What is Oracle Demand Planning?

Overview

Oracle Demand Planning is a powerful and flexible OLAP tool that supports the forecasting of demand for sales of finished goods. Oracle Demand Planning is part of the Oracle Advanced Planning Solution, and seamlessly integrates with the other Oracle E-Business Suite modules.

System Features:

Pre-seeded Dimensions, Hierarchies and Data Streams Flexible dimension, hierarchy and data stream creation Legacy data integration Demand plan scoping by line of business Revenue and quantity forecasting Statistical forecasting Product lifecycle and promotion management Service parts forecasting Dependent Demand Planning Exception management Internal collaboration Integration with APS and E-Business Suite Flexible reporting pre-defined and ad hoc

Forecasting: What, Why and When?

What is Forecasting?

Forecasting is a term used to describe the prediction of future events. Demand Planning is the process of recognizing econmic demands for products in the marketplace. Forecasting enables the future to be visualised, it creates a baseline against which actual data can be measured. The key in Demand Planning is to monitor the gaps between the anticipated (forecast) and the achieved (actual) and then, using forecasting skills and tools reduce that gap in the next cycle. The Forecast is an estimate of future demand. A forecast can be determined by mathematical means using historical data; it can be created subjectively by using estimates from informal sources; or it can represent a combination of both techniques. [APICS Dictionary]

Why Forecast?

There are lots of reasons why Forecasting is a good idea: ...Meet Market Demand...Manage Lead-Times...Reduce Inventory...Avoid Over Production...Minimise Stock-Outs...Support Budget...Inject Realism...Increase Buying Power...Schedule Resources...Guide Company Strategy...Define Promotions...Maintain Price Structure...Measure Awareness of Demand...Drive Continuous Improvement...Create Stability in the Supply Chain... The type of Forecast solution that is needed will vary depending in the reasons selected but the key drivers will probably be: To reduce uncertainty: Predict future demand and as such reduce uncertainty, leading to reduced costs, increased responsiveness and improved customer service. To anticipate change: Plan for future change (adjust for changes in the environment: competitors, new entrants etc.) and drive / support Company Strategy To improve communication: Between all departments and affiliates. Drive continuous improvement and migrate to intelligent models. Provide elements of Realism to reduce inventories driven by fanciful sales targets To increase knowledge: Project production capacity, project future costs and revenues, compare forecast and actuals to budget, evaluate marketing initiatives and decipher correlations between casuals and actuals To grow revenues and increase profitability!

When should you Forecast?

High level process just before budget cycle. Aimed at setting base for Budget discussion and high level capacity check (long horizon)

0-12 Months

Process to set sales targets. Set once or twice a year.

6-12 Months

Process which is aimed at providing accurate predictions of market potential. Data used to drive supply planning, allocation and production operations. Process (weekly/daily) used to obtain latest sales data from Sales (orders + expected orders). Applies to current month only

0-6 Months

1 Month

Forecast Types

Sales Forecast:

What Customers will Order ($) Sales forecasting is a self-assessment tool for a company. You have to keep taking the pulse of your company to know how healthy it is. Sales forecast reports & graphs enable you to analyze the health of your business. It can make the difference between just surviving and being highly successful in business. It is a vital cornerstone of a company's budget. The future direction of the company may rest on the accuracy of your sales forecasting.
Demand Forecast:

What Needs to be Procured / Made to meet the Sales Forecast (Qty)Demand Forecasting is the activity of estimating the quantity of a product or service that consumers will purchase and thus can be based on the Sales Forecast. However the Demand Forecast should strip out Inventory, apply supply chain constraints and accuracy factors.
Revenue Forecast:

Revenue Forecast: What Money will be made by meeting the Sales Forecast ($)The Revenue Forecast is an assessment of the profit that a company might make (gross - cost = net) providing a financial baseline to measure achievement of business strategy.

Dimension, Levels & Timeframes

When defining a Forecasting solution it will be necessary to decide which Dimensions to use, the levels of Data that will be organised and the Timeframe to manage.

Dimensons

Dimensions contain the core elements of your Data. Relational databases are by definition two-dimensional (rows and columns) but OLAP databases like Oracle Demand Planning and Demantra enable the introduction of additional dimensions which can be likened to perspectives or views of information. Basic Dimensions are: Time, Product & Customer. Useful: Organisation, Sales Channel & Manufacturer Others: Sales Account, Complaint, Service, Customer Satisfaction, Marketing Campaign, Sales Compensation. These Dimensions can be combined or separated as your Business requrements needs and your data allows.

Levels

Levels are the groupings that are defined in each Dimension. Balance the reward of extra detail with the cost of Master data setup and maintenance!

Timeframes

Timeframe decisions include: the amount of past data used to generate the forecast, the horizon time that is forecasted, the cycle time between each new Forecast and the period that is used to create accuracy analysis.

The Bullwhip Effect

The 'Bullwhip Effect' is one of the major issues for forecast-driven supply chains. The further away from the end customer, the more significant the levels of safety stock and exposed risk become. Some of the key 'Bullwhip' factors are:

Demand variability leads to significant supply chain inefficiencies:

y y y y

Data Sources (who is forecasting?) Forecast Method (quantitive, qualitive) Level of aggregation, allocaton and forecast generation Time Dimension (the further in the future the less accurate)

Demand planning difficulties:

y y y y

Evolving market conditions Global Markets Multiple Sales Channels Multiple Sources of Information

Demand distortions:

y y y y y y y y y y

State of the economy Promotions Competitor responses Lot-size discounts Fiscal Budget cycle Alloctations Panic & preceptions Hedging The weather Stock-outs

The 'bullwhip effect' effect can be reduced through:

Collaborative planning KPI: You get what you measure Improved data sources Improved forecasting methods

Demand Planning Inputs, Members & Values

Inputs are the "Data Streams" (ODP) or "Series" (Demantra) that can be collected and loaded into Forecasting Solutions. Inputs are pieces of information obtained from ERP, Legacy or Third Parties that can be used to create, compare and adjust Forecasts. There are virtually unlimited inputs to a Demand Planning Solution but some of the key examples can be identified as follows:
Historical Data Streams

y y y y y y y y

Bookings Shipments Invoices Returns Inventory Previous submitted Forecasts Distributor Actuals Syndicated Data

Forecast Data Streams

y y y y y y y y y

Statistical Forecast Annual Plan Revenue Forecast Sales Forecast Sales Pipeline Manufacturing Forecast Customer Forecast Backlog / Future Orders Incoming Supply

Causal Data Streams

y y y y y y y y y y

Stock-Outs Seasonality State of Economy Marketing Events / Promotional Activity Weather Labour Problems Price Changes Distribution Method Changes Consumer Earnings Production Capacity

y y y

Style or Fashion Direct and Indirect Competition Political Events

Members
Members are the Items, Customers, Warehouses, Sales Reps etc. that you view in your inputs. What Members do your Planners need to see? All Customers and Items or just a selection? The Standard collections process might bring into your OLAP cube things all "items" (Toilet Rolls, Coffee Vending Support, Scissors etc.) rather than the things you need to Forecast. Members will need to be grouped (see Levels) for Collection, Reporting and Forecasting.

Values
Values usually means the number associated with the Members that have been collected in the Inputs. Typical Demand Planning Values will be Qty and Price - but what price? Average Selling Price, Future Selling Price, Cost, Margin.... The more values required the more streams of data for the Planners to review not to forget the setup and maintenance.

More Inputs & Outputs

Having defined the Inputs, Members, Levels and Values of your Plans you will then find that more data is required: KPI's and Exceptions will need to be calculated. Promotional and Historical adjustment Activity will need to be measured for effectiveness. Forecast Variance to prior Months and alternative Forecasts will be necessary. Don't forget the final Outputs at the end of each cycle the Forecasts will be sent to Finance and/or Manufacturing and/or Purchasing. These extra streams of data will need to be defined and stored and might be wanted in more than one Value (Qty, Price, %, MAPE etc.)

Statistical Forecast Options

Forecast Type

There are many different approaches to creating a Statistical Forecast but the most common forms are: y y y Moving Average - Smooths data to create a less volatile line. Smoothed past data can be used to create smoother forecasts. Exponential Smoothing - Similar to Moving Average but uses the forecast error to drive forecast direction. Regression - Uses at least one variable to predict another.

Intelligent Systems

The advantage of the various Solutions that are available is that they can automate the analysis of hundreds of forecast types and their variations and then select the most appropriate. Two typical approaches are: Best Fit: Apply the most appropriate from a selection (ODP) Bayesian: Combine the best of many in one forecast (Demantra)

Forecast Level

The results of a Statistical Forecast will differ depending on the Level that it has been created at. Generally lower is better but issues like short lifecycles and unique customer products can force grouping levels. If you set levels too low the variations in data might be large causing decreased accuracy. Statistical Forecasts generated from higher grouping levels can produce better results than using the lowest levels but inaccuracies can be magnified at aggregate levels! If you set levels to high, the forecast might pick up localised trends and allocate them incorrectly.

The Dimension Levels that you choose should relate to your company data and processes and the needs of the business.
Forecast Time

The accuracy of the Statistical Forecast will depend on the length of History analysed and the horizon of Forecast that is created. 3 years of past data is enough to generate season trends but this requies product lifecycle stability or the creation of suitable levels to group like-items.

What is Oracle Demantra?

Overview

Demantra is one of the Oracle Value Chain Planning family of Applications. Fully integrated with Oracle R11 and R12, Demantra provides a Bayesian Forecasting capability in place of ODP's Best Fit approach. Other benefits range from Real-Time data to PDA functionality.

Demantra comes from: Demand is our Mantra

Yes, Yes, but what does it do?

The standard blurb can create more questions than answers can't it? Fundamentally, Oracle Demantra collects, stores and organises item, location & time information, generates statistical forecasts and enables comparions of actuals to forecast. Forecasts can be created and adjusted in Demantra, validated and then forwarded to other applications (such as Hyperion or ASCP) for action. You don't have to create Statistical Forecasts - you could just as easily use the previous cycle as a starting point.

And how does it do it?

Demantra is a web enabled tool that can be implemented in various ways: from a StandAlone single PC populated with data from Excel spreadsheets to a sophisticated solution integrated into your ERP System (Oracle E Business Suite, J D Edwards, Siebel, SAP etc.) with hundreds of globally distributed users. Planners interract with the Forecast (and related) data via worksheets that are quite simple to create and manipulate. Workflows and Methods can be created to automate functions and security features enable the creation of user specific interfaces.
Demantra Modules

Oracle Demantra is actually comprised of a number of modules that can be mixed and matched according to your business needs. The modules are: 1/ Demand Management (DM) 2/ Advanced Forecasting and Demand Modelling (AFDM) 3/ Real-Time Sales and Operation Planning (RTSOP) 4/ Predictive Trade Planning (PTP) - Deduction and Settlement Management - Trade Promotion Optimization

Demantra Pros and Cons

Demantra Positives

Fast collection & data loading process (much better than ODP) Has greater flexibility in Security / Setup / R&R Improved Forecast creation over ODP Improved Promotional, NPI and Event functionality Workflow Creation and Management Ad Hoc Hierarchy level value creation by end user Improved Graphical display functionality Customisation of Graphical display per user or user group Visible Comments and Audit trail

Promotions / Causals visible alongside Plan data Dependent Demand Functionality Other features: Off-Line working, Attribute Forecasting, PDA capability. Option for Real-Time rather than Batch mode transactions.

Some Demantra Negatives

Although data collection is fast filtered worksheets will be necessary for large volumes of data Relatively untried in Europe compared to the US Loss of ad-hoc Drill Down (requires embedded worksheets) End-user formatting restrictions in Worksheets: colour coding and resizing. More Modules = more analysis, setup and testing Series can really only be created by IT Colour Coding can really only be created IT Risk of RTS&OP module duplicating some ASCP functions Data only held at the lowest level means potential allocation and aggregation restrictions. Intersections need to exist before they can be seen in Demantra unlike ODP where all combinations are created by default.

Business Transition Considerations

Considering implementing Demantra? Here are some transitional thoughts:
1/ Series in Demantra cannot be User Maintained - Hierarchy Levels can be User Maintained

This is a swap of the current situation with ODP and therefore perhaps they balance each other out but.... There needs to be a far greater emphasis on creating a full set of correct Series prior to implementation. This means completing more thorough design reviews and more considerable training with End-Users on the significance / structure of Hierarchies and the layout of data in worksheets. Consider the processes required to request and create new Series post Go Live. The ability to create ad-hoc levels in hierarchies might seem liberating but will inevitably create data problems. Source data should really only be maintained at the ERP Source which mean that this functionality will really only be used for limited business needs and should be restricted to responsible super users. In reality, a successful Demantra implementation will need even more up-front analysis and preparation than ODP. The payoff will be a much more versatile, accurate and efficient Forecasting tool.
2/ Sales and Operational and Promotional become more significant link into or aspect of Planning

Need Quality data and process for the new module RTS&OP currently these aspects are underutilised in ODP. Gaining advantage from the new modules will require a new set of data to be collected and maintained.

3/ Simplicity versus Complexity

At what stage is your Planning Department? Many Companies are at entry-level stages of Demand Planning sophistication there is little to be gained and potentially much to be lost by replacing and overcomplicating any simple solutions.
4/ Bayesian Forecasting

The change from a Best-Fit solution to a Bayesian will mean a re-evaluation of the preferred Statistical method. A greater complexity in the Statistical Forecast will require a greater level of sophistication in end user understanding. Extra training and support will be necessary to bring the planners to the required level of expertise. Forecast level and reporting level can be different due to the dynamic forecast creation process in Demantra. This can mean a discrepancy in expectations and also provide frustration when a set forecast level is required.