Supply Push to Demand Pull – How much is IT putting in?

Supply push to Demand pull – How much is IT putting in?Companies, in the ‘historic’ days, based their supply chains on a simple formula: ‘Push in more and the supply is gonna make its own demand’. Today, supply has been deprived of this ‘demand-creating autonomy’. A lot of things along with higher availability are required to push up demands. Advertising, broadcasting, promoting and not to forget – twittering are some examples. Here, producers need to make more, stack up shelves and then wait till their makings get in the shopping carts.
However, if this model is turned upside down – the customer asks for a product, the company makes it and it is delivered – It becomes a demand-pull. Somewhat like a ‘dial-up’ pizza shop, where you can get the topping of your choice.
However, it seemed impossible for big companies to switch over to the demand-pull model, most of them were worried about drop in sales or high implementation costs. When Dell Inc. introduced the mass customization order processing systems, others followed. Since then, a large number of companies are giving preference to get customer orders first instead of piling up huge inventories.
However, getting orders is just a slice, making it fast and delivering it fast makes the whole cake. And to make this cake, companies require highly integrated business process workflows to keep their suppliers, distributors and customers at their tips – all at once.
This demand – pull model is not only helping companies reduce inventories, but also helps achieve high levels of customer satisfaction through responsiveness and flexibility. However, there are some drawbacks to going Rambo with ‘demand-pull’. Companies need to spend a fortune for a responsive – ad hoc system. Next is to persuade suppliers and distributors to join the ‘make-over’ – one of the most difficult matters to deal with.
What do you think –Is demand-pull applicable to every business model, or is there room for the supply-push strategy today? How can IT contribute towards automating both paradigms effectively, and efficient transition between them? Comments appreciated.


  • Pull system is a very powerfull structure, but I feel it cannot be applied to every business modal. Pull system is a part of lean managment system. For application of pull system customers need to be educated and customer have to wait more for the product. For details you can refer DELL business model.

  • Interesting question, Murtaza. I am not sure there is an entire supply chain (from base materials through to end consumption) that can be entirely pull-based with no ‘push’ at all. The logistical obstacles imposed by geography and transportation are too great. The supply chain needs some quantity of stock to get it primed.
    Instead, demand-pull solutions are enhanced with carefully placed stock in the supply chain to satisfy 2 main concerns: first, some customers simply won’t wait for the supply lead time, so stock must be available for them. Conversely, others will wait in many situations. The second consideration is demand variability. It is not clear just what demand exists, for which items, in which locations, in what quantities and when it is required. The forecast doesn’t help us at this level of detail. Fortunately, most supply chains have common components and materials that are used in a variety of products; they also have relatively few sources of supply, shipping to many points of consumption. This allows to aggregate consumption to a few places, where the forecast accuracy improves dramatically. In these places, we can more safely stock product, and respond only when real consumption has happened. We call it consumption-driven replenishment, however you can interpret it as a pull system with an initial push of supply.
    The next challenge becomes creating a simple mechanism to respond to changes in demand over time. Consumption-driven replenishment buffers use a few rules to separate noise from real trends, and adjust stock levels up and down according to trends. When demand dries up, stock targets are reduced, when demand accelerates, stock levels may go up as well (or if supply can keep up, inventory turns increase).
    Supply chain managers can push for better replenishment approaches than the standard offerings of min/max, safety stock, or electronic kanban. Check out the software link below for a constraints-management supply chain solution.
    Additionally, Eli Goldratt recently published a book called Isn’t it Obvious, and excellent read on how retail can benefit from demand pull.
    Peter Milroy

  • It’s really a balancing act b/w having to repeatedly incur order processing cost VS cost incurred in stocking the inventory… Read EOQ (Efficient Order Quantity) to see how this is quantified at least in theory.
    In practical supply chains however, as Peter has rightly pointed out, there are too many variables to contend with and stock buffers might have to be created and maintained at key points in the supply chain to achieve an optimum ‘consumption-driven replenishment’ specific to each supply chain. Also taking in to account the fact that some of the stock buffers might supply to multiple demand pulls and applying stock consolidation strategies around it.
    As far as IT is concerned, the key then is to set up an enterprise architecture (or enterprise service bus) that provides both ‘push n pull’ levers and services.
    An emerging middleware technology that facilitates this is called BAM (Business Activity Monitoring) which allows you to build a real-time operational intelligence of your supply chain. Using BAM, you can set different data and process sensors and thresholds at various points in the chain and take replenishment decisions accordingly.
    BAM works beautifully hand-in-hand with workflow automation technologies like BPEL (business process execution language) to allow for a dynamic workflow automation based on real-time operational intelligence.
    Hope that helps!
    Best Regards,

  • Great question and very enlightening by all. Thanks!!!

  • Murtaza,
    As a consumer goods professional, I ask your question about which strategy is more appropriate in a slightly different way. First, I want to know if I am dealing with a demand driven or supply constrained model (i.e. is there more supply than demand potentially available or is the reverse true). As I deal in the demand driven model more frequently, then I ask for the end consumer, how is the demand generated (e.g., impulse, outside event, long term process), what is the expectation for how fast it will be met, and what is the desire on how fast it should be met. In consumer packaged goods, the answer to the last two questions tends to be “immediately”, in which case the retailer/wholesaler/manufacturer supply chain must support that without a perceived increase in cost. But for a larger purchase, or a service that directly involves consumer participation, the answers might be different – and as cost becomes more important than delivery and driven by delivery time, the consumer answers will vary (what is one willing to pay for faster delivery). In a supply constrained model, cost is less important (because if it were too high, the constraint would shift), and maximization of the supply for the purposes of the enterprise (e.g., more sales, maximum profit, keeping the demand or “buzz” high) might become more important (for a case study on this, look at the Harley Davidson model until the mid-2000’s). Typically, in this case, push is appropriate, because there is already more demand than supply, and inventory does not need to be held (unless slowing down delivery is part of the strategy).
    In how IT can help, there are two areas. First, if all information were available to all parties (including the consumer) immediately (regardless of the model), then everyone would have the ability to act more efficiently and not need to spend time “tracking” and a lot less time on “planning”. The cost of this “nirvana” might be prohibitive, but technology professionals working with the business to see where this information adds real value at an appropriate cost can be invaluable – for both paradigms. Second, because IT professionals tend to be more process focused and tend to have more Project/Program management skills, they can often help the business determine where information and tools can add some real value, efficiencies, etc., put together cohesive and complete implementation plans, and manage their delivery. IT can and should help in identifying the key process and information differences between the models and the indicators that a product, business or service is moving or should move from one model to the other.

Leave a Reply