I went to get my car washed today in freezing weather, the day of a massive snow storm about to hit DC. Needless to say, nobody else was there. (Why did I do this? Because the car desperately needed to be waxed + interior cleaned, and I'm not in DC for long). The experience got me thinking about dynamic pricing and customer loyalty.
Businesses typically try to use frequent-purchase tactics to drive loyalty, like a "buy 9 get 1 free" card or, in the case of airlines, frequent flyer miles. But I believe there's a better way to drive deep loyalty while at the same time maximizing the revenue a business gets: Dynamic pricing, with a Hedge. Here's what I mean:
As I mention in the video above, to say it was a slow day at the car wash facility would be putting it nicely -- I must've been one of only a couple dozen customers they would have the entire day. It's expensive to keep a carwash open on a day like today, including paying at least 10 employees to sit around and do nothing.
Airlines suffer from the same issue: The plane needs to leave the gate on schedule whether it's half full or entirely full. Those empty seats are all lost revenue.
It's the same with most businesses: When they aren't running at their max or optimal capacity, the business is losing out on incremental revenue that would have a very low incremental additional cost associated with garnering it. Or to put it another way, the cost of water & soap at the car wash is only a few dollars, so if the car wash could have more cars passing through, any revenue above, say, its $10 variable cost base per vehicle (and that's being generous) goes right to the bottom line.
I paid $50 for the car wash service, which is crazy on a day like today. So what if the business introduced dynamic pricing? On a non-busy day, why not just price the car wash above the variable cost of offering the service. Maybe $15 or $25 instead of $50. That would drive more volume, which would mean more revenue, but more importantly, it would also mean that more existing customers would be interacting with the business and new customers would be trying it when they otherwise wouldn't have. Anytime a business can increase the number of touchpoints it has with customers, it wins because it's getting customers used to interacting with that business instead of a competitor's.
The same thing applies for airline seats -- if a seat is unsold three hours before the flight is due to leave, why not offer the seat at a deeply discounted price to fill it vs. having it fly empty?
I believe the reason dynamic pricing isn't used more often is:
Uber is a leading player in its use of surge pricing, where it raises fees when demand outstrips supply. And it's gotten plenty of negative press for that. But as Jeremy noted in this post, the use of hedging could help mitigate that.
And that got me thinking about how hedging could really drive deep customer loyalty.
As the technology behind setting prices becomes more sophisticated, I fully expect to see dynamic pricing to become more widely implemented. Reasons #1 and #2 above are based on fear, and fear is never a long-term strategy. The only justifiable reason prices have traditionally been static in the past is because we've lived in an analog world. It's been hard logistically to change the price on a bottle of soda when the price is printed on a sticker or on a store shelf. But that's changing, and some things, like stocks, have always been priced dynamically. So I consider dynamic pricing that equalizes supply to demand to be an unstoppable trend, and the businesses that lead with it will benefit the most. .
The problem with dynamic pricing is that humans are emotional beings. If you're used to paying $50 for a carwash, you'll be happy to get it for $15 on a snowy day, but that doesn't mean you won't be unhappy when it costs $75 on a sunny summer day when everyone else wants their car washed. This is where hedging comes in.
What if I could lock in a price (and my loyalty) at the car wash up-front by buying 10 car washes for $500? I'd be paying an average of $50 per wash. The business gets to pocket $500 up-front and keep the interest float on the cash before the service is rendered. When the actual dynamically priced cost of the wash is less than $50, like a snowy day like today, I don't use one of my pre-purchased washes. Instead, I pay the lower rate of $15. So I'm happy on a day like today. But when I show up on a summer day and the cost of a wash is $75, I use one of my pre-purchased washes. Heck, maybe I can still buy the 10 washes for $500 even on the summer day that a single wash costs $75. When I use one of my wash vouchers, I feel like a VIP. I just got a $75 wash for $50. I feel smart for having locked in a lower price. And the business is happy because it's already made $500 off of me.
In this way, hedging a high dynamically priced service could be a great way to create customer loyalty while locking the customer into the service. And using dynamic pricing, just above -- or even at parity with -- the variable cost when there's low demand keeps revenue coming in the door.
I'd love to know what you think. Why don't more businesses do this?
I buy a discounted book of tickets at my car wash and if I get another coupon I use that instead. Airline tickets would save a greater amount of money but if you could navigate that more complex system it would be remarkable. Frequent flyer miles are becoming more inconvenient to use. Clearly your premise of filling seats would benefit all.
Here's a great article about the pitfalls of dynamic pricing. It states:
"There is a thin line between what is acceptable in terms of dynamic pricing and personalized experiences. According to Werner Reinartz, a professor at INSEAD in France, for dynamic pricing to be successful these 5 conditions must be met:
1. Customers must have a difference in their willingness-to-pay.
2. The market must be segmentable.
3. Limited potential for arbitrage.
4. Cost of segmenting and policing must not exceed revenue increases due to customization.
5. Must not breed violations of perceived fairness.
With fixed pricing, businesses can be either missing out on profit or potential sales depending on consumers in the market. Next, businesses need to figure out a way to segment consumers by their willingness-to-pay. In Orbitz case, it is by device, while in the airlines case it is by time. If consumers can buy lower priced goods and sell them for profit, the company is missing out on potential revenue. New technology is allowing the cost of implementation to not exceed the increased revenue from implementing a dynamic pricing strategy.
Lastly and most importantly, the dynamic pricing strategy cannot be seen as unfair by consumers as it was in the Amazon case. Sometimes there is an emotional barrier to break before consumers will accept paying different prices."
During the planning of a recent international trip I asked the CSR for Virgin Australia some probing questions on the pricing of premium economy tickets, specifically why they are so much more expensive than they used to be. I got some insight into how the pricing is determined.. well.. actually I got no "insight" really other than the way the price was determined was a complete mystery to the CSR however suffice to say they used to discount fares as the scheduled travel date approached but found that people were waiting until the last minute to book because they'd worked out how to "game the system". It turns out that airline pricing (for example) is horribly complex. Often the price of a ticket with fluctuate wildly with variables that are not limited exclusively to the seats available on a given flight. So much so that it is almost impossible to predict the trough in a ticket's value.
Certainly the concept of dynamic pricing is woefully underutilized, however the real objective here is maximizing long term earnings and if customers can work out how to game the system there are clear risks involved. Which is why I assume the fear aspect plays, and why mitigating those risks would, I imagine, be a crucial component of a dynamic pricing system.
Yeah I get that but I have to think that the real reason airlines fly with empty seats is a technological one.
What I mean is this: Currently, they charge a huge premium if someone walks up last-minute to buy a ticket for a flight that's leaving in an hour. Which is basically the opposite of what I'm advocating. This must be because after plenty of trial and error, they've found that's how they can currently maximize revenue: Fly with empty seats, but charge an arm and a leg for last-minute bookings.
I don't have industry numbers, but I have to think that volume of last-minute ticket buyers is very low, and it's probably mainly limited to business travelers who have to get somewhere for a last-minute meeting.
But with the proper use of technology, airlines could more finely segment their customer base to create an entirely new category: Let's call it "The risk-taking leisure traveler" -- someone who's willing to wait until the last minute to buy a ticket to fill an otherwise-empty seat.
You mentioned that airlines found in the past that lowering the price of tickets was cannibalizing sales because people were waiting to book. I don't exactly buy that. There's a demand curve based on people's risk profiles. Some travelers will want to lock in a ticket early because they're risk averse. Others will be willing to wait. Right now what's happening is that airlines are not serving the latter group. They're only serving the former group and then betting that flying with empty seats is less expensive than creating a solution for the risk takers.
But with the proper use of technology, it would be possible to finely segment the groups to further maximize revenue. Some ideas off the top of my head:
OK first, I'm no airline pricing expert. Everything I say here is purely anecdotal and generally speaking, I don't know what I'm talking about. But that won't stop me...
I take your point on the risk profile of travelers but I feel like there's also a risk of underestimating both the complexity, and the current level of sophistication in airline pricing. For example, we've all seen the "warnings" about over booking. One can only assume that one of the tools in the airline's pricing belt is overbooking to create an overflow that fills last minute gaps. These last minute gaps may arise for a variety of reasons (illness, lateness, assholeness). Also it may be that the "last minute" traveller is not risk averse, but it also may be that they are desperate. I suspect many last minute travelers are not there by choice, but necessity, and therefore can be held to ransom.
Besides, some airlines do offer last minute deals. On a recent domestic flight on Delta (which wasn't nearly as bad as I was expecting by the way) I noted that for $50 one can "upgrade to first class". Now a first class ticket on a 1 hour flight is about as useful as a fly screen door on a submarine, but will still command much more that a $50 differential at retail rates. By shifting people into first class they are presumably opening up regular seats to either accommodate their own overbooking, or to allow for last minute (highly lucrative) sales.
There is also the issue of downstream sellers like affiliate sites (lastminute.com, travel agencies etc). I am completely ignorant to how they work from a pricing perspective but their very existence must certainly add a layer of complexity.
Also front loading the sales (selling early tickets more cheaply) may have planning and scheduling implications. That is, there is presumably a point at which an airline will cancel or merge flights if their is insufficient patronage. Knowing this early may be a material cost saving.
Having said all that, I can agree with you that from an end user's perspective there is a LOT of room for improvement in the purchasing process for tickets. Airlines do a wonderful job of making the average customer feel like a scumbag degenerate because they weren't willing to pay 5K for a domestic flight, or don't have "unobtanium status" on their frequent flyer bullshit crapfest assbasket program. But the real question is, what effect does this have on revenue? Would improving this really have a material influence on long term earnings? I don't know, but the fact that even the most progressive airlines suck at it tells me that either the industry is dominated by numpties, or it's just not worth the effort.
Perhaps the more interesting question is why the downstream sellers, who presumably are ONLY motivated by selling seats (as opposed to actually flying a plane) don't adopt a more malleable pricing system like the one you describe.
It seems Google is also interested in the space:
Yeah I hear you that it's an incredibly complex system that's very analytics-driven, and that's enough to make an innovator not want to touch it with a ten foot pole. And like you, I'm no industry expert so really we both have no idea what we're talking about -- but that's what makes it fun. It's the old saying that it's those who don't know that it's impossible that actually do it.
I guess the core of what I'm really saying comes down to this: Airlines still fly with empty seats. Yes, they have algorithms that overbook flights because they can predict how many people will be no-shows. But even so, I'd say that the majority of flights I take have at least a few empty seats. And occasionally, the flight will be massively empty. And that's an indication to me -- forgetting about all the complexity in the system for a moment -- that there's an untapped opportunity. An empty seat = a revenue opportunity, no matter how complex things are today. And one way to tap into that revenue oppty. would be via better use of technology to create more dynamic pricing, like deeply discounted last-minute pricing. (The discounted last-minute upgrades you mention are an excellent example about how this is already being done by airlines today and is exactly the type of thing I'm referring to. And I agree about the screen door on the submarine, lol.)
But my larger point is that while dynamic pricing can maximize revenue, what it can really be used for is locking in customer loyalty. And I don't see the airlines -- nor most businesses -- doing that well at all yet. Example: What if I could deposit $1,000 (or $10,000 or $100,000 etc.) as credits on, say, Virgin America. And by doing that, I could use those funds to "get my ticket at yesterday's price." How many times has the price of a ticket gone up, and you wish you'd bought it yesterday? It happens to me all the time. Why not exchange a pre-payment from a customer for a pricing hedge benefit? I'm way more likely to always fly VirginAmerica if I have $1,000 locked up with them ahead of time.
Below is a great post by Dharmesh Shah on building a startup sales team. You can read the full post here.
"Your sales force if your company's lifeblood. No matter how good your product is, it won't sell itself, no matter how much you believe otherwise. Establishing a competent, effective team to draw customers is often challenging for entrepreneurs, though, who would rather focus on research and development or chase VCs.
First off, a few disclaimers: I've never been a sales person. I've never even played a sales person on TV. All the points below have been pulled from startup sales teams that I think work pretty well (including the team at my marketing software startup).
1. Don't hire sales people too early. In the early days, the founders should be able to sell (and should be selling).
2. You don't need sales people, you need sales. Don't think VP of Sales - think "Revenue Engineer". (Not the greatest analogy, but just like you won't hire a development "manager" as one of the first 5 people in a startup, you shouldn't hire a sales "manager" either). Don't get caught up in fancy titles - focus on dollars in the door.
Below is a great post by Dharmesh Shah on building a startup sales team. You can read the full post here. "Your sales force if your company's lifeblood. No matter how good your product is, it won't sell itself, no matter how much you believe otherwise. Establishing a competent, effective team to draw customers is often challenging for entrepreneurs, though, who would rather focus on research and development or chase VCs. First off, a few disclaimers: I've never been a sales person. I've never even played a sales person on TV. All the points below have been pulled from startup sales teams that I think work pretty well (including the team at my marketing software startup). 1. Don't hire sales people too early. In the early days, the founders should be able to sell (and should be selling). 2. You don't need sales people, you need sales. Don't think VP of Sales - think "Revenue Engineer". (Not the greatest analogy, but just like you won't hire a development "manager" as one of the first 5 people in a startup, you shouldn't hire a sales "manager" either). Don't get caught up in fancy titles - focus on dollars in the door. 3. Don't hire several sales people at once. Your goal is to figure out the "pattern" of what kinds of people are best based on what you're selling and who you're selling it to. You need some feedback from the system so you can continue to iterate on your hires. 4. If you've never hired or been around sales people before, be prepared for a bit of a shock to the system. They're not bad people, they're just different. If you're an introverted geek like me, it's helpful to remember that your startup needs to sell stuff. 5. Resist the temptation to create complicated compensation plans. If it requires a spreadsheet to figure out the commission, it's too hard. You'll have plenty of time to confuse sales people later - start simple. 6. Agile methodologies can work in sales as well. Iterate! Refine your demo script, your slides, and any other collateral information. Capture the lessons learned by the best-performing people and spread it to the rest. 7. Sales people will generally act in mostly rational (but often surprising) ways based on incentives. The rules of the game define the behavior of the players. You were warned. 8. Always connect incentives somehow to ultimate customer happiness. If you reward just "deals getting done", you'll get deals - but at too high a price. You might get push-back that sales people don't control/influence customer happiness, but they do. They "pick" customers. They set expectations. And they control the degree of "convincing" applied. 9. Make sure you understand the economics of your business. Figure out your total COCA (Cost of Customer Acquisition). This includes sales people, marketing people and marketing campaigns. Quick example: Lets say you paid a sales person $10k, a marketing person $10k and you spent $5k on Google AdWords (for a total of $25k) last month. If you sold 10 customers last month, your COCA is about $2,500. Different businesses have different needs in terms of sales vs. marketing spend. Make sure neither is too far out of whack. 10. Your life-time-value (how much revenue you expect to generate per customer) should be higher than your COCA. (No, I did not need a degree from MIT to figure that out.) Once your LTV is a multiple of your COCA, you're ready to start turning the knob and scaling the business a bit (hiring more sales people). But, if your LTV is way lower than your COCA, proceed with caution. If there is no hope for LTV getting higher than COCA, you've got a problem. Don't try to hire additional sales people until the economics sort of make sense. If the car is pointed towards a brick wall, hitting the accelerator is not a good idea. 11. Track data maniacally (even if it's just in a spreadsheet). Information you will want includes: What was sold, who sold it, when, for how much, etc. This data will be invaluable later as you start to scale. For example, you should be able to answer the question: We had 14 customers cancel last month - who sold those customers? Is there a pattern? In the early days, you likely won't have the volume (or the time) to analyze the data - but you should at least capture it for future use. 12. Your pricing should be in line with your sales structure. For example, you can't expect to have an outside sales force (that meets with customers in person) if your average deal size is only $10,000. The math won't work. 13. Once you get beyond three or so people, running your sales in a spreadsheet will become painful. Start looking at CRM systems (like Salesforce.com). 14. Start watching the shape of your "funnel" as early as possible. How many leads are you getting a month? How many turn into opportunities? How many of those convert into paying customers? Once you understand your funnel, you can slowly start tweaking your system to fix the "leaks". That's all I've got for now. For those of you that have built early-stage sales teams, what are your ideas and insights?"
Organisations touch our lives everyday, in one way or another, whether you like it or not. We simply cant do without them. From schooling to healthcare, from entertainment to electronics, from food to shelter - organisations touch every aspect of our lives. There are charity organisations, non-profits, non-governmental and for profit organisations. Be it any kind of org, if it needs to be successful and be effective in carrying out its mission, its purpose, then it certainly needs some kind of organising plan. We can call it a business model (even though not all orgs are businesses; we can use the word business in its broadest sense).
The point of this post is not to elaborate on the pros and cons of business models. I simply want to draw your attention to the similarity (often unnoticed) between an organisation (say, a business org) and your career. I think its an interesting similarity.
Every business (and a business plan) has to focus on some key parameters or elements in order to make sense of its existence and be effective in achieving its goals. Like, for example, a business needs to identify who its key customers are, how it plans to engage with these customers and what customers actually want. It also needs to know what its strengths are, its key relationships, its value chain, its cost and its revenue model, among other things. Clarity on some of these key elements are necessary (though they may not be entirely sufficient) for any organisation to achieve its true potential. Typically, these factors and many other facets of a business are covered in great detail in a business plan. This document then becomes a guiding light for members of organisations. Of course, the model can be dynamic (at least the effective ones are) and will need to be modified to reflect actual experiences and changing expectations of all stakeholders.
Now consider your own career. How different is it to the workings of an organisation? Don't elements like customers, value addition, costs and revenues make sense from a career point-of-view as well? Well, I think so and let me explain. Just like any business, you too have customers (the company or organisation you work for). You have costs (in terms of your time spent and other opportunity costs). You certainly need to know your strengths and weaknesses. You need to know your network of key relationships, both inside and outside your organisation. These days you do need a marketing plan and a "personal branding" strategy. Your revenue model is quite simply your compensation (in all its forms). You would do well to have an idea of how you add value to your organisation. You do make investments in many forms like a college education. So, in effect, all those elements apply to you as much as it applies to a business. You are not much different from a business at all, are you?