*This DSiM FAQ was co-authored by Dan Hawker and Neil Bundy*
Forward thinking fashion companies, such as Top Shop, have enthusiastically embraced social media to generate buzz and business around their live catwalk events. These are highly visible, exciting events where the gliterati and fashionistas can share in the excitement of the moment, and tweet their thoughts on what’s hot and what’s not.
So what has this got to do with demand signal repositories, and more specifically, SAP Demand Signal Management (DSiM)?
Well, it’s all about giving CPG companies greater insight into consumer demand. In the world of fashion for example, consumers can buy online direct from a catwalk show. This presents an excellent opportunity to reduce inventories through the supply chain. In a globalised supply chain where lead-time, from picking the raw materials from the field, to a garment crossing the tills, can be measured in weeks or months, driving demand from the catwalk instead of in-store sales is a clear advantage.
For other industries such as food and beverages, doing something similar also offers benefits around:
Promotion and pricing advantage
Consumer engagement strategy
So what is SAP DSiM all about? , And what benefits does it offer?
1. What is SAP DSiM?
SAP DSiM is a SAP version of a generic concept called Demand Signal Repository. Demand Signal Repositories aim to gather different “signals” of demand into one place in order to harmonise them, and then reuse them wherever they are needed.
2. What is a demand signal?
A demand signal is insight into what consumer demand might be. For example, broadly in order of latency (more current signals first):
Weather forecasts. For certain products like beverages and some foods, upcoming weather patterns can signal changes in likely demand
Social media sentiment. If people are tweeting, blogging, posting about a brand on a social media platform, this can be analysed by any number of providers to gauge – in real time – brand perception. Some of it can also be attached to locations, giving a geographical lens on consumer demand. If they’re talking about it, they may be about to buy it!
POS data. If people are buying that brand, that’s a clear signal of demand
Panel data. Nielsen, IRI and other providers provide a different lens on consumer demand, that can be a more aggregated substitute for POS data
Sales in. Sales from a CPG company to retailers is a signal of demand
Distributor data. Sales from distributors are a signal of demand
Inventory. Movement of stock levels in the supply chain are a signal of demand.
3. What is a demand signal repository?
Some of these signals – the internal ones in particular (e.g. sales-in, inventory) will already be in a data warehouse within a CPG company. Others – POS, panel data – might be too, but with POS you’re starting to get into big data territory for larger organisations, so it may be in a different place from the internal signals. Panel data may be provided by agencies, along with some value-added analysis. Social sentiment, weather data, and so on, may be provided via a cloud service, or not at all. A repository is a large data warehouse where all of these things come together in one place, so will typically reside in a big data environment such as Hadoop, Teradata, or SAP HANA.
4. Why have a demand signal repository?
A number of processes in a CPG company rely on a good view of demand. Things like Sales & Operation Planning (S&OP), Integrated Business Planning (IBP), promotional planning, trade investment planning and so on. With a DSR, all of the different processes can source from the same enriched view of demand.
5. What is SAP’s DSiM offering?
SAP DSiM is pre-written content delivered via SAP’s data warehouse for big data – SAP BW on HANA. Find out more here.
6. Which organisations can benefit from SAP DSiM?
SAP DSiM is aimed at any organisation that sells via a retailer or third party such as consumer goods organisations, electronics equipment manufacturers, life sciences and pharmaceuticals, and automotive parts manufacturers.
7. What are the main business benefits of SAP DSiM?
Traditionally, consumer goods organisations and products manufacturers receive data on the demand for their products through response to purchase orders raised by retailers requesting more stock. The flow of this information from the retailer is often slow relative to the requirements of the demand and may take weeks for a manufacturer to have the information to respond to fluctuations in the demand.
There are also other demand signals, such as social media data and weather data that can offer more insight into product demand. Integrating these data into a single place gives a much clearer picture of true demand.
SAP DSiM allows data to be sourced directly from retailers, or in some cases, third parties, which may be available as early as the following day. This provides real insight into what’s happening at the store level.
Out of stocks
One of the most costly impacts to product manufacturers is out of stock situations occurring at retailers. While this obviously results in loss of sales, it can also result in additional sales to competitors. Through understanding stock levels not only of depots or distribution centres, but also retailers, helps manage and respond to these scenarios.
There are lots of statistics published on how trade promotions are often not profitable and fail to meet expectations. Sourcing and acting upon data from the store level helps give a much deeper insight into promotion effectiveness and profitability.
During promotions, retailers may “forward buy” and therefore stockpile products. This is done with the view to then selling the products at the standard cost after the promotion has finished. In the traditional flow of data, where all the manufacturer sees is a huge spike in demand, the manufacturer may increase output to meet the apparent demand. Having insight into store level stocks and sales enables the CG organisation to get a true picture of the demand.
8. CPG companies have to manage different formats of retailer POS data – are there any pre-configured rules that can be used?
SAP DSiM supports data from external sources in a consistent flat file format. Alternatively you can source data directly from BusinessObjects Data Services. Data Services can help put the data into a consistent format for loading into the DSiM data model.
If you don’t want to manage this yourself, there is a certified integration for certain retailers from Retail Velocity. Other providers will no doubt surface over time
9. Are there any early adopters of SAP DSiM?
Companies like Coca Cola and Colgate Palmolive are taking the plunge with SAP DSiM.
10. How mature does a company need to be to embark on a SAP DSiM initiative?
It depends. We have a framework to gauge how ready companies are, and what steps are needed to prepare successfully for one. Broadly speaking, if your company already has an established BI Competency Centre (BICC), and already experienced in processing POS data and sending it on into multiple commercial processes, then those are good start points. If there is a mature understanding of what social sentiment can offer, perhaps even integrating already into a SAP CRM system of record, then even better.
11. I have SAP Advanced Planner Optimizer (APO), isn’t that enough?
Without external data, SAP APO can only forecast demand based on stock requested from a retailer. By integrating data from SAP DSiM into APO, you can integrate demand data within as little as 24 hours of what has happened in store. Additionally SAP APO doesn’t provide insight into the profitability of promotions.
12. I sell my products to a number of retailers in each market that I work do business in. I also get market research data from a number of different suppliers by geography. Can SAP DSiM handle these complexities?
DSiM includes a whole data upload workflow called the Process Flow Control which is extensively configurable. This allows files with different naming conventions, different granularities and different types of files to be loaded. You define data loading steps for each file set or data delivery agreement you have with a retailer or market research company.
13. Does SAP DSiM have its own user interface?
Yes. DSiM uses the SAP NetWeaver Business Client to deliver the reports through its own reporting and analytics UI. In addition to reporting, data harmonisation, data quality validation controls and data mapping rules are available through a web based interface.
The application is set up to be ‘role based’ and provide reports to users, such as demand planner or marketing manager. These roles can of course be changed and customised to meet business requirements.
Alternatively these reports can be consumed in a web browser.
14. What reports are available out of the box?
With the current version, there are a number of analytical views available in the web analytics UI.
On-shelf availability by product
On-shelf availability by location
Out of stocks by product
Out of stocks by location
Sales analysis: including sales comparison for different products over time ranges and includes quantities or amounts
Promotion analysis: sell in and sell out data for promotion, sell out analysis, promotion type, sell out analysis by product, sell out analysis by location.
15. I have SAP BusinessObjects Enterprise 4.0 (Dashboards, Webi, Analysis for Office etc.). Can I use these reporting tools with SAP DSiM?
Yes, and I would encourage you to make use of other reporting solutions. The out of the box analytics are a starting point and there is scope for improvement.
SAP DSiM comes with some queries that are specifically designed to interface with BusinessObjects tools such as Analysis for Office.
16. What about mobile integration?
You can create your own mobile applications through the standard SAP reporting toolset. This could include tools such as Design Studio mobile dashboards, or BusinessObjects Mobile Dashboards.
17. Can I use the data that I have in DSiM to create my own customised predictive models?
Yes. With some customisation, you can expose the harmonised and cleansed demand data in DSiM to SAP’s other predictive analytics applications:
R (in HANA)
Further analysis in SAP HANA with the SAP Predictive Analytics Libraries (PAL) (in HANA)
18. What datasets do SAP DSiM currently support out of the box?
Data from your internal SAP ERP system (SAP ECC data is of course the easiest to integrate), sales order data, delivery data
SAP Trade Promotion Management (TPM) data. Native support and extraction for SAP TPM data is included, but with a bit more work and the right interfaces you can integrate TPM data from any TPM solution
Retailer Point of Sale (POS) data
Retailer stock snapshot data from stores and depots
Market research data. This could be data from organisations such as IRI or Nielsen or others. The standard interfaces for these data expect data in the retail panel format frequently used by market research companies to share these data
Social media data to show buzz and sentiment. This has to be provided through an approved API, such as through SAP Social Media by NetBase. You cannot directly load Twitter data into DSiM for example.
19. How can I consume the demand data in other SAP tools?
You can load the output from cleansed demand data into:
SAP Trade Promotion Optimiser, which in turn feeds back into Trade Promotion Management and then back into DSiM
Planned future integration with SCM Enterprise Demand Sensing
20. What skills are required for a successful implementation?
Ideally you would have prior experience of SAP DSiM and demand signal repositories. With DSiM being a new tool then this may not be possible. Here are some essential skills
Strong SAP Basis and HANA skills to configure and install the data model and user interface
Up-to-date BW skills including experience of SAP BW on HANA and the new data modelling approaches associated with this such as LSA++
In depth functional understanding of demand planning and retailer and CPG organisations and how internal and external data can be mapped and harmonised
Your implementation team will have to be very skilled. SAP DSiM has a very complicated data model and a lot of non-standard functionality. To make the implementation a success, your implementation team will have to be prepared to embrace these technical and functional differences, rather than trying to code/hack around them
You should also be aware that there is significant effort in setting up master data and metadata for the acceptance and loading of data.
It’s an exciting time for CPG companies, and for the development of BI within these organisations.
Along with the technology challenges, there are a host of cultural challenges that also need to be addressed. But the prize is significant.