Is the Right Selling Price an Optimized Price?
I think every person running a business needs to regularly ask themself the following question:
Given our operating constraints, what is the best mix of products and/or services for us to produce and sell in the period, and at what prices, to generate the highest expected revenue?
Whether you are selling penny candy or you are a retainer lawyer with a huge hourly fee, the basic concepts are the same. You need understand your constraints (cost structure, liabilities, resources), your market, and your target customers to formulate a plan to attain maximize performance. Your pricing strategy is going to be a big part of whether you are successful.
Today's Landscape of Selling Online
Big Data has been really hot the past few years and it's only growing. Now more than ever, we continually initiate more actions that end up as data facts in someone's database. There is a huge urgency for companies to perform data mining/analysis to better understand the our interactions. The end goal usually is to understand our behaviors, better market something to us, and get us to buy something.
We have established the synergistic relationship between price and demand in an earlier post. The more organized retailors (Amazon, Target, Walmart) and verticals (grocery chains, airlines, car rentals, insurance) make heavy use of this data in determining the overall demand for a product, your potential demand for that product, and the appropriate price point to offer it. Ther processes often allow them to use predictive analytics to make decisions about their product mix and inventory levels.
As its name suggests, demand-based pricing is a method that uses the buyer’s demand, based on an estimate of a good’s or service’s perceived value to the buyer, as the central element in setting price. Pricing strategies are most important because they can have a disproportionate impact (positive and negative) on a company’s bottom line. Managing prices has always been an activity of keen interest, but it has become even more so over the past decade as a result of the constrained pricing environment.
Price and revenue optimization (PRO, or Optimized Pricing) is a business discipline used to effect demand-based pricing; it applies market segmentation techniques to achieve strategic objectives such as increased profitability or higher market share. PRO first came into wide use in the airline and hospitality industries in the 1980s as a way of maximizing returns from less flexible travelers (such as people on business trips) while minimizing the unsold inventory by selling incremental seats on flights or hotel room nights at discounted prices to more discretionary buyers (typically vacationers). Today, it is a well-developed part of any business strategy in the travel industry and increasingly used in others.
Methodologies for determining Product Prices
Cost-plus pricing
Cost-plus pricing is a description of several types of pricing methods used by companies that require relatively little information. In cost-plus pricing, you calculate the cost of the product and then an additional amount is included to represent the profit percentage.
The method uses the direct, indirect and fixed costs of the production of a product or service, although one of the failures of the model is that it includes all costs, whether related to the actual production, marketing or sale of the product or not. The method also requires that the firm select a desirable markup percentage, which will then be applied to costs.
One equation used to come up with cost-plus prices is: P = (AVC +FC%)* (1+ MK%), where P = price, AVC = average variable cost, FC = percentage apportionment of fixed costs, and MK% = percentage markup. Variable costs are those that vary as the output level varies, and fixed costs are costs that do not vary with the level of output. Fixed costs include costs such as property, equipment and labor. AVC or average variable cost is variable cost divided by level of output.
Although this method pretty much ignores demand and many market factors, it is extremely common by less sophisticated sellers for its ease to implement and understand.
Yield Management
Yield management is a business practice used to maximize revenue for resource-limited goods and services. Yield management is particularly suitable when selling perishable products, i.e. goods that become unsellable at a point in time (examples scenarios are seats for a baseball game, seats on a airline flight, hotel rooms, and tables at a restaurant). Because the seller has fixed costs (baseball stadium, the airline flight, hotel or restaurant has to stay open and keep staff paid for the entirety of their business hours), the more resources that get used, the better.
Yield management has one specific goal: getting customers to buy a resource that might otherwise go empty at the highest possible price.
A good example of Yield management is how movie theaters offer discounted prices for the first shows daily, as these times are less popular and they would rather make some money then have empty seats.
Another good example is how baseball tickets may be raised when they are in high demand - such as when when Yankees play the Red Sox at Boston. The Red Sox home stadium (Fenway Park) is the smallest in the majors and has sold out every home game for the past 7 years, so tickets are in extremely high demand when they play the Yankees (their biggest rival).
Dynamic Pricing
Dynamic pricing is one method of price discrimination, which is the practice of charging different prices to different consumers for similar goods. This is part of the producer's intent to capture what economists call "consumer surplus"--the difference between what a consumer is willing to pay for a good and the amount they actually have to pay. Economists refer to the price that a consumer is willing to pay as the "reservation price", and if producers could find out a way to calculate what a specific consumer's reservation price was for a good, they could charge the exact highest amount that the consumer would pay for the good before walking away, capturing all of the consumer surplus. However, as it is highly difficult for firms to judge individual consumers' reservation prices, price discrimination is more about separating consumers into groups than aiming at individual consumers.
For retailers, the motivation behind dynamic pricing is a bit different than yield management. The object of their game is to get as many goods out the door and into customer's hands as possible, while making the maximum profit.
I found a new laser printer on Amazon.com that I was set to buy (directly through Amazon, not from a seller in their marketplace). I had to take a break from the computer to feed my son (for under an hour). When I checked back to make the purchase, the price had gone up by 23%. Amazon uses dynamic pricing algorithms to measure a product's market and demand, adjusting their prices accordingly.
Time-Based Pricing
Time-based pricing refers to a type offer or contract by a provider of a service or supplier of a commodity, in which the price depends on the time when the service is provided or the commodity is delivered. The rational background of time-based pricing is expected or observed change of the supply and demand balance during time. Time-based pricing includes fixed time-of use rates for electricity and public transport, dynamic pricing reflecting current supply-demand situation or differentiated offers for delivery of a commodity depending on the date of delivery (futures contract). Most often time-based pricing refers to a specific practice of a supplier.
Time-based pricing is the standard method of pricing in the tourist industry. Higher prices are charged during the peak season, or during special-event periods. In the off-season, hotels may charge only the operating costs of the establishment, whereas investments and any profit are gained during the high season. Another example is all the sales you see around Memorial and Veterans Day at major retailers.
Identity Based Pricing
Identity-Based Pricing is offering different prices to each customer. Online commerce sites are also experimenting with identification-based pricing, which prices items based on what is known about the customer, such as their buying history and browsing behavior. This can have both good and bad implications for the shopper.
On the plus side, if a site recognizes that this is the umpteenth time the same customer has window shopped for a particular item, an algorithm may try a lower price, just for that customer, to see if that will close the sale.
Newegg and Amazon will send offers to active buyers tailored to their buying habits, often with discounted prices.
Auction Based Pricing
Some verticals have moved to the auction model for pricing, and I am not just talking about selling stuff on eBay. In this scenario, buyers are asked to submit offers (bids) and they get compared to offers of other potential buyers.
When you want purchase advertising via Google's AdWords product, you actually have to supply a minimum bid you are willing to pay for each click. Your bid gets compared to that of other advertisers wishing interested in the same keywords, with the highest getting the preferred placements.
Considerations for implementing Optimized Pricing
Yield Management, Dynamic Pricing, Time-Based Pricing, and Identity-Based Pricing are all methods for Optimized Pricing, where sellers can use data analysis and algorithms to regularly find optimal price points.
Segmentation Strategy
Another way to extend these methodologies is segment one's product line, offering multiple products and price points. Optimization can help sellers adjust prices and to allocate capacity among market segments to maximize expected revenues. This can be done based on different Key Performance Indicators:
- by goods (such as a seat on a flight or a seat at an opera production)
- by group of goods (such as all the seats to a ball game against contending rivals)
- by market (such as sales from Seattle and Minneapolis for a flight going Seattle-Minneapolis-Boston)
- overall (on all the routes an airline flies, or all the seats during an opera production season)
In this way, PRO's overall aim is to provide an optimal mix of goods at a variety of price points at different points in time or for different baskets of features. The system will try to maintain a distribution of purchases over time that is balanced as well as high.
Good PRO maximizes (or at least significantly increases) revenue production for the same number of units, by taking advantage of the forecast of high demand/low demand periods, effectively shifting demand from high demand periods to low demand periods and by charging a premium for late bookings. While PRO systems tend to generate higher revenues, the revenue streams tends to arrive later in the booking horizon as more capacity is held for late sale at premium prices.
By doing this, they have actually increased demand by selectively introducing many more price points, as they learn about and react to the diversity of interests and purchase drivers of their customers.
When to Re-Price
The key to much business success is to understand the competitive environment you are in and react accordingly. Accordingly, a good time to consider price changes is when opportunity for revenue maximization is seen/discovered.
At any time, we have 3 choices: prices can go up, prices can go down, or prices can stay the same.
- Dropping Prices - Sellers faced with lack of pricing power sometimes turn to PRO as a last resort. After a year or two using PRO, many of them are surprised to discover they have actually lowered prices for the majority of their opera seats or hotel rooms or other products. That is, they offer far higher discounts more frequently for off-peak times, while raising prices only marginally for peak times, resulting in higher revenue overall.
- Raising Prices - But if we‘re willing to lower our prices when demand is low, we should be willing to raise them when demand is high. I have seen conferences and presenters who were delighted to sell out completely months in advance. Form a revenue perspective, selling out isn‘t good news. If all the tickets sell out months in advance, it tells me the price was too low.
- No Change to Prices - Often the right strategy is to keep prices where they are. It certainly simplifies the entire sales process for sales people and for buyers, and may avoid logistical complexity.
So how do we go about capturing the value of ALL the tickets? One way is the method an extremely iterative approach. This allows tickets to be assigned to prices as they sell. The drawback? It‘s fairly labor intensive to set up and manage, and assumes a very flexible logistics and point of sales systems. Here is an example:
Build the seating projections for each season, reserving rows/sections of each price off-sale. After a period of sales reveals trends in demand, these rows are priced according to response, either as part of the lower-priced seating block or as part of the adjacent higher-priced section, depending on which price band is selling faster.
Alternatively, you can build sales curves or charts of sales for comparable performances. Taking a historical look at the speed of ticket sales in the run-up to events will generally reveal patterns. You can set up the charts to record ―T minus x‖ days (or the number of days in advance of the event) and set T to 100%. Overlay several comparable performances and you will begin to understand how your current sales compare. The advantage to this method? We have seen over-performing example events at 30% or less of the stadium sold, and been able to ratchet prices up to match the curves to capture extra revenue on most of the tickets sold. It can also revealed under-performing events long in advance of a crisis point, which is also useful.
Patterns & Observations
In the change to dynamic pricing, we often see customers treated as the people who should be the last to know. Sellers will try hide/obfuscate their prices until the time of the purchasing decision.
The practices (and results) of five typical US companies that currently use dynamic pricing are illustrated below. Their schemes have several things in common:
- These companies don‘t publish prices (other than on their websites, which are constantly updated)
- They rarely raise the lowest available price
- They report few or no patron complaints
- They have weekly reviews of sales and pricing
- They report total season revenue increases of up to 4%
AND they differ in a number of ways:
- Some companies apply small price changes broadly across many performances, while others impose quite large changes infrequently
- Some organisations spend just an hour or two a week reviewing pricing, while others have staff managing prices nearly full-time
- Some adjust prices for specific seats or sections, others for all unsold seats.
We see PRO clearly dominates static pricing when consumers do not behave strategically. This fact means that the presence of naive buyers in our marketplace will only increase our revenues when we price dynamically. So what can take from this analysis?
- When your supply is higher than the demand, post one set of prices and stick with them or advertise your price changes from the outset. With the right prices to begin with, you’ll maximize revenue.
- When demand is higher than supply, dynamic pricing will maximize your revenue.
- The presence of non-strategic buyers will increase sales at higher prices. You may find that over time, some patrons wise up and act early to save money. This will, however, not stop others from paying more.
Example for Pricing Optimization
A good example to further analyze is tickets for a sporting event on a specific date. Revenues from the event are based on the finite number of tickets that can be sold.
For years, ticket prices were static, only varying by seat location. They were a set price set months in advance. Often a small discount is offered if tickets are bought in quantity or well in advance.
This looks like a classic opportunity for yield management, An obvious option would be for organizers (sports teams) to look at the supply and demand for tickets, reacting accordingly. Imposing price increases when a set percentage of the tickets are sold can result in some extra revenue. If there is genuine excess demand (economist-speak for buyer enthusiasm) for a product or event, you‘ve let 80% of your tickets go at a too low a price. You could have sold all (or most) of those tickets at the eventual higher price.
An article from ESPN last year shows that some Major League baseball teams are in fact using PRO for setting ticket prices.
Proceed with Care
In the ESPN article, the author writes the following excerpt:
But the other reality is that a team not only wants to make as much as it can on that game, but it wants to foster a long-term relationship with its fans.
Dynamic pricing can’t all be one way -- up.
Prices need to be adjusted down, too, offering bargains to fans.
“Teams are very much aware that the fans are driving the team,” Eglen says. “They don’t want to engage in gouging every last dime they can get out of the fans. They want a long-term positive relationship.”
- If you consistently discount often, custiomers will become conditioned to expect these discounts.
- It's important to CONSIDER THE LONG TERM RELATIONSHIP.... Don't gouge and piss off your customers.
Some More Considerations
There are quite a few factors that may not be obvious, but can definitely influence pricing strategies:
- There is cost of holding inventory.
If you are selling a physical product, you likely need to consider the logistics associated with storing the inventory and delivering the item. If it is a seasonal item, does it make more sense to heavily discount than to store the item for an entire year. You'll often holiday items go on sale after the holiday passes and snow shovels discounted after the winter.
- Accounting methods can influence pricing.
The method a particular uses to value inventory can drive price. If you got a great deal on buying additional inventory of a specific product, companies using First-In-First-Out (FIFO) will likely recognize their cost structure differently than companies using Last-In-First-Out (LIFO). - Product Segmentation can cannibalize other segments or selling opportunities.
Companies need to make a strong case for customers to buy something other than the lowest priced offering. The value add must be clearly apparent and actionable.
Conclusions
We live in an information economy, where information can be the most valuable asset your company has in the quest revenue and efficiency maximization, With that said, optimized pricing really makes sense for a seller in many ways. It really amazes me that the vast majority of sellers in this country do have the knowledge/resources to take advantage of PRO and take the Cost-Based Pricing approach.
A new kind of Marketplace fully based on PRO
The popularity of eBay auctions have been a really amazing phenomenom to watch. The bidding process naturally pushes for a dynamically determined price by the auctions closing time. What would be amazing to see is dynamic environment of eBay, but without a set end date. It would make shopping for commodity goods much like investing in stocks and trying to gauge financial markets.
Earlier I spoke about the way the printer I was interested in buying from Amazon changed prices in a short period of time. While some items in storefronts (generally considered fixed price marketplaces) may sometimes shift, they are largely consistent. Imagine if a marketplace (the size of Amazon or eBay) continually shifted with a PRO approach, it would be an incredibly interesting and fun to shop as you would only be able to speculate whether the price at that point in time was a good value.
There are many examples where applying gaming strategies to various consumer experiences has worked to improve user interest and spending. I think gamifying the shopping experience based on PRO would be really cool and could be a successful business differentiator if executed correctly.