Markdown management in practice

It is interesting to observe how markdowns are managed in practice. Here is an example of two blog posts documenting the dynamics of the optimal markdown discovery. The first uses data from several retailers and argues that markdowns should be reduced or eliminated citing the negative impact of markdowns on revenue and recovery of the economy. The second presents the case of Macy’s for larger and more straightforward markdowns as opposed to the convoluted coupon discounts. So, are markdowns good or bad?

Academic research on the topic had a similar evolution of ideas. Early stream of papers recommended that markdowns should be large, 50%, or sometimes even more. Then, researchers realized that deep markdowns pull consumers into the strategic waiting game, where consumers wait and gamble to get a product at a low price. The result of this realization was in justification of everyday low prices, or very small markdowns. The current state of the debate, including the results from one of my papers, is that some markdowns are good. In fact, there is a sweet spot for markdowns – they certainly should not be as large as 50% also not as small as 10% either. There is a benefit of offering a reasonable (~20%) markdown and bringing in the value-oriented shopper, while keeping those ready to buy at the full price buying. In that sense, Macy’s seems to be doing the right thing.

Advertisements

Always low prices. Always?

walmartIt is a beginning of teaching semester and this blog inevitably gets more active. One article caught my attention recently. It is related to my research, and as one can probably guess from the title, it’s about retail pricing. Particularly, Walmart and its everyday-low-prices (EDLP) policy. It turns out that Walmart’s (at least online) everyday-low-price policy becomes “everyday-adjusted-price” policy.

The article suggests that the reason for the Walmart’s pricing policy change is competition from Amazon.com. Citing the e-commerce data analytics firm 360pi, it reports that on 15% of the products, Walmart changed prices daily (very close to that percentage at Amazon.com). Furthermore, Walmart prices closely track those at Amazon.com generally staying within 5%.

Competition with Amazon.com may be one of the reason, but to me it seems it is not the only nor the main one. Retailers are in the business of converting goods into revenue, and dynamic pricing simply generates more revenue. The revenue lift is of course conditional on the fact that consumers keep buying at regular (high prices) and not just patiently waiting for bargains. This is where understanding how consumers decide whether to wait or buy becomes important.

Two recent research papers speak directly to this point. One, by my colleagues, shows that human behavior provides rationale for markdown or dynamic pricing over EDLP pricing.  The other, by me and co-authors, shows that markdowns can be set even larger than the current methods prescribe, leading to substantial revenue gains. From that perspective, Walmart is doing exactly the right thing by adopting the dynamic price policy.

Is Amazon.com cashing on consumer behavior?

I am very much tempted to answer – YES. The genesis of this post is the recent report by NPR’s Stacey Vanek Smith; the case under consideration is Amazon.com potentially raising the cost of its Prime membership from $79 to $99 per year or more.

Prime membership gives consumers free shipping on their orders and free access to numerous books, movies, and TV shows. The volume of Prime subscribers is estimated to be well over 10 million and growing rapidly. Could it be that Amazon.com has decided to curtail the growth of membership? This could be sensible if Amazon has reached capacity limitation for shipping or content streaming. Neither seems likely, though. So what is the logic behind the (possible) decision?

The report offers this explanation:

But the rationale for raising prices, may not be fast cash, speculates Michael Levin, co-founder of Consumer Intelligence Research Partners in Chicago.

“At Amazon, nothing is ever what it seems,” he laughs. “If they charge more, I think customers are probably going to spend more. So quite ironically, by raising the price of this membership, they may end up getting people to shop there even more.”

A more expensive Prime membership equals a customer who is all the more motivated to get his or her free-shipping’s worth.

Having done some research on the topic of consumer behavior, my explanation for this phenomenon is the so called Sunk Cost effect. This is a well-known phenomenon in Behavioral Economics. In simple terms, the effect occurs when consumers continue to use products that have become obsolete. Moreover, the usage increases if consumers spent more money to purchase these products. As such, this behavior is irrational in the classical economic rationality sense: a decision to use a product should depend only on the current or future cost and benefit, but not on a past one.

So by raising prices of the Prime membership, Amazon.com could be just targeting our irrational tendency to recover sunk costs. Even if there is an identical product couple of dollars cheaper, Prime members would still buy from Amazon, driven by the (sunk) cost of membership.

How to profit from no-show train passengers?

According to this NYT article, Amtrak is rolling out the system where conductors will be scanning passengers’ tickets with iPhones.

By late summer, 1,700 conductors will be using the devices on Amtrak trains across the country, the company said.

With the new system, passengers will be able to print tickets or load a special bar code on their smartphone screens for conductors to scan, and conductors will be able to keep track of passengers on board, Amtrak said.

A digitized check-in process for trains seems long overdue in a world of online concert tickets and flight reservations. But the industry faces a particular challenge in that passengers hop on and off at different platforms at different times, unlike at an airport, where people check in at one gateway to board a flight, and then stay there until the flight arrives.

Both airlines and railways profit from no-shows if they overbook. However, for a railway the above mentioned challenge of people hopping on and off at different stations might actually be a blessing in disguise, especially if one can track no shows in real time. That’s exactly what this new technology is doing.

Imagine a train departing from Boston to DC. Continue reading

On the power of randomization: Selling diapers to pregnant women

I posted before on how Priceline uses randomization to conceal the minimum price they would be willing to accept in the Name your own price channel. Here is another example on putting randomness to work. This one is from retail.

Whether we like it or not, retailers do accumulate a wealth of data about our purchasing patterns. Every time you use a credit card, coupon, or a loyalty card, transaction data is logged and stored. The data is obviously used to send you more coupons, ads and peddle new products. Now, how much can retailers learn from this data? It turns our quite a lot. As this article from the forthcoming NY Times magazine describes, Target can actually predict whether a woman is pregnant just by analyzing change in her shopping patterns.

Pregnancy is a sensitive matter, however, so when a week after shopping a bunch of coupons for maternity clothing and baby products arrives in mail, women can get upset. It also looks awful lot like spying – does not it? Lesson #1 – one has to be very careful with this kind of data. Lesson #2 – even if data suggests something, do not necessarily pursue the opportunity at full speed. Here is what Target does: they randomize. Put a coupon for wine glasses next to diapers and office furniture next to pacifiers. Combinations seem to be quite ironic. But it does seem to work – the same article reports substantial growth in maternity and baby product sales for Target after they started doing this.

Selling ‘opaque’ product

Since I mentioned Priceline in my previous post, it seems like the right place to resurrect one of old posts on the subject. Everybody knows how Priceline works: they sell you a product with certain characteristics, without revealing it at the time of purchase. So the product is opaque. In exchange they let you bid for the product and if the bid is high enough, it’s accepted and product is sold. Or is it?

Continue reading