by Robert Jackson (11/16/ 2010)
A basic screen would show that OpenTable (OPEN) is overvalued by several conventional measures. The slightest hiccup in the company’s performance could send the stock careening back to something closer to traditional measures of fair value. The company has made money by hawking empty tables online for restaurants, themselves being low margin businesses, many of which close after a few years. Since 30% of the float is already sold short, you would not be alone in thinking that OpenTable is a faddish website business with limited ability to fit into its high market cap. I too was formally short OPEN. My mistake was that I didn’t consider the pace with which quality management would innovate and exploit economies of scope. Now, I am long OPEN and would like to explain why the commonly circulated short ideas are wrong, lack industry perspective and offer poor risk/reward. I am not going to reprint the short viewpoints or identify the publicized short ideas as this information is easy to locate by the interested reader.
Instead, I will refute the major short themes and present other optionality for OpenTable which is still too early to model but exposes the short seller to unnecessary risk. The most commonly cited reason as to why OPEN’s valuation makes no sense is a limited market opportunity. This typically involves a calculation of the number of possible global restaurant clients and
a best estimate of how many diners will reserve a table through the system. Indeed, modeling trailing twelve months revenue by this method would not justify the current valuation. Regardless, sell-side analysts have chronically underestimated earnings since the IPO. Why? This has primarily been due to the growth in mobile reservations which have exceeded expectations as smart phones and their geolocating apps continue to penetrate the market. Thus, it is somewhat difficult to anticipate the growth trajectory of seated diners and the likelihood of achieving market penetration of 24% as has been achieved in San Francisco versus 8% in North America. The network effects of increased restaurants and increased seated diners is difficult to model but is present as evidenced by the number of seated diners per installed restaurant being relatively constant over time which means that despite the increase in table inventory, these tables are being filled through increased diners and increased dining frequency.
The growth rates in restaurant clients and seated diners have demonstrated great execution and could be clarified by management’s discussion of restaurant churn rates given the volatility of the industry. This would help with understanding long run cost of sales. The balance of this discussion will reveal that OPEN’s historical revenue model has very little to do with future revenue projections. First, we have to deal with competitive threats. At the outset, it is worth noting that OpenTable is 12 years old. It took a long time to build a brand which attracts the interest of diners while at the same time keeping enough restaurants engaged in the service to make it useful to diners. Since OPEN may have now educated diners and restaurant operators on their model, in theory, new entrants would be more efficient in their growth despite the existence of a dominant competitor. However, the industry is highly fragmented geographically and is primarily comprised of independent restaurants with a handful of chain restaurants (i.e. Ruth’s Chris, Fleming’s, etc.). The sales process is serial and restaurant owners are notoriously hard to reach relative to other retailers. OpenTable is effectively a natural monopoly media company since the dining audience reach and restaurant (advertiser) must both exist for the
model to work. Ninety percent of OpenTable’s traffic originates from its own website so it would be naïve to suggest that other entrants could instantly partner with other media to gain reservations. Since
OpenTable is actively engaged in the reservation function, the requirement for operational integration is a massive barrier to entry. Restaurants have to abandon their old process and re-train on the OpenTable process. The operational function must be maintained and managed by OpenTable and the restaurants on a moment by moment basis. The economies of scale for this aspect are not appreciated by new entrants as witnessed by UrbanSpoon’s half-hearted entry having after 1 year gaining 1.2% of the restaurant count of OpenTable and a far lower reservation count with muted network effects. Real competitive threats will require significant capital and time.
When these competitive threats come they will likely choose to compete on price. Is price a reasonable strategy to dislodge OpenTable? No. Restaurants meal periods are sort of like commercial airplanes flights. Assume that fine dining restaurants are open 6 days a week and offer 3 seating times per evening (i.e. 5PM, 7PM, and 9PM) and that the restaurant seats 75 people. Thus, the maximum number of people that can be served is 225 (3 seatings times 75 people). The load factor is essentially the proportion of seats filled and is a good predictor of profit or loss. With the exception of raw food, most costs are essentially fixed (i.e. lease, prep work, line cooks ready to cook, management, equipment, cleaning, etc.). Since the servers are working for tips essentially on commission, it’s important to have enough work for them or they quit and go somewhere else. At least airlines don’t have this problem and are able to pay the flight attendants enough to show up and allow the flight to leave. The restaurant operator is challenged to keep the restaurant as full as possible which requires great food, great service, great marketing, and as much demand forecasting as possible. Thus, the relationship of online travel sites to the airline, hotel, and rental car industries is a reasonable proxy for understanding OpenTable.
Even with very few suppliers (i.e. airlines, etc), there are limited number of media properties that resell most of the inventory. Restaurants chronically suffer from low working capital and have limited credit with their suppliers so every meal (i.e. flight) requires a good load factor to maintain cash flow. Reservation suppliers that meet this need will earn a premium fee while traditional branding media (i.e. print, radio, T.V.) which cannot be accountable to an ROI goal will continue to be less relevant to restaurants. Assuming a 3% of sales ($15b, see below) advertising allocation, OPEN is reaching into an advertising budget of about $450m per year among all of their restaurant clients. Also, some restaurants are already seeking to be destination restaurants with lower leases by leveraging the OpenTable community and relying less on walk-in diners available only in high traffic areas. Another favorite argument of shorts is that there is declining revenue per restaurant. In fact, I would completely expect a monopoly business to engage in price discovery through price discrimination. These effects are largely driven by r restaurant demand in the case of OpenTable. There are many more restaurants which do not have a large need for reservation systems or a large number of reservations but who still desire to market their restaurant to OpenTable users. Thus, the OpenTable Connect service was designed to meet the needs of these restaurants with a lower monthly subscription fee which is essentially a simple advertising fee (no ERB hardware) plus a higher per seated diner fee of about
$2.50. I would expect this model to evolve however as they discover the dynamic which works best for both OpenTable and the restaurant client. All media companies offer tiered pricing to meet the needs of different advertisers (i.e. 30 second vs. 60 second commercials).
What if OpenTable actually started to sell meals directly to the consumer instead of just booking the reservation? This is what they are now doing with the Spotlight feature but for some reason shorts do not consider the potential of this aspect of their business. OpenTable has a tremendous opportunity to perform e-commerce for an industry which essentially obtains zero e-commerce revenue. A simple way to think about this opportunity is to estimate the amount of product that is sold by their restaurant partners by assuming $1m gross sales per restaurant across 15,000 restaurants which is $15b. The Groupon-like model is being beta tested currently and anyone can look at the gross revenues for this business on their website. My modeling shows that OPEN can achieve $25-50m EBITDA in 2011 using different assumptions based on the frequency of deals in different US cities.
What if OpenTable helped their restaurants to do the very thing that makes OpenTable so popular in the first place? To recap, OpenTable exists because restaurants are not good at answering the phone, particularly not during a meal period. Also, restaurant staff work hard, long hours and although administrative assistants across the country would like to speak to someone in the restaurant to make a reservation at 8:30AM, the chances of doing that efficiently are quite low, so they go online and use Open Table or call back later. OpenTable recognizes that 92% of all reservations are taken via the restaurant’s telephone. Having recently acquired a company that was in the business of being an outsourced reservationist for multiple fine dining restaurants, OpenTable is now expanding this transaction space to their platform. This is a profitable model ideal for a call center application since call times are short and minimal agent training is required. Of course, not all restaurants will participate but based on a 10% participation rate, I estimate a $6.5m EBTIDA contribution in 2011. Shorts do not discuss or understand this initiative.
Shorts do not recognize the propreitary value of the OpenTable software. OpenTable is both an electronic reservation book and marketing platform. These sorts of applications have a legacy of being inherently sticky in restaurant operations. For example, the new Flex mode upgrade allows the reservation system to understand requests for larger parties where tables must be reset and combined. This is useful because many hostesses are typically the least experienced members of the restaurant staff but who have to make critical decision about seating guests. Several national vendors provide cash register and food production systems for restaurants at very high cost (i.e. Micros). These systems are critical to operations and P&L reconciliation for controlling food cost and labor. Similarly, the ERB is a critical revenue management tool which has inherent value independent of the reservations generated from the OpenTable users. Restaurants and OpenTable work together or independently to maintain guest databases and communication. From personal experience, I have dined in about 100 fine dining restaurants over the last 10 years or so. When it was easy for me to provide my email address, I did so and yet my experience has been that only 1 restaurant out of 100 has sent regular email communication.
Only 3 out of 100 have sent any communication at all. Restaurants essentially have no idea how to communicate with their customers and do very little marketing to their guest base. Shorts do not consider these factors when discussing barriers to entry or the value of the software embedded in the monthly subscription fee.
Other revenue options include restaurants upgrading placement on the site and advertising by 3rd parties, both of which increased by at least 100% in the most recent quarter. Management is providing little guidance on these revenues at this time and is not breaking them out so there is little clarity on this aspect of the business. OpenTable possesses as unique and proprietary walled garden which provides substantial advantage to their restaurant clients. There are several additional transactional services that OpenTable may offer in the future which they are not currently offering. For example, they can sell full price restaurant gift certificates and assist restaurants with the fulfillment of those orders for a small fee. Most restaurants do not offer online gift certificate sales through their website. OpenTable can leverage other social networks such as Facebook or LinkedIn to organize group meals such as through an RSVP based invitation. OpenTable can provide incentives to diners to accept off-peak reservations to assist their restaurants with better revenue management which is de rigueur in the airline industry. I have described key fundamental characteristics which explain the seemingly high valuation for OpenTable. If these new revenue streams develop as projected, the current valuation will be shown to have been much closer to fair value than anyone is suggesting which is why shorting OpenTable makes no sense.
The writer can be reached at