods used by each of the “boxes” that make up the Production and Distribution chain.
The only distortion that is allowed is the detection and acceptance of “exceptional
orders” from the box that is the next one downstream.
WALTER’S is totally adapted to all types of company,
whether with centralised responsibility or not,
thanks to its capability to handle distributed operations, and to consolidate both data and
functions.
APPENDIX
Demonstration using your data
We need six files, preferably in ASCII format, though any format generated under Dos or Windows © can be
handled.
Description of the basic files required to run WALTER’S
For the basic files:
- the fields marked “*” are obligatory: the others may be completed, dependent on the data
available.
- The Alpha-character fields are aligned on the left
- The Numeric fields are aligned to the right
- Each entry is confirmed by CR/ENTER
- No delimiting punctuation between the fields
Item File (NAME= ARTI.TXT)
Product Code* : 18 alphanumeric characters
Description : 40 alphanumeric characters
Supplier (code) : 10 alphanumeric characters: supplier or manufacturing source.
Product Line :10 alphanumeric characters: gondola? Counter? or manufacturing line
Unit managed : 2 alphanumeric Characters: Kilos, M,….
Price : 11 numeric. The last 3 digits are after the decimal point.
Min Reorder : 8 numeric.
Batch : 8 numeric
Date* : 8 numeric: YYYYMMDD date of data extract
Sales history (NAME=SLHIS.TXT)
For setting up. Thereafter, the file of sales movements will enable historical data
update.
Product Code* : 18 alphanumeric characters
Number of fields* : 2 numeric Each field = 1 month: so e.g. 10 months = 10 fields
Sales per month* : 8 numeric: one field per month, as many months as you
wish.
Last month : 6 numeric YYYYMM
Or last week YYYYWW
If you send us 18 months of sales history, ending for example on the 30 November 1998, each record will consist of:
product code
number of fields (18 in this case)
18 fields, each one containing the sales registered month by month. The field furthest to the right will contain November 1998, with October to its
left, and so on.
/CODE/ No Mths/………./………/ …………. /………/………/………../ DATE
18 Sept Oct Nov 199811
/……………………………………………………………../
18 Fields
Months with zero sales: you may leave blank, or enter a Zero.
Movement files ( NAME=MSTM.TXT)
Product Code* : 18 alphanumeric characters
Customer movement date : 10 alphanumeric Characters
Quantity : 8 numeric
Sale price : 11 numeric The last 3 digits are after the decimal point.
Date data extracted : 8 numeric YYYYMMDD
Stock files (NAME=STOC.TXT)
Product Code* : 18 alphanumeric characters
Stock* : 8 numeric
Date data extracted* : 8 numeric YYYYMMDD
Supplier orders file ( NAME=SUPO.TXT)
Product Code* : 18 alphanumeric characters
Quantity on order* : 8 numeric
Delivery date : 6 numeric YYMMDD
Date data extracted* : 8 numeric YYYYMMDD
The fields STOC.TXT and SUPO.TXT must have the same date of data extracted
Customer orders file ( Name= CUSO.TXT)
Product Code : 18 alphanumeric characters
Quantity on order : 8 numeric
Delivery date : 6 numeric YYMMDD
Date data extracted : 8 numeric YYYYMMDD
File returned by WALTER’S for inclusion into your information system
Proposed supplier orders (NAME=PSO.TXT)
Product Code : 18 alphanumeric characters
Quantity to order : 8 numeric
Delivery date : 6 numeric YYMMDD
Supplier code : 10 alphanumeric characters
Run Date : 8 numeric YYYYMMDD
***