To understand why, let’s first take a look at the economics that underlie the value proposition for restaurants.

Assumptions on average deal:

• Restaurant profit margins after variable costs – 20%
  • Sales less COGS, labor
• Bill size (for two) – $50
• Groupon – $25 off $50 purchase
• Restaurant payment to Groupon – $10

Value to Restaurant from Groupon to New Customers

Assuming a bill size of $50 and margins of 20%, the restaurant gains $10 every visit.  On the first visit, the restaurant loses the initial $10 (payment to Groupon) and $25 (discount on bill), so it nets a loss of $25.  Therefore, the customer needs to go back to the restaurant at least three additional times on average before it even breaks even.

Value to Restaurant from Groupon to Existing Customers 

If the customer was going to go to the restaurant anyway, then it nets a loss of $35 from the payment to groupon and the discount on bill.

The table below illustrates what these economics mean for restaurants. 

These numbers suggest that if a restaurant is already very successful with existing customers who tend to be very loyal AND if the restaurant has unused capacity, then using Groupon can be a profitable endeavor.  But even those restaurants require a very high mix of coupon purchasers to be new customers.      

Customers don’t decide from whom to receive an email, Groupon or LivingSocial… they sign up for all of them!  Customers decide which coupons to purchase.  When customers have more choices, they are more likely to purchase a coupon from a restaurant they already go to.  So whereas a Groupon deal may have once yielded 75% new customers, the emergence of so many competitors has resulted in a much lower mix.  Personally speaking, around 25% of the coupons I buy are for new places, and there is no way I’m going back to those places 14 times.  I’m sure most restaurants will eventually realize these economics don’t work for them.

It’s obviously a problem when potential customers choose to buy their coupons from competitors instead of with them.  But it’s an even bigger problem when the emergence of competitors serves to undermine your entire business model.  Groupon (or any other Groupon clone) should forget about restaurants completely and build a brand around being the premier provider of coupons for higher margin service businesses.

Instead of targeting advertisements based solely on location, advertisers are able to use behavioral information as well.  If a user frequently visits a restaurant, then a promotional ad targeted to him is likely to be much more effective.  This is good for both advertisers and the consumer.  Advertisers no longer waste money on people who would never be interested in their offering, and consumers are actually happy to get deals from places they already visit.         

Advertisers will want to carefully decide the best way to advertise to each specific customer.  They will want to take a different approach to increase loyalty with existing customers than they will use to attract new customers.  There is a fine balance that needs to be achieved with the ad between adding value to the vendor and adding value to customer.  In the above example, Starbucks runs the risk of discounting a product that would otherwise have been purchased at full price.  Instead, they may want to consider offering a complementary product at a discount.  If robust data shows that people who order Chai Tea Lattes are also inclined to order Banana Chocolate Chip coffee cake, then it may be more beneficial for Starbucks to offer a coupon to redeem it for free if that person has never tried it.

It is much harder to attract new users.  But if advertisers have data on what kinds of purchases people make, then they can provide much more personalized incentives to get these people in the door.  For example, suppose a customer has a propensity to order cheeseburgers at restaurants.  If your restaurant has great cheeseburgers, then it would be a good idea to offer a huge discount or even a free burger knowing that if they like it, then they’ll likely come back many more times.

The prevalence of mobile devices has made location-based advertising a great way for advertisers to extend their reach and effectively target customers.  With LBS applications, such as foursquare, advertisers can couple location with the customer’s penchant for certain restaurants, bars, or stores.  When loyalty deals move from unconfirmed checkins to legitimate purchases, LBA will begin to cross the chasm.  The advertiser will be able to target based on location and purchase behavior, and the customer will receive ads that are so good that they are no longer ads.

foursquare’s and Gowalla’s approach to manually require checkins lets users opt-in, thereby mitigating concerns with GPS inaccuracy, intrusiveness and privacy.   Loopt and Brightkite are other popular LBS applications that have taken the same approach.  Even Facebook plans to add location functionality within the next month that supposedly incorporates a manual checkin.

The checkin feature alleviates certain concerns for users; however, it is still limiting for potential advertisers:

• Checkins are unconfirmed – as long as I’m near the vendor, I could checkin without even visiting
• There is no visibility into how long a user is at a particular vendor
• A visit does not necessarily mean a whole lot if the user does not purchase anything

Therefore, it is particularly difficult to offer loyalty programs.  In fact, some vendors I’ve talked to have refused to offer “Mayor” deals on foursquare because the checkins cannot be validated.  And trying to confirm with GPS, cell-tower triangulation, and WiFi access points will give unreliable location results.

The only way to truly confirm a checkin is if a purchase is made.  Therefore, the mobile applications that are successful will incorporate some sort of payment component.  They need to allow users to link credit card information to their LBS accounts.  This addresses the issue of fake checkins.

• Vendors will be more inclined to offer even better deals knowing every visit is legitimate.
• Loyalty programs can extend to other loyal customers beyond the Mayor.
• Vendors will offer deals based not only on the number of visits, but on the number of purchases or the amount purchased.
• More vendor deals will provide greater incentive for users to join which will in turn incentivize vendors to offer more deals – creating a virtuous circle.

With this information, advertisers can target in unprecedented ways.  For example, it’s great for Starbucks if they know I’m in the vicinity of their shop, but it’s even better if they know that I like ordering their Chai Tea Lattes.  Instead of a display ad saying “Come to Starbucks” they might consider shooting me a promotional ad with fifty cents off a grande Latte.

Location-based advertising (LBA) is effective; LBA coupled with behavior and demographics is extremely effective; but LBA coupled with purchases is the Holy Grail.  At the end of the day, purchases are all that matters for advertisers.

1. Create your own curriculum

Mark wrote down about 15 skills that he thought he would need as an entrepreneur, and he figured out how to get them.  Sure, some were “take this class,” but others were “join this club” and “intern at this startup.”  Mark also prepared himself for a startup by participating in the MIT $50K Business Plan Competition where he was a semi-finalist.

2. Get a lot of experience

Most students interested in entrepreneurship don’t have much experience working at a startup.  Typically a summer internship is the chance to gain this exposure, but Mark recommends doing more.  He took the minimum course-load possible, and did an internship at a startup every semester.  Ask VC’s if they have any startups looking for part-time work.  Find an advisor at school who can help with the job search.

3. Choose your startup carefully

It is crucial to do your due diligence on the firms for which you are considering to work.  Think about it.  VC’s are serious about their diligence and they have 10+ companies in their portfolio, but everything is riding on this one company for you.  So shouldn’t your research be just as thorough?  Mark also recommends joining a company as an equity holder or else joining a later stage firm that already has customers.  If the firm has customers and has been funded, there is still an opportunity to “get the startup experience,” without a lot of the risk that younger firms have.

You can follow Mark on Twitter @markroberge and LinkedIn

Now what about the other 50% in E&I?

Complementing a post by Charlie O’Donnel (First Round Capital) about what MBA’s should be doing to get a job at a startup,  Rob Go (Spark Capital) recently made some great suggestions on how B-schools can prepare their students to become entrepreneurs.  Rob says that “MBA entrepreneurs shouldn’t be looking to join great startups.  They should be looking to start great companies.”  And I couldn’t agree with him more.

Unfortunately, most of us don’t have  ”great” companies to start.  So what happens?  We try to get a job at a startup, thinking this will give us the experience we will need to start a company later.  The logic makes sense, right?  I buy it, and I’ll seek out an internship at a startup (while heeding Charlie’s advice, of course).  The problem is knowing we have this option when we graduate can often lead to our complacency.  Research by Ed Roberts showed that the median age for MIT alums starting a company was 28 in the 1990′s.  That’s trending down from 41 in the 1950′s and 35 in the 1970′s.  One way to interpret the data is it is becoming increasingly more difficult to start a business the older we get.

As Rob implied, if you’re really serious about starting a company, there needs to be a sense of urgency.  We have two years to start something up (really it’s more like 1.5 years because, let’s be honest, you can’t count core semester).  That’s not a lot of time to take advantage of all the resources MIT (or other business schools) have to offer.  We better make it count!

• 60% on-campus interviews with consulting firms, banks, or fortune 500 companies
• 20% externships at relatively small companies hoping to parlay that into summer internships
• 15% traveling abroad or going home abroad to see family/friends
• 5% pursuing actual side interests by taking not-for-credit courses offered by MIT faculty and students

I fall into this last group, as I took a bunch of web programming classes on PHP, HTML, CSS, etc.  One class provided a great overview on the semantic web, which is all about linking data on the back-end of various websites to provide even more relevant results on the front-end.  A highlight included a talk by Tim Berners-Lee, founder of the World Wide Web.  Hell, I also dropped $200 on programming books from Amazon.  I know what you’re thinking -this doesn’t sound like much of a winter vacation- so let me explain…

You see, I left my gig in management consulting (@ Mars&Co) last year for one reason, and one reason only – to start a company in the Web/IT space.  Of course there was just one little problem:  I had no idea what this startup would be.  Lucky for me, I was accepted to MIT  with the hope that when the “big idea” came, I would have the resources available to make it a success.  Well, that idea has come, and taking advantage of these IAP courses was the first step.

The truth is my IAP spent on web programming may not have been as fun or interesting as an externship at a hot startup (or traveling around Southeast Asia for that matter).  But I know I made the right decision because the knowledge I have gained will help lay the foundation for doing what I came here to do.  I may not ever be able to build a compelling website from scratch, but if I need to outsource or interact with a programmer, I’m ready.