jump to navigation

The costs of the “Free Shipping” option at Amazon November 29, 2007

Posted by The Armchair Economist in Business, Economics, shopping, Technology.
Tags: , , , ,
12 comments

This morning, I caved and bought the Nikon D40 D-SLR that I’ve been fantasizing about on and off for 6 months. In any case, I ordered the camera with an extra 55-200mm VR zoom lens in addition to several other accessories (which came to almost free taking advantage of various promotions). Being the cheapass that I am, of course I opted for the ‘Free Shipping’ option.

I wouldn’t call myself a frequent buyer from Amazon, for one, I feel that their prices aren’t all that much better than elsewhere (especially on textbooks and review books, where there is often a 0% discount, go figure.. they have less overhead but charge full price?). But on occasion, a deal crops up where I’ll bite the bullet and buy from them, always using the free shipping option. I noticed that up until maybe 2 or 3 years ago, anytime I used the free shipping option, it would ship within a few hours. However, within the past 3 years, my “free shipping” items would often be delayed a few days before it shipped. This coincided with the “Amazon Prime” offering, where you could get free 2nd day air if you pay them 79$/year. I figured the delayed shipping on the “Free” option was just their strategy to artificially inflate the ‘value’ of Amazon Prime. This time is probably a record, the ship date is 5 days AFTER i placed the order.. and all of my items are in stock and are available for 1 day shipping. This got me curious enough to do a quick google search and I info (posted below)… apparently its from an insider and provides some details into the free shipping/ estimated date of shipment algorithms..

I used to work at Amazon, in fact I was responsible for a lot of the algorithms used in Amazon’s logistics network in use today. I’ve read a lot of FUD about Amazon order fulfillment so I have to try to add some facts to these discussions.

Amazon’s order fulfillment process is insanely complex and involves hundreds of different variables which impact customer satisfaction and costs. Since keeping costs (and thus prices) as low as possible is also a major factor of customer satisfaction, it’s important to understand that the underlying principle for all of the algorithms used is to maximize the customer experience. But that’s very complicated. Without disclosing any trade secrets, here are a few of the types of parameters which go into the order fulfillment decisions:

* What has the customer paid for? If they’ve paid for next-day delivery, Amazon assumes they want their order as quickly as possible. Likewise, if they’ve chosen super-saver free shipping, they assume it’s not needed that quickly.

* Where is the inventory? Amazon has many distribution centers scattered across the U.S.A. and not all items are available at each center. It obviously costs less to quickly get an item to a customer from a nearby warehouse. But it might be necessary to ship from further away.

* Are inventory levels as low as possible? To maximize cash flow and utilization of its warehouse space, Amazon is very aggressive about managing its inventory levels. This may mean that rarely ordered items are not kept in inventory and may require time to source from a supplier. It also means that Amazon tries to move inventory out to customers as quickly as possible. Usually, delaying shipments has inventory costs. Believe me, Amazon has no desire to own inventory one minute more than it has to.

* Where is there space for inventory? Related to the above point, space can get very very tight in the Amazon warehouses, especially before the holidays. They just may not have room for dozens of 60″ plasma TVs in every distribution center.

* Does the item need to be ordered from a distributor or manufacturer? If an item is not in Amazon’s inventory it may already be on order and due on a future date supplied by the source or it may need to be ordered. Obviously it’s cheaper to buy large volumes directly from a manufacturer, but those items may not be available quickly. If a customer has paid for next-day delivery though, Amazon will pay to get that item to them.

* Which distribution center has enough people working today? To help keep employees happy, Amazon tries to smooth work loads. Employees prefer to know when and how many hours they will be working next week. When there are high volumes of orders, this may mean delaying some low-priority orders.

* Which distribution center has capacity in their automated systems? It’s much cheaper to fulfill an order using Amazon’s automated facilities than to do it using manual labor. But only so many orders can go through these automated systems a day. Orders may be routed to a different fulfillment center, or delayed to minimize the handling costs.

* Are there any automated fulfillment lines? For big ticket items (think new Harry Potter book) Amazon will often dedicate mechanized packaging and shipping hardware to just that product. It may be much cheaper to wait a couple days until one of these mechanized production lines is in place before fulfilling an order.

* Which shipping company is cheapest? Amazon utilizes just about every common carrier in the U.S. and can even use them together to minimize shipping costs and reduce delivery times. These rates change regularly and service levels can vary from day to day. It may cost 50% less to get a non-next day order to a customer by waiting a day or two before shipping it.

* Do the shipping companies have capacity? Many of Amazon’s distribution centers are in low-cost rural areas. It’s not uncommon for Amazon to max-out the capacity of UPS, DHL, USPS, FedEx, or other shipping companies from one of these centers. If that happens, an order for New York City may end up coming from Nevada rather than near-by Pennsylvania.

* How quickly can Amazon be paid for an order? Legally, Amazon can’t charge you for an order until that order has shipped. To maximize cash flow, Amazon normally tries to get orders out as quickly as possible so they can be paid as quickly as possible.

* How good of a customer is this? And yes, if you’re a good customer, Amazon will spend more to keep you happy. In extreme cases, I’ve seen Amazon send employees to local Walmarts or other retailers to buy an item which is then couriered to a great customer. Believe me, this isn’t cheap. This cost and effort doesn’t go into a first-time buyer’s Super Saver order.

These are just a handful of the types of decisions which have to be made for every order. During peak periods Amazon handles over a million orders a day; trying to balance and optimize all of these constantly changing variables is a difficult job but I believe the company does a pretty good job of it. And yes, I believe there is intellectual property in the software created by Amazon to make this happen.

In general, I think Amazon’s customer service is great. But when I hear that some individual customer service rep, who is probably sitting in West Virginia or India and has never seen the algorithms used for fulfillment or even been to one of Amazon’s fulfillment centers explains how free shipping works, I just have to giggle. They really have no idea.

digg story

Fixing health care in the US? Part I November 16, 2007

Posted by The Armchair Economist in Business, Commentary, Economics, Fix Health Care, Health, Medicine.
3 comments

Huge task… I know. I realize that I always come up with ideas to ponder but I’m discouraged in pencilling it down because it means taking hours away from my day to flush out my ideas that will likely be a unintelligible tome. Rather, I will post short segments/thoughts (open to discussion if you feel like commenting) of problems I’ve identified.. as well as possible solutions. Search under the category ‘fix health care’ for future installments.

Problem: The oft cited ‘America spends XX% of their GDP and health care still sucks’
The reason for our high costs are obviously multifaceted…

  • We have the newest technologies, newest procedures, newest pharmaceuticals.. All of this costs money. If I was deathly ill, I definitely want to be in the US with all of its resources at its disposal to cure my illness, over any other country.
  • We subsidize the rest of the world’s health care. The cost of most new technologies are not regulated in the US (ie: manufacturers can charge whatever they want) while prices are heavily regulated by all other nations (ie: thats why drug prices are so much cheaper in socialized systems than in the US), otherwise these nationalized systems would be bankrupt. (likewise, allowing competitive pricing in other countries would likely drive down the prices in the US)
  • Preventative health care is not a priority in the US: Since there are so many insurance companies out there (all focusing on the bottom line) there is no guarantee that the expenditures they lay out today (for a disease you do not have and may not get), would benefit them in the future, since you can always go to another insurance company thus negating their investment in your health. Their solution? Reduce emphasis on ‘low return’ preventative health care, and try to ‘manage risks’ (ie: minimize sick patients on their rosters.. even to the point of canceling policies in the middle of expensive chemotherapy as elucidated by Health Net recently)

Yet for all of our costs, our health doesn’t seem to be any better..

  • Our health isn’t a failure of our health care system, it is a failure of priority and accountability. With a cultural priority on wealth accumulation, work, and leisure time becomes a valuable commodity. Increase in middle class and disposable income means more money to spend on dining out and convenience foods. A cultural focus on higher education is migrating our workforce to a more sedentary service oriented economy. Health care is a means for us to enjoy those steak dinners and burgers yet maintain a ripe old age. (ie: 1 surgery to make up for 30 years of culinary indiscretion, hit me)
  • Our infant mortality rate sucks (#17 according to Michael Moore’s Sicko) Two key reasons that come to mind: the definition of ‘infant’ is different. In the US, infant mortality counts an infant exhibiting any signs of life, regardless of gestational age or size. (ie: high risk, premature babies <28 weeks or less than 1000gr would not be counted as infants in some countries). Additionally, the technology allows us to birth high risk babies that would likely not be carried to term elsewhere. Basically this is a numbers game.. the question is, if you were going to have a high risk baby, what country would you want to be in?
  • Lies, damn lies, and statistics: Statistics can be massaged to support any statement. Lets look at objective statistics.. for exampling by match risk factors (ie: obesity) with outcomes by country.. (I’m not suggesting that US would come out on top)

Solution:

  • Use our political and economic leverage to allow less restricted pricing by socialized nations This will obviously face substantial resistance. Arguably, we are suggesting to raise the health care costs of other nations so that we won’t have to pay as much (although rightfully, we have been subsidizing R&D for the rest of the world)… probably not the most popular proposal.
  • Increase the accountability of health to the individual: With the exception of genetics, your health is your decision. Your decisions can influence your health in 50 years. Each cigarette you smoke increases your risk for emphysema, lung cancer, coronary artery disease, etc.. why should your decision affect my financial wellbeing (in terms of higher insurance premiums.. to take care of your sick ass in 30 years). Each bag of fries, canister of pringles, slab of filet mignon, and can of soda will increase your chance of obesity, diabetes, hyperlipidemia. How can we influence peoples decisions and make them financially accountable for their decisions? Consumption tax… or Fat and Sugar tax if you will. Taxes collected can be used to subsidize healthcare costs as well as fresh fruits and vegetables (not necessarily to pay for farm subsidies though).
  • Prohibit health insurance companies from being for profit entities: There is an inherent conflict of interest in making money in the insurance industry. While the industry argues that it increases competition and reduces costs to the consumers, all it does is reduce corporate accountability. Ideally insurance works by pooling risk, in a profit motivated world insurance works by reducing costs. By becoming mutual companies where profits are distributed to the policy holders, there is a form of checks and balance.. money saved goes back to people who payed the premiums, and any policy changes that might reduce benefits affects only policy holders.
    • Shift the responsibility of verifying information accuracy on applications to the insurance companies: In response to the Health Net fiasco, insurance companies shall be prevented from changing coverage once they begin accepting premiums. Insurance companies argue that this will foster fraud (ie: people will underreport existing conditions).. I’m not sure why no one has thought of this before, but rather than sitting back and collecting premiums, insurance companies should be responsible for doing due diligence on the health of a prospective client BEFORE they begin accepting premiums rather than while they are undergoing expensive and lifesaving therapy. (I would classify this responsibility as ‘risk management’) This is not as much of a hardship on insurance companies as you would think, insurance companies can protect themselves by easily sharing patient information with each other to verify the accuracy of applications.


Thats all I have for now. I reserve the right to continuing updating this entry for additional points, accuracy, and hopefully citations. I welcome your thoughts to fixing the system.

digg story

Making Recycling Work November 15, 2007

Posted by The Armchair Economist in Business, Consumerism, Economics, Environment.
add a comment

I’m probably not the only one to wonder where the bins of plastic bottles and newspapers I painstakingly separate out are indeed recycled. Here’s an interesting article from The Economist that elaborates on the inner workings of how recycling works.

This article got me thinking.. how the hell do Americans generate almost 800kg’s of municipal waste a year? One reason that comes to mind is the underhanded practice of raising prices by selling a decreased amount of product in similar packaging. In one example, after years of stagnant sales, coke tried (unsuccessfully) to raise their prices, but came to realize that 0.99 was a mental barrier to their 2L bottles of soda, thus reduced their bottle size to 1.5L, retaining the same pricing point). Another recent example is soap manufacturers cutting ‘grooves’ into their bars to ‘make it easier to handle’, but at the same time you have ~7-8% less soap per bar, while the size of the packaging remains the same. Perhaps the US should adopt a rule from the Germans, ie: shifting the responsibility for the entire life cycle of packaging to producers… but oh lord.. that means corporate responsibility!! god no.

Further evidence that Health Insurance companies should not be private, for profit entities November 12, 2007

Posted by The Armchair Economist in Business, Ethics, Health Care, Medicine.
add a comment

Woodland Hills-based Health Net Inc. avoided paying $35.5 million in medical expenses by rescinding about 1,600 policies between 2000 and 2006. During that period, it paid its senior analyst in charge of cancellations more than $20,000 in bonuses based in part on her meeting or exceeding annual targets for revoking policies, documents disclosed Thursday showed.

The simple view is that corporations exist to make money for their shareholders (ie: their fiduciary duty), insurance companies make money by paying out less money than they collect in premiums, thus it would make sense that companies would try to reduce their risk by eliminating clients that need paying out… which is contrary to the whole idea of insurance. All the more galling is that these companies can collect premiums for years, but when they are called to pay out claims (ie: their customer gets sick), only then do they do a careful review of the customers application to see if the customer is ‘eligible for care’ (shouldnt this be something they figured out before they started collecting premiums?).

The question is how can we fix this conflict of interest? Perhaps requiring insurance companies to be mutual companies rather than for profit entities, so that all balances (if there are any) at the end of the year are returned to policy holders (ie: there is no ‘for profit’ motive, and money are returned to the people who paid out the premiums in the first place). Another idea might be to place the burden of ensuring accuracy of medical information on application the insurance companies responsibility, thus once they accept premiums from a customer, they are on the hook for their care as long as the patient continues to pay their premium (ie: they cannot walk away from a contract). Additionally, even IF the patient neglected to mention a preexisting condition, claim payout denials should only be restricted to those relating to the omitted condition (ie: the article discusses a woman whose insurance company who denied coverage for breast cancer surgery and chemotherapy claiming she omitted having a heart condition). Or perhaps, we can do without health insurance companies all together, or have one quasi-governmental, non profit insuring agency.. having all policies under one roof would aid in information sharing (ie: no need to verify ‘preexisting conditions’), reduce overhead (ie: $$ spent on marketing), and likely improve medical care (easy access by medical professionals to the patients past medical care).

Any thoughts on whether these solutions would work? How other solutions to this problem?

read more

Follow

Get every new post delivered to your Inbox.