To On Demand or Not To On demand - apps4work/co.a4w GitHub Wiki
That is the question. Whether there is an On-Demand sector in any particular product category is a function the nature of the demand, and how the factors of demand play into costs and value, and how technology can change the costs.
The critical factors of the demand are
- Variations in the product
- Predictability
- Immediacy
Immediacy can be a show-stopper for On-Demand. The requirement for it is a function of product and the consumer. No matter how many variations there are in snake venom antidote, nor how unpredictable the demand, when you need it, you need it right now.
But for some products "immediacy" is malleable and is changing. On-line commerce has changed "immediacy" from "I can pick up right now in the store" to "immediate" is "overnight", and then, as Amazon competes against Bricks and Mortar, to "same day", and to "two hours". For those who find shopping trips a chore to be procrastinated against as long as possible (ie. me) , and live just the right distance from population centers (ie. me), buying from Amazon is usually more immediate that buying from a store.
But I still have a hankering for a hands-on experience of some products prior to the purchase decision. I want to know if that tool feels as strong in my hand as it look in the on-line picture. For a long time I regarded Borders as a place to sit in comfortable chairs, drinking good coffee, sampling books; better coffee than the library and a wider choice of books than Starbucks. And, while sitting in Border's comfy chair, I might order my chosen book from Amazon Prime to save me carrying it home. I'm not surprised that Borders is no more.
Bricks and Mortar store are going to have to figure out the value that the customer goes there for - and how to appropriate a share of it. Lowes, Home Depot and Walmart are trying hard to blend the immediacy of the store with the product range of on-line. On-line is not necessarily On-Demand, but it lays the path. Once the immediacy requirement is worked around, the issue is only cost of on-the-(distribution center)-shelf vs On-Demand.
And, of course, the immediacy requirement isn't satisfied if what you want is not on the 'shelf'. And I willing to wait 2 hours for Amazon to deliver the right antidote, or overnight for the dress in my size? Do I have a choice?
On-Demand is favored when there are a lot of variations, when demand for any particular choice is unpredictable, provided there is no over-whelming requirement for immediacy.
Long run, "on the shelf", always delivers better immediacy - it always takes longer to manufacture the product and hand it to the customer, than it does to just hand it to the customer. Long run is favored by few variations each with predictable demand. There is always a economy-of-scale cost benefit to long run. Predictable demand reduces the cost of over- or under- stocking.
As the number of variations increases, the predictability goes down and the economies of scale diminish. The habits of the reading public are sufficiently stable that the year-to-year number of sales of copies of books is relatively predictable, but you can't do a long run of "books". You have to do a long run of a specific book, and forecasting the sales of a specific book is a lot harder. There is no customer value from correctly predicting 90% of the ISBN of the book that the customer. It is either right or wrong. You either have the book she wants, or you don't.
Variation without customer value is a self-inflicted wound. If the customer values the paperback and hardback equally - and will buy whichever is available at the same price - the variation is pointless. But she doesn't. A segment of the market is willing to pay a significant premium for the hardback over the small incremental cost of manufacturing it, and both hardback and paperback have high margins over costs.
The gross margin over costs has a significant impact on the cost of "on-the-shelf" in terms of under or over stocking. You may have noticed that French bread stores always run out of bread. Bread is a low-margin product; the cost of discarding a single loaf of overstock wipes out the profit on a lot of loaves that were sold. French bread, in particular, because the French regard bread that's two hours old to be stale. US Supermarkets don't sell out of Wonderbread because a large segment of US consumers don't appreciate real bread, but supermarkets do sell out of 'fresh baked bread'.
Garment stores overstock because of the high margin over costs. Publishers are willing to take 50% returns for the same reason.
So, for example, Starbucks has ready-made ("long run", "on-the-shelf") coffee-of-the-day. It takes the barista less time to brew a pot of coffee and serve one cup at a time than to make each cup to order. It's reasonably predictable that they will sell coffee-of-the-day and it sells at a high margin over cost so throwing away some isn't that big a deal. Whether or not the customer values the extra immediacy is debatable, as is whether they will tolerate the occasional extra wait for the coffee pot to brew. But the huge variations of coffee and other beverages available - that are not just coffee-of-the-day - and all made-to-order because their predictability is close to zero.
The housing market shows a similar stratification. Starter homes are typically built on spec ("on the shelf") in developments of very similar, if, no longer identical, houses ("long run") using specialised crews gaining economies of scale in crew-overhead time.
The starter home market is under price pressure - lower margins. Developers give in only slightly to customer pressure for uniqueness (or at least absence of obvious 'sameness') with slight variations on a theme in the products. A paint-color or a finish there, or a choice of one of three floor plans.
Trophy houses are ultra-customized and always never built "on spec". And between the extremes, developers vie for the most profitable combination of long-run/on-the-shelf - economies of scale, immediacy - against customizing and uniqueness. The Trade-offs vary with the state of the housing market because of perceived variations in predictability.
On-Demand Manufacturing
On-Demand is a technology driven opportunity. The technology is the control over the manufacturing process by computers, which enables the process to produce individual unique products for about the same unit cost as it can make a Long Run of identical products, although typically that's a higher unit cost than equipment that manufacturers only identical products.
The Trophy House market and Saville Row Tailoring are on-demand opportunities that do not have computer control over the manufacturing process. Computer control flattens the cost curve for variations and uniqueness, which allows unique products to be sold at lower prices, but, necessarily, only where the uniques is valued by the customer sufficiently to cover the extra costs.
"On-demand Manufacturing" is the technical requirement for an On-Demand business. If you can't manufacture it from a "Digital Product" specification, you can't do any of the clever stuff of customization etc. - you can't do "an on-demand business"
So for us to do an On-demand business for a product category, it has to be possible, or at least plausible at this stage, that we can do "on demand manufacturing" of it.
Manufacturing is "assembly". where some parts get "manufactured" from raw material and others part are just manufactured by someone else, and all these parts are assembled into the final product. for example, Garment Manufacturing is "assembly". cloth parts get "manufactured" by cutting rolls of fabric, other parts (like buttons, thread, closures) just get bought, and we assemble all the parts (sewing them together) , into a product.
It would be dumb for me to claim that I know what "parts" we can make from raw materials, or we can "assemble". But that's what constrains the answer to the question, and it defines it more than the "product category".
So if we can manufacture plastic parts, and assemble plastic parts, we can go into an On Demand Business on "anything that's made from plastic" . We'd a 20x20 and it would be "certain kinds of plastic parts" and "certain ways of assembling plastic parts".
And as long as the digital product impies actions that are in that 20x20 we can do it with any product category -- hair combs and kids toys and spectacle cases. etc.
or If we can manufacture leather parts (ie cut leather) and assemble leather (ie sew leather) we can do anything made of leather. Handbags and gloves and luggage and and horse tack. [ all subject to its 20x20 ]
or if we can manufacture metal parts (cut metal) and assemble metal (weld), we can make product made of metal.
or PCBs and electronics components
and so on.
(and if we can do several of these we can make several of th product categories - things made out of cloth and leather and metal and electronic parts).