Dynamic Pricing Ain’t All That

Dynamic pricing has come up quite a bit lately, driven in part by interest into whether the practice was used extensively during the holiday season, and a continued interest in whether it will be a big deal in 2016.

My take: at least half the hype around dynamic pricing comes from a misplaced definition of what dynamic pricing actually is.

I didn’t realize this was an issue to such a degree until Brian and I met with a client right after NRF. A pricing solution vendor, they too were concerned about the actual definition of dynamic pricing, and in fact have been exploring terms like “high frequency pricing” as a way to split off what is a true fad from something of longer term value.

Let me explain what I mean. The literal definition of “dynamic”, according to Google, is “characterized by constant change, activity, or progress.” In that sense, what Amazon does on its site – change prices multiple times during the day (although it’s important to keep in mind that the number of items that receive this treatment is still pretty small) – is, technically “dynamic” pricing.

What is required on the technology side in order to enable this kind of dynamic pricing? Ultimately, nothing. You could literally have someone sitting at their computer, manually changing a price, and then hitting “update”. If you’re changing prices 8-9 times per day per item, it wouldn’t really even have to be a full time job, especially if you made it possible to bulk upload and execute multiple price changes. Your only limitation is how long it takes the site to refresh to reflect the new prices.

You could potentially get a little more tech-enabled, and actually use some kind of optimization to make intra-day price change recommendations. But it would be based on aggregated demand, and (I imagine) something reactionary to competitive pricing in the market.

Ideally, you’d want situations where competitive price intelligence detects that a competitor has made a significant price change, and also that your conversion rate on the product in question is now also taking an unexpected demand hit. Or, conversely, you’d want your competitive intelligence to let you know that competitors are stocking out (but you aren’t) and you have an opportunity to raise your price and capture more margin because you have availability and they don’t.

You definitely don’t want to react to a situation where a competitor drops a price, but your own demand is unaffected. If you’re still selling as expected, there should be no need to react to the competitor – that’s just giving margin away.

The challenge with all of this is, what if you’re a store-based retailer? You have all of this activity going on online and no way to bring it into stores, not without confusing the heck out of consumers – or angering them. In fact, I wrote about it three years ago, and it’s still a very popular article on our site. I’ve heard two scenarios around bringing dynamic pricing into stores, and while one of them kind of makes sense, it’s such a niche application I can’t see it making it in the wider world.

The niche application is end-of-day pricing of perishable goods. You know, automatically discounting the tuna salad at the end of the day to make sure it all gets sold instead of thrown out. This kind of application works for two reasons. One, it definitely has a value proposition for both retailer and consumer (though, really, I have no idea how substantial a business case actually exists) – the tuna salad gets automatically marked down based on how much is left and how quickly it needs to go. And two, the sale typically involves a scale and a dynamically printed price tag, so even if you printed the tag and the price changes between when the consumer picks up their order and when they get to the front to pay for it, you can still charge them the price they expected – the actual, calculated price on their printed label.

The other scenario is more disturbing. Imagine the toy aisle of your local superstore, all decked out with electronic shelf labels. A shopper picks up a Barbie for $19.95 and puts it in her cart. That Barbie is tagged with an RFID label, which means that specific item has a specific unique identifier. Readers in the store register that the item is no longer on the shelf and, in fact, moving, and thus “fixes” the price for that item at $19.95 until it either goes through POS at the front of the store, or gets placed back on the shelf. That way, the reasoning goes, when the shopper gets to the register, she’s not surprised by the price going up or down, when she had already committed to buying the product at $19.95. That just seems insidious to me, and I think shoppers are sophisticated enough that they’ll figure out what’s going on soon enough – and the retailer will be punished for such behavior in social media spaces. A PR fiasco.

These scenarios are where dynamic pricing leads us to today. It’s not very smart, it’s only valuable if it is used carefully and in concert with competitive price AND inventory data, and it’s undifferentiated. If someone was willing to pay a higher price, they still get to benefit from the lower one.

This is a far cry from “personalized” prices, which is what some in the industry mean when they use the term “dynamic” pricing. Make no mistake: dynamic pricing the way it is used today is not personalized by any stretch of the term.

Personalized pricing has its own pitfalls, and retailers who employ it also need to be just as careful they don’t wade into PR fiasco territory. But, especially when it’s backed by clear rationales that are made at least somewhat visible to consumers, it’s a much more comprehensive, and ultimately more valuable strategy than dynamic pricing. It enables retailers to have a consistent price across all channels – “the” price – while simultaneously (dare I say dynamically?) offering different prices to different customers.

Look at how Jet.com does it today. You add items to your cart, and depending on what you add, each item gets repriced right in front of you. The more you add, the more you save. But you might save more by adding certain items over others – it depends on the deals Jet has with suppliers, and the supply chain efficiencies of the items you’ve selected. There is “the” price, and there is “your” price. Your price may depend on supply chain efficiencies, like with Jet, or it may depend on your behavior, whether that is localized behavior, like adding something to your cart and then not buying it, or behavior over time, like offering discounts to high value shoppers to expand the categories they shop.

As long as retailers can justify how those offers are made, so as to avoid accusations of bias (pricing by demographic would get you into a lot of trouble fast), then retailers can manage variable prices by explaining that there is “the” price, and there is “your” price, and this is how your price was calculated. This strategy, by the way, also makes competitive price tracking of your own prices more challenging. Because the list price will not necessarily be the price your competitors have to compete against, but it will be the one their CI systems pick up.

So – dynamic pricing, as it exists today, as it is defined today, ain’t all it’s cracked up to be. I think personalized pricing is the future, and that requires a lot more capability than simply changing prices many times over the course of the day. That other meaning, the one in play today, is just a fad, and will ultimately be out-gamed, whether by competitors, or by consumers themselves.

‘CamelCamelCamel’, anyone?