Professional Documents
Culture Documents
Sensing
February, 2016
Objectives Slide
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 2
Agenda
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 3
Create a Forecast Model for
Demand Sensing
Create a New Forecast Model for Demand Sensing
Set up a new forecast model for the demand sensing algorithm via the
Manage Forecast Models app.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 5
Create a New Forecast Model for Demand Sensing
Time Settings
Mandatory: Periodicity needs to be set to Daily for the demand sensing algorithm
The periods are entered in days when the periodicity is set to Daily.
Historical Periods: As best practice from past customer projects, we recommend using 52 or more weeks
as historic data input. This showed very good demand sensing results.
Forecast Periods: This depends on the business case. Demand sensing is usually used in a forecasting
range of 4-8 weeks.
There is a minimum of historical periods that needs to be used in order to expect good results.
This depends on your overall settings. The UI will give recommendations upon saving the model
when the number you entered is too small.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 6
Create a New Forecast Model for Demand Sensing
Preprocessing Step Promotion Sales Lift Elimination (optional)
As of SAP Integrated Business Planning for demand 6.1, you will find a new
preprocessing step called Promotion Sales Lift Elimination . This step needs to
be used in combination with Demand Sensing (Full) in case you have
promotional sales lift information in your system.
Preprocessing steps other than Promotion Sales Lift Elimination do not work with
Demand Sensing algorithms.
The Promotion Sales Lift Elimination preprocessing step is not limited to Demand
Sensing in its usage. It can for example also run standalone.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 7
Create a New Forecast Model for Demand Sensing
Preprocessing Step Configuration of Promotion Sales Lift Elimination
Outlier Multiplier: A decimal number by which the system is to multiply the range of accepted values
when the variance test is used, thus including additional values in the range or excluding a set of values
from it.
Sales History: Key Figure that contains the Sales History including the sales lift from the promotions
(total). For Demand Sensing, this is usually the Confirmed Quantity or the Delivered Quantity.
Planned Sales Lifts: Key Figure that contains the planned promotion uplifts or downlifts in the history
and future. In case of SAP6, this is PROMOSPLIT (Promotion Uplift)
Consensus Forecast: Key Figure that contains the Consensus Forecast excluding the sales lift from
promotions! In case of SAP6, this is CONSENSUSDEMAND (Consensus Demand w/o Promotions)
Save Results In: Output Key Figure in which the results should be stored. This would provide the Sales
History exluding the promotions
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 8
Excurse: Promotion Sales Lift Elimination Preprocessing
Step
Searches for correlations to identify the periods when the outliers were
Step 3 actually caused by promotions
Eliminates the values from the sales history where a correlation was found
Step 4 between outlier and promotion periods
Stores the results for sales history, cleansed by the sales lifts in the output key
Step 6 figure
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 9
Create a New Forecast Model for Demand Sensing (Full)
Forecasting Steps (1)
Main Input for Forecasting Step: Key Figure that contains the data that acts as the main sales
history. For Demand Sensing, this is usually the Confirmed Quantity or the Delivered
Quantity.
Note that the Main Input for Forecasting Steps needs to be the sames as
Either one of the inputs for Historical Ordered Quantity or Delivered Quantity for the
Demand Sensing (Full) Model.
Input for Algorithm on the Preprocessing Step for Promotion Sales Lift Elimination
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 10
Excurse: Demand Sensing Full and Update Runs
A Full Run
Is normally scheduled at the beginning of a week to get the regression weights.
In most cases, a full run in demand sensing would follow an updated weekly mid-
term forecast and, subsequently, a snapshot that has been taken based on that.
These outputs are then used in the final sensed demand calculation.
An Update Run
Is scheduled whenever there are delta changes in the inputs
(new open orders/shipments) during the week.
Thus, it takes into account any new information for orders/shipments during the
week and then calculates the sensed demand based on the regression weights
obtained as part of the full run that was done at the beginning of the week.
An update run cannot be scheduled until and unless a full run is already done.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 11
Create a New Forecast Model for Demand Sensing (Full)
Forecasting Steps (2) Please refer to the naming
conventions for certain key figures
that are used as input for demand
Add the Demand Sensing (Full) algorithm to the model sensing.
This is described in detail in the
SAP6 Planning Model slide deck.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 12
Create a New Forecast Model for Demand Sensing (Full)
Forecasting Steps (3) Key Figures
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 13
Create a New Forecast Model for Demand Sensing (Full)
Forecasting Steps (4) Further Settings Bias Horizon: Bias in SAPs Demand Sensing can
be better defined as the consumption or ordering
patterns of the customers. It determines or predicts
what the future consumption looks like, based on
Consensus Demand Snapshot Suffix: calculated weights from the past buckets.
Needs to match the suffix that is used
when creating the forecast snapshot Bias Horizon 1|2|4|6|8|9 as in the screenshot is an
(see chapter 2) example and varies depending on your individual
data. Maximum is 10 as the highest bias horizon
entry. The last 2 entry fields can also be left blank in
case that is needed for the specific data set.
These four values are used for capping the final sensed demand
numbers. See next slides for more information.
These thresholds are checking the result of the Demand Sensing
Run (before putting uplifts back) against the Consensus Forecast
without promotions.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 14
Threshold Settings
The maximum forecast increase / decrease settings are used to cap the demand in case of exceptions in the data and smooth
the results. The user can define absolute and percentage values.
The threshold settings cap the outcome of DS before the order adjustment.
It is a global threshold for all planning objects that are planned with the forecast model, and thus has to be set very carefully.
In case you have items with very different volumes, one might think of creating different sensing profiles to avoid setting too
aggressive or too limiting thresholds.
Example:
Max Forecast Increase (absolute value): 50 PC (MAX_INCREASE)
Max Forecast Increase (percentage value): 30% (MAX_PCT_INCREASE)
The consensus demand is 100 PC (CONSENSUSDEMAND)
The sensed demand that would be calculated without taking the forecast increase factors into account is 300 PC
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 15
Example: Daily Average Calculation Horizon
The Daily Average Calculation Horizon is a rolling window (number of weeks in the past) over which the demand sensing
averages daily shipments to determine the daily shipment profile.
In this case, for 4 weeks into the past, the shipments for each day of the week are averaged. For last four Mondays,
Tuesdays etc. This in turn is used in calculating the shipment profile of each day.
Lower numbers than 4 weeks will influence the results in a negative way. We recommend using a value equal or higher than
4. The default value is therefore set to 4.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 16
Create a New Forecast Model for Demand Sensing (Full)
Forecasting Steps (5) Workdays
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 17
Create a New Forecast Model for Demand Sensing (Full)
Forecasting Steps (6) Minimum Data Points and WMAPE Threshold
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 18
Create a New Forecast Model for Demand Sensing (Update)
Forecasting Steps
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 19
Create a New Forecast Model for Demand Sensing
Postprocessing Steps
Error measures are not calculated during the demand sensing run
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 20
Manage Snapshot
Configuration
Introduction What is a Snapshot?
A snapshot is the current picture of any time key figure for any time range.
In case of demand sensing, it is for the mid-term forecast (e.g. consensus demand)
to learn the patterns of demand and correlate forecast with demand for different
lags
Historical snapshots are needed by the demand sensing algorithm in the calculation
of the sensed demand regression metrics. Note that it is mandatory to load historic
snapshot data. It is a mandatory input factor to the demand sensing
algorithms.
Historical forecast snapshots are loaded manually before the first demand sensing
run.
After the first run, the snapshot creation needs to be scheduled on a regular
basis via the SAP IBP add-in for Microsoft Excel (e.g. weekly or monthly,
depending how often your mid-term demand plan is generated)
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 22
Generate Snapshot Key Figure for demand sensing
A new snapshot key figure is automatically generated for your planning area
when a snapshot configuration is saved (see next slides)
is used to save the historical snapshot values during the initial system set up.
Note that historic snapshot values are needed by the demand sensing algorithm to
calculate high-quality results. So a certain snapshot history needs to be loaded
during the initial system setup (see chapter 3) and while productively using demand
sensing, the snapshot creation needs to be scheduled regularly.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 23
Generate Snapshot Key Figure for demand sensing
Time
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 24
Manage Snapshot Configuration (1)
Create a new snapshot key figure via the Manage Snapshot Configuration UI:
On the next screen, select your planning area via the dropdown list.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 25
Manage Snapshot Configuration (2a)
Mandatory:
Snapshop Type =
Change History
REV
Forecast Model:
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 26
Manage Snapshot Configuration (2b)
This is the range of periods that the snapshot should be taken for. Snapshot Frequency in weeks depends on whether you
want to take the snapshots weekly, monthly, etc. for weekly
These numbers represent days. snapshots use 1, bi-weekly use 2, for monthly snapshots use
From = 0 means that the snapshot will be taken always from the 4 as input. The frequency depends on how often your mid-
current day, the snapshot run is executed, onwards. E.g. if the term forecast (e.g. consensus demand) is being updated.
snapshot is scheduled every Monday, the data included in the At least one snapshot should be taken every four weeks!
snapshot starts from that respective Monday onwards. Example:
To = 63 means that the snapshot will be taken for the next 63 days Sensing Horizon = 6 weeks
(9 weeks) in the future. Snapshot Frequency = monthly
To = [6 + 4 - 1] * 7 = 63 days
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 27
Manage Snapshot Configuration (3)
Once you save the snapshot configuration, a new key figure will be generated
for the respective planning area.
Note that the planning area is set to Inactive now, so you need to reactivate it.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 28
Load Historical Snapshot
Data and Schedule Snapshot
Jobs
Historical Snapshot Upload
Prerequisites
Procedure:
Create a new snapshot key figure for the planning area
Check and adjust mandatory settings in the planning area
Pay attention to prerequisites during data upload of master data and key figures
Load Statistical Forecast qty/ Consensus Demand Plan qty for the past periods
that you want to use as historical periods in your forecast model.
Either load Statistical Forecast qty / Consensus Demand Plan qty for the future
(at least one data point is needed) or generate it via a statistical forecast run.
Load the historical forecast snapshots before the first demand sensing run
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 30
Mandatory Settings in Planning Area
Change History
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 31
Prerequisites for Master Data Upload
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 32
Upload the Historical Snapshot via the WebUI
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 33
Sample Snapshot File
Revision of
April 27, 2015
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 34
How Many Historical Snapshots Should Be Loaded?
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 35
Troubleshooting
Error: Cannot import data because period & planning object data doesnt exist for
the selected revision, Invalid Revision for Planning Object or Period.
Solution: This means the UoM (Unit of Measure) templates are not imported or
imported in the wrong order.
So, reload all master data types that have Attributes as key figures and then
reimport the snapshots (see the Prerequisites during Data Upload slide).
In the SAP6 template, there are two Attributes as key figures used for UoM and
currency conversion:
UOMCONVFACTOR
EXCHANGERATE
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 36
Set up Batch Job to Automatically Create New Snapshots
for Future Periods (1)
Schedule the snapshot operation to run automatically via the SAP IBP add-in
for Microsoft Excel
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 37
Set up Batch Job to Automatically Create New Snapshots
for the Future Periods (2)
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 38
Run Demand Sensing from
the SAP IBP add-in for
Microsoft Excel
Demand Sensing Run
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 40
Run Demand Sensing in Batch Mode(1a)
Log on to your respective planning area in the SAP IBP add-in for Microsoft
Excel and create a daily or a weekly view.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 41
Run Demand Sensing in Batch Mode(1a)
Optional step: Filter based on any attribute (e.g. Product ID, Product Group,
etc).
Note: The demand sensing algorithm will then only run on the filtered subset
of data.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 42
Run Demand Sensing in Batch Mode(1a)
Pick an appropriate Unit of Measure for the conversion of these key figures
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 43
Run Demand Sensing in Batch Mode(2a)
Start the full demand sensing job after picking the forecast model that
includes your demand sensing algorithm
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 44
Run Demand Sensing in Simulation Mode(1b)
Log on to your respective planning area in the SAP IBP add-in for Microsoft
Excel and create a daily view.
Note: The demand sensing algorithm will only run on the filtered subset of
data in case filters were applied.
Simulation of DS can only happen on a Daily View since the
DS algorithms are always set to a periodicity of Daily in the
forecast model (Slide 12)
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 45
Run Demand Sensing in Simulation Mode (2b)
Run demand sensing from the Simulate drop down after picking the forecast
model that includes your demand sensing algorithm.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 46
Demand Sensing Run (3)
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 47
Demand Sensing (Full) Outputs (1)
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 48
Run Demand Sensing (Update) 1
Situation: Future ordered quantities (Current Open Order QTY) were modified
in the middle of a certain week. In this example, new orders came in during
the middle of the week and the current open order quantity was increased
from 340 to 600 pieces.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 49
Run Demand Sensing (Update) 2
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 50
Lesson Summary
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 51
2015 SAP SE or an SAP affiliate company.
All rights reserved.
No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP SE or an
SAP affiliate company.
SAP and other SAP products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of SAP SE
(or an SAP affiliate company) in Germany and other countries. Please see http://global12.sap.com/corporate-en/legal/copyright/index.epx for additional
trademark information and notices.
Some software products marketed by SAP SE and its distributors contain proprietary software components of other software vendors.
These materials are provided by SAP SE or an SAP affiliate company for informational purposes only, without representation or warranty of any kind,
and SAP SE or its affiliated companies shall not be liable for errors or omissions with respect to the materials. The only warranties for SAP SE or
SAP affiliate company products and services are those that are set forth in the express warranty statements accompanying such products and
services, if any. Nothing herein should be construed as constituting an additional warranty.
In particular, SAP SE or its affiliated companies have no obligation to pursue any course of business outlined in this document or any related
presentation, or to develop or release any functionality mentioned therein. This document, or any related presentation, and SAP SEs or its affiliated
companies strategy and possible future developments, products, and/or platform directions and functionality are all subject to change and may be
changed by SAP SE or its affiliated companies at any time for any reason without notice. The information in this document is not a commitment,
promise, or legal obligation to deliver any material, code, or functionality. All forward-looking statements are subject to various risks and uncertainties
that could cause actual results to differ materially from expectations. Readers are cautioned not to place undue reliance on these forward-looking
statements, which speak only as of their dates, and they should not be relied upon in making purchasing decisions.
2015 SAP SE or an SAP affiliate company. All rights reserved. Early Knowledge Transfer Customer 52