Sunny forecast - Planning/Forecasting/S&OP report June 2015

Manufacturing & Logistics IT spoke to a number of key spokespeople from the vendor and analyst communities about recent developments in the world of planning and forecasting-related software – including those involving mobility, Cloud/SaaS and Big Data.

The pace of technological change has been considerable in the demand planning and forecasting arena in the recent past. So, what have been some of the most prominent sweet-spots in terms of enhancement or innovation? From an Advanced Planning & Scheduling (APS) perspective, Barry Drummond, business development manager at DynaSys, makes the point that the development of in-memory processing is helping to deliver true 'real-time' planning and collaboration possibilities for APS users. "Vendors who are still relying on cumbersome import and export of data between functional modules, or lengthy waits for aggregation and disaggregation of data are being left behind in this area," he stressed.

Eric Kimberling, president of Panorama Consulting Solutions, considers that many current developments have come in the form of more complex workflows and pushing more data to the end-user for better decision-making. "Required inputs (e.g. Demand Forecasting & Planning, Advanced Planning & Scheduling, Finite Capacity Scheduling, Workforce Planning and Integrated Business Planning/S&OP. BI/Analytics systems etc.) have been augmented to include seasonality trends, organisational and customer constraints, and extended variable factors that are specific by industry," he said. "Including customer-specific information beyond sales (seasonal buying trends, carrying capacities, and regulatory reporting requirements) enables a business to enhance their services offerings to customers."

Kimberling points out that one of the largest national malt brewers, Rahr Corporation, readily admits that its success is due to Customer Service. He commented: "Alerts in changes of buyer habits are pushed to the end-user for review; is the customer launching a new product? What is the reason for such a change in orders? These insights are being integrated into other aspects of ERP such as CRM, Contract Management and Pricing."

Kimberling believes having a closer relationship with your suppliers is also critical. "Solutions now offer direct communication with suppliers via EDI and can align production/distribution schedules," he explained. "Businesses now have the ability to plan further out with all aspects of their supply chain in order to be 'proactive' when traditionally 'reactive' had been the baseline.

Supplier insights enable the supply chain workflow to run smoothly, with organisations gaining product availability and new product campaigns. More importantly, solutions are instituting Demand Sensing and Demand Shaping techniques that flow downstream from raw material intake through the production schedule and to distribution. If the supply chain is an oil pipeline, these demand techniques are grease that keeps it flowing smoothly. Grouping items with common demand requirements and having a balanced stock will provide a higher ROI, with software's ability to 'detect' potential variances in SCM and planning the end-user gets better information faster."

By integrating communications with suppliers and customers, Kimberling maintains that businesses can increase agility and flexibility in their rapidly changing markets. "More accurate 'what-if' scenario planning, using Social Supply Chain Networks for increased collaboration and accurate input data, and considering macroeconomic factors enable organisations to be respond faster than ever," he said.

In terms of user needs, Chris McDivitt, vice president, supply chain technologies leader at Capgemini, comments that demand planners do not necessarily want to spend time looking in hindsight at developing a statistical forecast. He elaborated: "The planner should be working with the most recent set of market intelligence that's available, having information about competitors, having information about how product is going to be promoted; leveraging collaborative relationships with their retail community (which is probably one of the big distribution channels)." McDivitt believes that by leveraging those different marketing intelligence channels – whether it relates to their customer service, marketing or sales organisation – companies can come in with the best demand signal before they walk it through their S&OP process.

McDivitt added that one of the advantages in this regard is in not having to put so much energy behind things such as auto assignment, statistical and the parameter settings. "There's still some work to do in this area, but essentially there are technologies available today that can help in this area," he said.

Additionally, McDivitt believes the market will see leveraging consumer insight as a component of market intelligence to help shape demand. "It's really about leveraging this consumer data to help sense demand better," he said. "If you can sense demand early on and know what that consumer sentiment is you could actually then shape demand by influencing the way you promote." However, he added that companies would have to do this in a very collaborative way with the retailers.

McDivitt explained that Capgemini is seeing this being put into practice particularly among manufacturers in the food and beverage space at the moment. "They are the ones that are blazing the trail," he remarked. "This is really an example of how manufacturers will increasingly leverage data and insight in a more significant way versus doing this in the traditional supply chain planning manner."

Tim Payne, research vice president at Gartner, reflects that, because of the way many companies' supply chains are evolving and the way their businesses are developing their planning processes need to be more integrated – i.e more joined-up sales & operations planning (S&OP), demand planning, inventory planning, replenishment planning, production planning, manufacturing scheduling etc. "Many companies and their suppliers now have supply chains that are more global," he pointed out, "and so these supply chains have become more uncertain, unpredictable and complex. For example, there are often more products involved, with customers having increasing expectations in terms of product choice or buying channels available to them."

Payne adds that the supply chain is also becoming more multi-enterprise. "It's hard enough for some companies to plan at an enterprise level – with joined-up thinking between the demand and supply side of the business – but then if companies outsource or want to work more with tier 1 or tier 2 suppliers or outsource part of their manufacturing operations etc. they are moving into a multi-enterprise environment where naturally the level of supply chain complexity increases."

Payne explains that the technology required to enable that will need to be able to work with more granular information, operate at greater speed and scale and have greater collaborative capabilities. "This multi-enterprise dimension becomes very important and that's why we're seeing a lot of the Cloud-based solutions moving more into this area," he said.

According to Payne, the type of planning many companies want to do is also changing. "We can certainly see that with the leading companies," he said. "These companies tend to be more mature in terms of what they are able to do with their supply chains."

Payne observes that there are hundreds of software vendors and hundreds of different planning solutions available on the market. However he believes there are in essence three main categories of planning capabilities or components to consider – configure, optimise and respond (COR).

Configuration (C) refers to technology or processes that help companies design their supply chain and, from a segmentation perspective, help them consider how they are going to postpone, determine which part is push and which part is full, and know whether they are going to be efficient and agile enough etc.

Optimisation (O) is about putting in place an optimal demand plan and optimal supply chain and trying to ensure the two are relatively well-balanced – or, if they are not, that the company has the right inventory in the right place to bridge the gap. This is really traditional batch-oriented sequential planning.

Then there is Respond Planning (R). "Having created an optimal plan reality is going to intervene," said Payne. "Companies know their forecasts and plans will be wrong to some extent; it's a question of how wrong. So they need a capability or mechanism by which in the short term they can respond to those changes in planning and execution events but in an intelligent way – in a way that keeps them as close to their original goals as possible."

Drummond reflects that things don't have to be new to be 'hot', and the continued importance of Sales and Operations Planning (S&OP) is proving that. He added: "Customers are engaging with us from different points of maturity in their S&OP process, with some companies just getting underway with the basics and others looking to develop their existing S&OP to encompass the themes of Integrated Business Planning (IBP) and Integrated Business Leadership (IBL)."

Drummond's view is that a good S&OP process will inevitably evolve over time; from simple summaries of demand and supply information, to 'what-if?' scenario planning and financial analysis. "The key message here, in terms of systems, is flexibility," he said. "Typically the initial choice of system is Excel in order to get the process up and running, but once you've outgrown this then you need something that can take you to, not only the next level, but the one after that and one after that."

Richard Watkins, managing director at the Delos Partnership Ltd., considers that many companies are confused on what Integrated Business Planning or SOP is. "Arguments rage around that somehow they are different, that IBP is SOP with steroids and an advance on SOP," he said, adding that, in principle they are the same thing. "Many companies still rely on a business planning model which rests on a process of creating an annual budget, take six months to do it, and then spend the 12 months of the financial year trying to achieve something that was put together up to 18 months ago. A review is carried out at most quarterly. Measurement of success is based on achieving the budget. The fundamental difference that IBP should bring is a formal monthly review of all of the business plans, and making sure they all join up. From a measurement point of view there are 7/8 key performance indicators which will help deliver the long-term strategic plan. This formal process happens monthly and works through all of the parts of the business."

Underlying this, according to Watkins, is a series of weekly planning activities which underpin the monthly process. "If there is a difference between IBP and SOP then it is that many companies who implemented SOP tended to focus on Sales/Marketing and Supply Chain," he said. "These projects are typically driven by the Supply Chain director trying to get a better forecast. IBP also covers the integration of Finance, Human Resources, Information Technology, Research and Development etc. But in truth the original SOP process as developed in the 1980s was meant to cover all of this."

Critical to the quality of this IBP process is good Demand Planning, maintains Watkins. "In this area there is a definite divide between ERP systems (which cover Order Entry, Inventory Management, Planning, Purchasing and costing) and specialist forecasting/demand planning systems," he said. "Few of the ERP vendors have a forecasting system which is entirely integrated with the Production Planning and Order Entry systems.

Those that do generally do not have the functionality of the specialist forecasting systems. To make matters 'worse' the specialist forecasting system vendors have seen 'opportunities' to add value to their packages, and so in turn have developed 'modules' like Rough Cut Capacity Planning, Advanced Planning and Scheduling and Finite Capacity Planning. So companies then implement these separate systems around or interfaced to their 'basic ERP' system."

Also critical to the quality of this IBP process, states Watkins, is a good reporting tool, which enables a company to aggregate the data into families, and then provide management with graphs, reports and commentary. "Here there are a number of Business Information tools which are stand-alone reporting tools," he explained. "These have to be configured to get the information required, and then need to be interfaced to their ERP systems to extract the information. In truth many people do not do this, and instead rely on Microsoft's Excel package to drive the information – which relies on dozens of peoples' individual abilities to programme macros and write complicated formulae – with all the probabilities of corruption, wrong formulae etc. causing potential damage to a business' financial position."

Watkins added that some companies are also trying to sell 'SOP' software or 'Integrated Business Planning' software, which is a further advance in this area, as it is built on a recognised process, does the adding up, and also is built around emails telling people to get the data ready for the monthly meeting. This, he believes, is another key development.

According to Watkins, Integrated Business Planning also needs companies to manage project type activities well, and also manage multiple projects (whether they are new product development, or new machinery and equipment). "In this are there is little in the way of integration of project management systems into ERP systems," he said.

Additionally, Watkins makes the point that managing the interface with the customer around enquiries, contracts, order entry, marketing activities and the like requires a Customer Relationship Management system. This, he states, should then be integrated into the Demand Planning process and the supply planning process.

However, Watkins adds that this is way off happening for most companies. He believes that, slowly, there is a move to provide Integrated Business Planning with the platform that it needs. "But there is still a long way to go before organisations can have a truly integrated high quality system to support Integrated Business Planning," he remarked, adding: "Back in the early 1900s Henry Ford famously came up with the idea you could get from A to B with a car, and drive. Cars now are very different things, with Sat Navs, automatic parking, anti-lock braking systems and so on. But the basic principle is that you can get from A to B because the engine, gear box, steering wheel, chassis and body and seats are all properly connected, and if the driver turns the wheel the car will go in an expected direction. To get to where we want to go we need to know where we are going, a plan of how we are going to get there, and a measure of our speed, petrol consumption, and fuel levels to make sure we get there safely."

Watkins makes the point that Integrated Business Planning is like the 'knowing where we are going', and having a plan to get there with measures to know how well we are doing. "But in many manufacturing organisations companies still need (if you look at the systems this way), to buy engines, gear boxes, chassis, Sat Navs and bodies from different people. There are still some motor car enthusiasts who build their own cars, but they are a minority. That is not the case with manufacturing companies. Certainly very few people would operate like Excel and take all of the individual parts of a car and assemble them themselves in their own garage."

Nick Castellina, research director for the Aberdeen Group's business planning and execution practice, comments that one key talking point in the industry at the moment concerns the integration of tools such as MES with ERP to get a more full view of a company's plans for production in the near future, and how that integrates with demand planning. "This type of integration is driving down the amount of time companies' inventory is staying on the shelf, and this is hugely important for driving costs down," he said.

Castellina added that Aberdeen Group is also seeing a lot more embedded analytics capabilities built within the solutions to make intelligent decisions. "Certainly, as organisations grown larger they are more likely to be taking advantage of these types of solutions," he said. "Some of the smaller organisations are having difficulty; they are still running their shop floor on paper. Even if they have these software tools they are still producing one thing, planning for another and going back at the end of the day and making those changes within the solutions. So I think there's a need for organisations, or the leaders within these organisations, to commit to using these tools that are available to them, particularly after they've purchased them."

Hans-Georg Kaltenbrunner, VP industry strategy EMEA, JDA Software. Interactive planning, makes the point that Agile Control Tower (ACT) capability enables companies to detect and respond to significant supply chain events rapidly and profitably. He added: "This capability allows planners to move from 'What do we do now?' to 'What is the best alternative of those that were tested and reviewed?' With guided resolution options, decisions are effectively made and implemented, and using in-memory power, clients are able to rapidly respond to disruptions during execution."

Kaltenbrunner then moves on to the topic of business & operational scenario planning. "Strategic and medium-term scenarios can be modelled for operational validity, and the results made visible across the teams," he said. "Moreover, the different scenarios can be optimised for cost and business outcome. Utilising advanced simulation models, planners are now able to 'try out' multiple options without committing to plans – and intelligently decide on – and execute against the most effective and profitable paths."

In terms of the subject of integration and ease-of-use, Kaltenbrunner explains that this facilitates effective user adoption and reduces change management requirements. "Users expect simpler interface, a more 'social' style of interaction and want information and insights 'now'," he remarked. "Planning solutions must adapt to user needs and deliver friendly, social aware offerings. Recognising that not all organisations will utilise best of breed, being able to provide connection into and through other systems is essential."

With regard to ERP integration, Kaltenbrunner considers that planning solutions need to be ERP aware while at the same time agnostic. "Many organisations operate multiple ERP systems, particularly those with historically high levels of M&A activity," he explained. "It's important to move beyond just being able to connect, and into having guided pathways and processes to minimise cost, impact and risk."

Barnaby Speller, product manager for SIMATIC IT Preactor at Preactor International – A Siemens Company, considers that the concepts of Industry 4.0, digital factories and Manufacturing Operations Management (MOM) mean that Advanced Planning & Scheduling (APS) is no longer being seen so much as a bridge between level 3 and 4, where it communicates directly with both MRP and MES, but rather as a component of the complete and integrated MOM portfolio within level 3. He added that the SIMATIC IT Preactor APS range is part of the Siemens MOM portfolio, which includes among other solutions, MES, to better meet the end user's needs. "This means that the APS system can, if an MES system is present, take advantage of the more manufacturing-centric data from there to produce schedules that accurately reflect the reality of the shop floor," he said.

Motivation for change

What has driven these types of trends and developments? Castellina thinks it really comes down the speed of doing business to increase competitiveness. He remarked: "Every company is subject to a large amount of competition due to the way that people buy things today; ordering over the Internet as well as in-store. A company's competitors aren't necessarily in the same town, same state or even in the same country; they could really be anywhere in the world. So I think being able to produce goods quickly and getting them to customers when they need them is essential."

Payne observes a trend towards bigger supply chains from country to region, from region to global. Coupled with this trend, he sees greater levels of uncertainty and unpredictability in terms of commodity prices, exchange rates, weather disruption, political unrest etc. "Therefore, companies need better visibility across the supply chain from a planning perspective," he said. "The greater complexity of supply chains – involving more product options, routes, multi-enterprise dimensions and the need to obtain better insights – means companies potentially have a lot more data to work with." This, he believes, is where idea of Big Data and Big Analytics could increasingly play a key role. Payne added that with more granular information regarding point of sale data, partly due to different buying channels, the need for a greater level of analytics comes into play in order to get the appropriate insights required to make more informed decisions.

Drummond reflects that there are still many companies that are trying to get the basics right. "They have usually been burned at least once on a failed APS implementation and are wary of starting another," he said. "Meanwhile the forecast accuracy stagnates and they struggle to get a step change improvement in customer service or inventory levels." Drummond adds that vendors with good R&D and strong Supply Chain consulting skills are able to offer a simplified APS package in order to help companies get the 'quick wins' by establishing a robust process based on industry best practice. "This can then be used as a platform to develop the system and solve more complex issues once it is embedded as part of day to day business," he pointed out.

According to Watkins, the end-user requirements have changed in part over the years. He commented: "Customers and consumers have wanted more products more quickly. Shareholders have wanted more profits, or to keep their current profits. Marketing activities to keep a competitive edge have meant more rapid changes to demand, and required more responsiveness. So whereas 20 years ago an MRP system might have recalculated once a month, now an ERP system has to do so in 'real time'. Once upon a time all cars could be black; now they need to have all sorts of different options fitted to them. The Lean philosophy has helped considerably in forcing people to simplify, but 'mass-customisation' is the buzzword the people follow. This means you make a pink one followed by a green one followed by a yellow one all in the same day. And you can only make a pink one providing it does not follow one that has nuts in it – because of regulatory requirements. Hence there has been a real expansion in Advanced Planning Systems and Finite Capacity Planning Systems to cope with complexity."

Watkins pointed out that some software companies – because of the development of 'Integrated Business Planning' – have started to market 'IBP' systems in response to their clients wanting integration. However, he explains that there is a real gap in this area, which is that few clients and companies have really bought in to the idea of true Integrated Business Planning and Management – "thus they have an ERP system which is linked to the Forecasting and Demand Planning system, which is linked to the Project Management system, which is linked to the Customer Relationship Management system," he said.

Watkins added that because many end-users do not understand and buy-in to this concept it is still the case that organisations have the ability to pick and choose the different systems, and hence develop disintegrated processes with which to run their companies. "Education in and acceptance of a completely integrated business model – in terms of systems, processes, behaviour and organisation – is still a dream," he remarked. "Phones and watches that not only allow you to call people but also to monitor how healthy you are have become the must have objects. That is because the inventors have looked ahead and developed things that consumers might want. The same is not true in the manufacturing IT world."

Jules Morgan, head of insight and analytics, consumer products & retail at Capgemini Consulting, points out that in the US and Europe there is currently a substantial amount of money being spent on promotions. "Grocers, the retail industry and the consumer products industry are under a lot of pressure to discount at the moment," he said. "This is pushing a lot of interest in how can these companies can target promotional dollars to the individual rather than to the mass, and to adopt technology available today, as well as keep an eye on technology trends and opportunities that may be just on the horizon. Take concepts such as Big Data, which might encompass social data and mobile data and a host of other sources of structured and unstructured data; Natural Language Processing (NLP) technologies have been available now for some time but the market demand to bring this data together, integrate it and understand it is very compelling now."

Morgan added that Capgemini has partners who have machine learning capability embedded in their tools, and Capgemini is able to build machine learning into tools so they can constantly learn about consumer vocabulary in the social media world, and constantly learn ways to better predict consumer buying trends. "There are waves and layers of technology that can be exploited in this space; even in terms of Mobile Device Management (MDM) and virtualisation," he said. "What is now possible in those spaces are things that people would have dreamed of ten years ago."

Another thing that manufacturers and retailers need to bear in mind, according to McDivitt, is that there are so many different channels through which consumers can buy goods today. "This is really being driven by digital forces such as the Cloud, social, mobile and Big Data," he said. "Therefore, there is not just a case for putting in place better sensing technology to monitor and anticipate purchasing habits related to certain types of goods, there is also the importance of determining which channels will most likely be deployed by the consumer when they decide to make a purchase."

Kaltenbrunner considers that today's user community is demanding a more interactive way of planning compared to the black box optimisation of the past. Another driver for change, according to Kaltenbrunner, is BYOD (Bring Your Own Device) and the proliferation of mobile devices, which has created the trend towards mobile applications with access anywhere anytime functionality. "Having applications and offerings being mobile aware and responsive is a must have," he said, adding that solutions able to deliver quick insights through dashboards, KPI's and high level reports are essential for the mobile executive – as well as the board level who look for the overview.

Kaltenbrunner also reflects that consumer centricity is critical in understanding what different types of consumers buy, the influences that drive those buying behaviours, as well as how and where they buy. He commented: "Responding to these needs requires evolving strategies supported by segmentation; with the proliferation of the omni-channel world, customer demand and volatility – being able to develop and manage multiple supply chain 'instances' is critical to supply chain success."

Speller observes that the APS market has matured considerably over the past five years. He maintains that users have a much clearer understanding of what is available in terms of functionality in the marketplace and higher expectations than ever before. "With today's global competition, the pressure to increase production capacity and shorten lead times while lowering costs and inventory, forces companies to find ways to optimise operations and reduce delays," he said, adding: "In some cases, we've seen customers who have tried to achieve this with their existing ERP/MES systems before turning to a SIMATIC IT Preactor APS system, which then enabled them to succeed in reducing or eliminating non-value-added activities and improve production efficiency, in order to better meet market demand and deliver on-time."

Best of breed

Is there an argument for best of breed solutions in favour of planning functionality built into systems such as ERP? McDivitt believes there remains a rational set of reasons for both to exist in the marketplace. He commented: "There are certain pockets of best of breed that play very well within a supply chain planning solution; one example would be a manufacturer's need to get closer to the shelf and understand what the likely profitability or performance of a certain assortment of products on retailer shelves. So there is currently quite a push in the market around not only having in place a traditional category management function but actually doing it in a local and collaborative way with their retailers. To achieve this really requires best of breed technology versus the traditional ERP route."

He adds, however, that ERPs are getting stronger. "One of the key attractions of ERPs is that they have the access to real-time integration through orders and inventory by bringing in their own set of planning capabilities; typically on the same kind of enterprise architecture as their ERP environment – so, they can offer more seamless integration." According to McDivitt, another argument in favour of ERP becomes apparent when a company's IT department asks 'why do we want to manage multiple vendors?'"

Morgan believes the question of system maturity is important. "We have big clients that have elements of ERP doing part of the S&OP process but have holes that could be filled within their ERP solution such as reconciling the finances at the back end of a trade promotion," he said. Also, he considers that these big solutions don't always finesse the answers to problems as well as a best of breed. "In terms of trade promotion management an ERP will give you the probability score out of TPM, but when you pull the data off and do the analysis through a best of breed solution or do the analysis through data science you'll find inaccuracies – you'll find new trends and new triggers that just aren't in the ERP tools," he remarked.

Payne makes the point that there isn't really a complete supply chain management suite available that can do everything, including more advanced analytics; at least not in an integrated way. He commented: "In order for companies to get their planning operations to work as effectively as possible they need a strong planning foundation – what we call the planning system of record – which includes demand planning, inventory planning, replenishment planning and manufacturing planning as the main components. You really don't want to break that up and approach one vendor for demand planning and another for replenishment planning and manufacturing planning." He added that if a company did this it would likely suffer problems such as latency issues etc.

"What you really want is an integrated suite," he said. "Then, if you want functionality such as demand sensing for advanced promotions management, doing analytics regarding sales uplifts and the amount of profit that you're going to make, then that's going to require additional technology which you could use in conjunction with the planning system of record."

Payne added that one could think of the planning system of record as providing the 'pipework', collaboration, visibility and the integration backwards and forwards from a planning perspective. "You might then wish to put specialised third-party applications on the top that will pull data from it to do the more advanced analytics," he said.

Castellina points out that, typically, Aberdeen Group research finds that users are better off having one single comprehensive solution that can handle as much of their operations as possible. He commented: "There are certain cases that many of today's ERP solutions can handle, such as sales. For scheduling and planning the preference is for this all to be built into the ERP system and I think ERP vendors are doing a good job of identifying the needs from industry to industry and building those needs into their solutions. In the enterprise ERP space those type of functions are often built by the vendor's value-added resellers and partners, while in the case of the smaller ERP vendors they are identifying which industries they are going after and building solutions specifically around the requirements of customers in those industries."

Drummond's view is that ERP functionality still lags far behind that offered by a best of breed APS. "We do on very rare occasions see organisations whose operations and products are very simple and they can plan with the rudimentary features of an ERP system, but this is very rare," he said. "The key input to all planning functions is the forecast, and the accuracy of this depends on the process as much as the system. An APS offers far greater functionality to support the process in terms of data cleansing, lifecycle management, aggregate level forecasting, collaboration, measurement and exception management."

Kaltenbrunner comments that leading companies have selected ERP solutions for different use cases and different user communities. "In today's world of an increasing number of M&A transactions, change of ownership and corporate ERP strategies, a planning layer and platform can help to bring the new combined organisation and supply chain together very fast," he explained, adding: "Acting with a single face to the customer is often seen more critical to business success of a merger than bringing the set of books together. S&OP and Integrated Business Planning can actually help to re-align on new corporate goals and supply chain strategies, better and faster as these solutions are independent for actual financial ownership of the assets."

Kaltenbrunner points out that having in place a segmentation strategy enabled by an order promising solution following the same strategy and rules is enabling customers to offer differentiated services levels to individual customers or channels. "In fact, the requirement to manage large scale and complex supply chains has gone up in many industries which require a more sophisticated solution approach compared to what usually is available by other solutions, e.g. ERP," he said.

Kaltenbrunner added that, with segmentation, organisations move away from a static 'one size fits all' effort. "Instead they can establish multiple value chains to handle different markets, business volumes, geographies and channels, as well as differing customer expectations," he pointed out.

Speller's view is that whilst many ERP systems have improved built in APS functionality recently, there is still a place for best of breed solutions. He commented: "These solutions are still set apart in terms of depth of functionality, performance and flexibility, depending on the user's requirements for adding further constraints and customisations. The recent trend for acquisition of best of breed APS solutions by manufacturing software vendors bears this out."

SaaS/Cloud

Has the Software as a Service (SaaS) model, and the Cloud concept in general, had a notable level of impact on the planning-related software solutions market so far? Watkins believes SaaS/Cloud solutions are beginning to take root, but are at an early stage of development. His view is that security of data is the principle stumbling block – although he recognises that there are clearly plenty of security systems and protocols.

According to Watkins also made the point that, with regard to on-premise ERP, the pricing mechanism is changing. "For years the model has been pay a large licence fee – say £1 million including named user licences, and then pay a maintenance fee of 20 per cent – £200,000 per annum. On top is the consultancy and implementation cost. Now some companies are charging a fixed monthly fee to cover licence and maintenance. This can ease the purchasing position, as companies no longer have a large capital expenditure to make, but can treat it as revenue expenditure. In the past many companies have been put off implementing a 'proper' ERP system because of the big capital cost. If this becomes less of a barrier then that is going to help."

However, Watkins maintains that the real challenge to the IT vendors will be to develop a completely integrated system in the Cloud; then each company could have their data managed anywhere at any time. He commented: "For this to happen, then every company will have to accept that the system in the Cloud has the right logic and works in the right way, supports good demand forecasting, good supply planning, and that they cannot have their own tweaks to make the system run in a slightly different way. That will be far more difficult, and is a barrier to a single ERP Solution in the Cloud."

Drummond observes that the Cloud model is still polarising opinion in the market; with organisations either definitely for or against it when it comes to APS. "Companies with an established Cloud strategy are actively encouraging the adoption of Cloud-based Supply Chain Planning tools, whereas those who have yet to make the move do not see the implementation of an APS as a driver for moving to the Cloud," he said, adding: "There is still work to do in educating customers on the cost-benefits of Cloud. It is a much more transparent model as the total cost of the software and IT support is provided upfront. However it can be initially viewed as more expensive when compared to an on-premise solution where the IT costs are hidden within the fixed costs of the organisation and soft costs such as people's time are overlooked."

Kaltenbrunner makes the point that utilising Cloud deployment allows businesses to focus on their business, and be able to grow and evolve without having concerns about their technology scaling. He added: "Deployment in the Cloud has become the default delivery method for JDA offerings. JDA has been offering these capabilities for almost a decade and has proven pedigree with more than 300 Cloud deployments to more than 350,000 users across retail, manufacturing, wholesalers/distributors and transportation worldwide."

According to Speller, the main change Preactor has seen is new requirements in terms of integration to accommodate offsite solutions. "I believe that over the next few years there will be increasing pressure on APS vendors to provide Cloud-based offerings, both to meet with customer expectations and to meet the requirements of global multi-plant companies," he said.

In terms of planning capability, McDivitt cites two areas worthy of analysis: Cloud/SaaS Planning and Cloud/SaaS ERP. "In terms of Cloud/SaaS Planning, we're definitely seeing momentum in this space; more so in retail than in manufacturing but it is certainly becoming increasingly deployed in both areas," he said. "Cloud SaaS planning is gathering interest because the Cloud SaaS players around planning are basically delivering optimisation capabilities across functions on a single enterprise solution platform. For example, companies can create a merchandise plan that feed into a company's demand plan that feeds into the way it replenishes and allocates. The company can also do its 'open to buy' and promote and price in a single enterprise platform. This is quite an attractive proposition, and that's why we're seeing momentum."

In the Cloud ERP world McDivitt is seeing Cloud/SaaS solutions gain momentum across sectors such as manufacturing high-tech and retail. He commented: "We partner with one of the larger Cloud/SaaS providers, and the rational for deployment includes the scenario whereby you might want to extend your ERP capabilities across different regions from the 'mother ship' of the solution framework. Many companies want to do this in as nimble and light a way as possible, so we're seeing momentum here. We're also doing deployment around Cloud/SaaS ERP globally and specifically within certain retail enterprises within the US."

Castellina believes the Cloud is certainly here to stay. He commented: "I think it's shortly going to become the preferred solution for manufacturing and also distribution organisations; the latter currently being faster to adopt Cloud/SaaS than manufacturers, who are a little slower to come over. However, many manufacturers are getting to the point where they have identified the need to replace their existing solutions and often this leads to them adopting a Cloud solution. There are a lot of benefits in the Cloud beyond just cost; for example, the ability to access information wherever you are in real-time can prove visibility in collaboration."

Payne explains that if one visualises a 'Cloud sandwich', the lower slice of bread could be thought of as all the functionality that provides the visibility outside of an organisation; the end-to-end supply chain visibility showing what is happening with a company's suppliers, contract manufacturers and so on. "Today, most people expect this to be Cloud because we're talking about the multi-enterprise and there's a community aspect to it as well," he said. Then if we think of the upper slice of bread as S&OP this, says Payne, can also involve collaboration with other parties; key customers etc. so there can be an external element to this as well. Therefore, this area is also fairly Cloudy. Then, according to Payne, the filling in the sandwich could be thought of as operational planning – production planning and scheduling. This, he said, is still predominantly on-premise for most companies.

Payne added that Cloud solutions tend be multi-enterprise, fast and scalable. Because of this, he believes more and more companies are looking to invest in Cloud applications and expects to see above market growth rates moving to Cloud in all categories; including operational planning. However, he points out that there are industry differences. "For example, in process industries that rely heavily on their operational planning and scheduling – chemical plants, pulp and paper manufacturers etc. – they often think it's still important to have operational planning tightly integrated to their ERP and MES system," said Payne. "Also, a defence company might say it can't have its data outside its operations due to perceived security concerns." Nevertheless, in general, Payne believes the uptake of Cloud is on the rise.

Big Data

Is Big Data having an effect on the development and benefits of today's forecasting and planning-related solutions? First, Watkins puts the Big Data concept in context by pointing out that there is masses of data out there and companies need to be able to process it – and by processing it they can come up with more valuable information which can save them time and money. "All Demand Planning and Forecasting systems must rest on assumptions," he said. "One basic assumption is that the price you charge for a product will influence demand. I see that companies can now collect information across multiple stores and multiple customers/consumers and can analyse this data to see if this assumption is true."

Watkins also commented that demand is also influenced by promotions. "There are lots of different promotions [BOGOF, end of aisle, price marked make, etc.] and now there are promotional evaluation systems which allow you to collect lots of data to then analyse if these actually have an impact and what impact they have," he said, adding that all of this goes under the heading of 'Demand Sensing'. According to Watkins, one of the areas that is slowly developing is the ability in the FMCG field to be able to collect data on sales by supermarket store, and then work out a forecast by store for each product – then translate that into a distribution plan from factory through distribution centres and into the store.

Watkins added that we have not yet quite got to the point where RFID tags on individual packs in each consumer's fridge will communicate when the customer will want a product because they have run out, but he believes this could happen one day. However, he believes the key in this area is that there is a major distribution company that is committed to developing forecasting and planning systems such that they will be able to predict exactly when a customer/consumer will want a product slightly before they need it. "That will only be possible by being able to analyse and process Big Data," he said.

Castellina believes that for the organisations that are able to afford a solution such as SAP's HANA there is certainly a user case in terms of having more data and being able to analyse it within a quicker timeframe. "I think this is something that is going to be increasingly important in the near future," he said. "It really comes down to whether your organisation can handle that extra data from a sophistication standpoint, and whether you can pay for it."

Drummond considers that the danger of Big Data in an APS is too much information can cause 'paralysis by analysis'. "Planners spend too long trawling through information and not enough time communicating and collaborating on the forecast, where the big benefits are," he said.

Kaltenbrunner explains that many FMCG companies are looking into the subject of Big Data and trying to utilise the condensed outcome into their mid- to long-term forecasting as one of many factors. He commented: "It is our understanding that this input is mainly used to detect early change of consumer buying preferences or to predict seasonal changes. B2B companies are using Big Data analysis to better understand their customer's consumer markets and to be ahead of the game where buying behaviour and product characteristics are likely to change. In some cases B2B suppliers take the lead in product innovation for the end product to suggest new features and functions. We've seen examples in the semi-conductor and automotive industries for these innovations. From a development standpoint, solutions need to be able to incorporate and respond to the changes seen and captured through Big Data."

Kimberling reflects that the initial concept for demand planning and forecasting was mostly inventory-based; and that confidence in the appropriate inventory level has always been fundamental in the simplistic supply and demand economic model. "Too little inventory and you risk losing sales, too much inventory and you risk higher costs," he said, adding that this required specific data inputs (historical sales, projected inventory).

However, Kimberling explains that recently software solutions have enabled businesses to not only capture mass quantities of information from Big Data but manipulate and integrate that information back into Sales, Pricing and more effective Distribution practices. He commented: "Big Data has become the new undervalued commodity with an oversupply of information that has not been fully utilised. That being said, the value in Big Data is tremendous. The increase in access to information enables groups to have more insight, visibility to trends, and as a result provide a valuable service for a profit. With the American economy heavily weighed on services over manufacturing there has been a growing demand for 'service-oriented' business planning and Supply Chain Management."

Speller sees the relationship between APS and Big Data as very much that APS can be a source of data that can be aggregated into Big Data and then analysed by tools such as Siemens' Omneo Performance Analytics to deliver product performance intelligence. He pointed out that Omneo monitors data across the entire supply chain and customer experience, using data from sources including field service, manufacturing, CRM, ERP and the Internet of Things (IoT).

"Big Data is to some extent influencing what the vendors do," said Payne. "The idea of Big Data and Big Analytics are very much on the business intelligence side and the general analytics side of most organisations. However, it would be fairly rare for a company looking for supply chain technology to say I need a Big Data demand planning system; it is more likely to say it wants a very scalable demand planning system with demand sensing capability because the company has a lot of sophisticated data to look at and so needs a big repository, and wants a system that is fast. So there's a lot going on in terms of scale, speed, granularity and the growing demand for things such as in-memory databases."

Payne adds that another area where Big Data is becoming more prevalent relates to the idea of end-to-end analytics; which might include things such as end-to-end design work. Also, Payne points out that S&OP providers such as Steelwedge are helping companies to work with a lot of planning data, and have technology roadmaps that incorporate various types of functionality to help with the Big Data side of things.

McDivitt believes Big Data is definitely having an impact on the market intelligence side as well as in terms of more and more organisations leveraging the data signal repository capability and the demand sensing capability. "And I think what is also coming is being able leverage the unstructured component that would allow companies to sense demand to an even greater extent," he added.

In terms of the definition of unstructured data, Morgan points out that this could include things such as weather patterns, which could, for example, have an impact on the sale of certain types of food or beverage products. However, he adds that, interestingly, weather data could also be structured to some extent. "We are increasingly seeing Natural Language posting (NLP) tools structuring that otherwise unstructured data to make sense of it," he said, "and when this is done the data actually behaves in the same way as structured data. There are still huge volumes of this data, but it will increasingly become more manageable."

Planning the future

What do our commentators believe will possibly be some of the key developments in the world of planning-related solutions over the next year or two? Morgan believes the individualisation of promotions and the ability to more fully exploit that data is definitely coming. He commented: "We talk to big CPGs and FMCGs and most if not all of them over the next five years intend to keep records on around half their target base – and they will be able to exploit this information in planning, forecasting and promotions in a variety of ways."

McDivitt's view is there is going to be an increasing challenge in trying to predict which buying channels the consumer is going to utilise. "This to some extent challenges the relevance of forecasting, or is going to be more reliant on predictive analytics than traditional historical sales forecasts," he said. "And that's because the market is now so dynamic around the consumer in terms of where they are going to be and how they are going to be buying goods." McDivitt also foresees greater movement of learning how to leverage Big Data and predictive analytics in trying to sense demand from the perspective of what channel that consumer is going to use.

Payne believes that, from an end-user perspective, we are close to the tipping point in the market where the old view of planning that started with MRP and then moved to MRPII and ERP – preparing a demand plan, passing this information to distribution and production etc. – is no longer going to be good enough for end-users' needs.

"Companies that are at a low stage of maturity could still start here, but some companies that have been relying on that traditional model for many years are now saying they've pretty much exhausted what they can do with it," said Payne. "This is understandable because, as I mentioned earlier, many of them are now operating in a more global, complex, uncertain and volatile environment." This, explains Payne, is why Gartner put together the configure optimise respond (COR) model. "We see that companies are increasingly wanting to be able to integrate those things together," he said.

Watkins makes the point that traditional forecasting and planning solutions tend depend for their existence on creating Time Series analysis, and also some statistical analysis. He comments: "They require marketing intelligence to produce a valid forecast. But that marketing intelligence should be based on a real analytical engine that enables you to look at the data, summarise, understand the relationships between things and then influence the demand to meet business objectives."

However, Watkins believes the real requirement should be to provide real integration between all the separate parts of the business so that Integrated Business Planning is truly enabled, and companies can get away from relying on spreadsheets with which to run their businesses. "Once the system side is sorted, then we can get to the point where we can integrate people in the organisation, because many companies still operate in cars without sat navs, engines from one maker, bodies from another, gear boxes form another all put together in their own garages," he remarked.

Drummond foresees a continued focus on supporting S&OP and its evolution into Integrated Business Planning (IBP) and Integrated Business Leadership (IBL). "Companies will want this functionality integrated with their APS where the demand and supply information already exists without bolting on yet another system," he said.

Drummond also anticipates more scenario planning and 'what-if' capabilities within APS to see the impact of changes in demand and supply and plan accordingly. He also sees the further development of functionality to support collaboration across the supply chain with internal and external stakeholders. Additionally, he foresees development in the area of web-based user interfaces for reporting and adding intelligence to the plan, coupled with full-feature 'rich-clients' for traditional planning functions.

Additionally, Drummond maintains that the improvements in technology must be matched by an investment in training and development of staff working in supply chain functions so they can truly maximise the value from their APS systems.

Kaltenbrunner believes the key innovation drivers will continue to be user enablement, ease-of-use and mobile front-end to offer interactive planning as well as sophisticated optimisation across organisational boundaries and countries.

He added that implementation of advanced collaboration options among trading partners through Flowcasting is particularly interesting and will drive tremendous value across the supply chain. "By connecting the planning and execution processes to what is actually happening with store level demand, inventory visibility can be increased and products can more effectively flow through a synchronised supply chain," he said.

Kaltenbrunner also points out that by utilising technology looking at point of sale data, partners have the ability to collaborate effectively to drive processes for better product availability and less excess inventory on the shelf. "By reaching consensus across trading partners based on true demand, a true single, agreed on forecast can be developed and executed against," he remarked.

Other areas where Kaltenbrunner anticipates further enhancements include greater flexibility of solution configurations to enable quick changes to correspond to ever-changing business requirements, and consistent business rules from supply chain strategy definition to operations planning and optimisation down to order promising and fulfilment.

Speller comments that the recent acquisitions of APS solutions (including Simatic IT Preactor APS) by companies like Siemens with wider portfolios in the MOM space, point to APS being considered as much more closely tied to the execution layer and an important part of a complete digital factory solution such as the Siemens MOM portfolio can provide. "This brings benefits in terms of quality of data for the scheduling model and also a tighter loop for shop floor feedback, giving a better representation of the production status when creating and amending schedules," he said, adding: "The move to Cloud and centrally administered solutions will also see the expectation that scheduling solutions will be made available in this way, particularly across large multi-site companies that have other business systems set up with a central implementation serving the appropriate data to multiple sites."

Due to the exponential growth rate of technological advances and information availability, Kimberling reflects that it is difficult to forecast future developments. However, with so many users on social networks he believes it is easy to see the logic in Microsoft buying Yammer for US$1.2 billion in 2012. "Social Supply Chain Networks are bringing collaboration, information and availability to all individuals within an organisation or association," he said. "First, we had to capture and gather information. Next, organising and interpreting captured data. Now we can manipulate and simulate data for even better business decisions. Moving forward based on current internet user trends, it is logical to assume that sharing all that information with anyone in the world will become ever more critical."

Castellina thinks the future will largely be about driving more embedded analytics into the solutions. He considers that Big Data is certainly a part of this; as is In-Memory analytics. "So it's really about being able to make decisions based on large amounts of data very quickly," he said. In terms of adoption, Castellina believes part of this is simply going to come down to the commitment of those on the shop floor in the line of business to ensure they can actually drive value from the solution. He commented: "If you see data regarding the average age of employees in manufacturing organisations and look at the individual groupings to see where the largest percentage is, the majority of them are reaching retirement age, while the smaller percentage is just entering the workforce. Therefore, I think manufacturing is going to have to do a better job of identifying talent and getting that talent on board in order to impact on the organisation."

To sum up, Castellina remarked: "It's really all about the convergence of technology with operations, and making a commitment to utilise the available technology that will impact on manufacturing the most."