I was in Groupon’s office in December 2010 when they received an offer from Google to buy the company for $5 Billion. They ended up turning down the offer and taking the company pubic at an even higher valuation, only to have it crash later when it became clear the company couldn’t sustain itself.

The top complain I’ve heard on why Groupon ‘failed’ was that its core product of 50% off with 50% margins did not work for merchants, and as soon as they made it work for merchants (by making the discounts less ‘real’) it stopped being compelling for consumers.

While I won’t disagree with that, I think the reason they failed was subtly different. I believe they failed because they forgot about the power of scarcity.

To understand that power we need to go back to before the first dot.com boom. Back to when Hotels.com was still 1-800-HOTELS.

Here is how 1-800-HOTELS got started:

They would launch in a market (a city). They would call all the hotels in that city and say: “We would love to put you up as an option when people call 1-800-HOTELS. Here is how it will work: You need to offer the consumer 50% off the rate they would get if they called you directly, and we will take 50% margin.”

That’s a 75% discount! What hotel would sign-up for that?

Turns out the smart ones would. A hotel is very high fixed cost and very low marginal cost. So if a hotel has a single empty room, they can make more money by filling it at a 90% discount than if they leave it empty. If they filled all their rooms this way they would be in trouble, but if they haven’t managed to fill rooms, then filling one at a huge discount makes sense.

But wouldn’t the smart hotel fill all their rooms? Nope. If you are a smart hotel and all your rooms are consistently selling out you have your prices wrong. You should be raising your prices and ending up with some vacancies (and have rooms available at the last minute for people who are price insensitive).

So most hotels will have empty rooms (at some price). If they could fill them with 1-800-HOTELS they should.

The problem is most hoteliers haven’t internalized the last few paragraphs. They aren’t quantitative data-driven marketing experts. They are operators who are much better than I am at actually running hotels. They haven’t put in the time to fully grasp the marginal economics. And they definitely didn’t get it 20+ years ago when someone from 1-800-HOTELS gave them a call asking them to join the program.

So why did 1-800-HOTELS work?

Because they didn’t need all the hotels to sign-up. They didn’t need most of the hotels to sign-up. They likely needed about 10 hotels per city to make their model work.

Here’s how it would work from the consumer side:

You call 1-800-HOTELS. You speak to an operator and tell her which city you are looking for and maybe what type of hotel (4-star for example). The agent pulls up her list. She says, “Yes. It looks like I do have a 4-star hotel available. In fact I have it at 50% off. Would you like me to book it for you?”

You may ask for more options, but would it matter if she only had one? 50% off is a pretty great deal. Maybe if you were a business traveler and someone else was picking up the bill you might demand a specific location, but for the rest of us it’s a pretty good solution. Chances are you don’t know where you should stay in the city anyway most of the time.

So 1-800-HOTELS doesn’t need all the hotels. They need one or two in each ‘category’ (in this case I’ll bet a ‘category’ is a general price point and location). So about 10 per city.

They don’t need all the hotels to ‘get it’. They need about 10 of them.

That is the power of scarcity.

 

Note that power doesn’t exist in classified sites. Hotels.com today needs as much inventory as they can get. It’s been shown that more inventory is the number one driver of both short-term conversion rate and long term success for classified-type businesses. Craigslist doesn’t have a very good product (as many entrepreneurs tell themselves as they try and create CL-killers), but they have all the inventory – and you can’t beat inventory.

Unless you can build a model that doesn’t need inventory. Something that relies on scarcity.

 

Priceline did it.

1-800-HOTELS continued to grow their inventory over time and eventually got bought by Expedia. Now Expedia had all the inventory. How was a new entrant supposed to compete.

Priceline changed the game – basically by re-inventing 1-800-HOTELS for the internet age. On Priceline you put in a ‘bid’ for a hotel, and only after you commit to staying at a 4-start hotel with a pool in Boston for $83.50 do they tell you if (1) You bid was accepted, (2) Where you are staying.

A lot has been said about the auction system for Priceline (and how it differs from the set-price of Hotwire), but to me the big winner here was that they created scarcity: Expedia needed all the inventory, and Priceline needed about 10 hotels per city (willing to deep discount).

 

Once you recognize this concept you see it everywhere. Google has effectively unlimited ad space, but they only have one top spot for any given keyword. By creating that scarcity they let the market bid up the value for that spot. If Google had just created a bunch of ad units and then sold them at fixed prices they would not be the giant they are today.

Which brings us back to Groupon.

 

When Groupon launched they offered one single deal a day per city to their subscribers. They had to start out in hard-core sales mode negotiating those deals with skeptical merchants. But they had the power of scarcity on their side. They only needed one deal a day to make the model work (and there are a LOT of merchants in any given city).

In fact the model worked so well that they soon had waiting lists going out 6-months or more in many cities. At the same time consumers LOVED the deals (who wouldn’t – they were great deals at this point) and were willing to get more than one deal a day. This wasn’t lost on other entrepreneurs who started competitors to try and pick off demand on both sides that Groupon couldn’t fill.

This is where Groupon took a wrong turn.

They saw these competitors popping up and said, “We already have users that would be willing to get more deals. And we have a six-month waitlist for merchants, why don’t we start giving out two deals a day?”

In fact there is no real limit to how many deals Groupon could give out to users. What started with 5 deals a week turned into 7. Then into 14. Then 21. Now it’s off the charts. I just checked out their Seattle page where there are 16 DAILY deals (that’s 112 if it repeats all week which I expect it will).

What does this reduced scarcity do?

It allows the merchants to negotiate for ‘less good deals’. Groupon’s margin decreases, and the quality of the deal to the consumer decreases. Instead of deals like “50% off one of the best restaurants in the city any time”, you now see deals like this:

29% off River Rafting, only available June 7, 14, 15, 21, 28, 29 or July 12 (Even better: They have the same discount on their website: “ONLY $99 (Normally $139) with code ‘RAFTNOW’ at checkout!)

Or

$35 of $60 worth of food at Ipaneme Grill (where the price for their core offering is $50 all you can eat meat, so in practice you are either paying $35 for $50, or more likely $75 for $100 for two people)

 

Basically Groupon went from have a deal every day that was so fantastic it made sense to stop and read it every day and buy it if you were at all interested, to a place that distributes standard coupons. These coupons have been around forever. Some of them are good, but most aren’t, and the time and trouble it takes to figure out if it is worth it usually isn’t worth it.

But what if they had done it differently?

What if, instead of going from one deal a day to 16 deals a day, they had stayed at one deal?

All of a sudden the arguments about whether or not Groupon was good for merchants would become moot. Only 365 merchants a year would be able to offer deals. If you offered a ‘standard deal’ (50% off at the right price point with 50% margin to Groupon) you go the back of the queue. Your deal may go live in a year. If you want to do something faster than you could offer a better deal to skip the queue. Groupon could develop tools to determine in advance how good a deal was (not just discounts – you could imagine a good deal at a top in-demand restaurant is more valuable than a great deal as an average place). Another component to the deal is how many vouchers you would allow to be sold (more is better).

Basically Groupon would need to develop a ‘quality score’ for each deal to rank them. And then let the highest quality deals be shown first (the second best can wait in line, the third best sit there only to be used if they have a gap in inventory for some reason).

Now this method may mean that at first Groupon does not sell as many deals as their ‘lots of deals’ model. Initially when they moved to 14 deals a week, they likely saw an uptick in sales. They even created a special team whose job it was to determine what the ‘lead deal’ should be for each used based on their demographics and purchase history. You have to believe that if, instead of a Nail Salon deal only, they had a Nail Salon AND a restaurant deal they would get more sales.

The problem is, as soon as you go down this path you start messing with the very thing that made the business successful: Scarcity.

 

My idea of staying at one deal per city per day may be too extreme. Maybe the right answer is to split the city into east and west side, or split the customer base into women and men (or ‘feminine deals’ and ‘masculine deals’). Groupon should have been able to make some smart judgments at this point – if the wait list for a ‘good deal’ was 6 months then maybe start 2 deals/day. If the it was only 1 month than maybe pull back on the number of deals. They could test the impact on inbound inquiries and deal quality scores based on the length of the wait list.

Instead what was done (I believe) is they tested the revenue impact of going to 2 deals (which was significantly positive the first day it was done). Then they tried 3 and that worked too. Then they started extrapolating how many they could do and they started hiring sales people to fill the now enormous queue. All of a sudden instead of being order takers of a very valuable scarce product they were salesman trying to convince merchants to let them distribute coupons.

It was the beginning of the end.

 

The Message:

Scarcity is a very very powerful thing. You want to do everything in your power to create it when you are building a business. And once you have it, you want to hold onto it and monetize it the best you can. Do not be tempted to take the short term benefit of selling more if it means giving up on that exclusivity.

 

There are lots of examples of companies using scarcity (or artificial scarcity) to their advantage. If you can think of some, feel free to comment below.