Description

TripAdvisor was founded by Steve Kaufer and Langley Steinert in 2000. Kaufer, the story goes, had difficulty finding unbiased opinions about hotels and vacation destinations, so he decided to build the solution. The small team started out by collecting reviews and opinions found in guidebooks, travel magazines, newspapers, and blogs and entering them into a database. The company did not have a consumer-facing website at the time, but rather was trying to sell a service to travel sites and search engines that could query the database and retrieve the content for their users. After that approach was unsuccessful, the company built a website and allowed readers to add their own reviews. Traffic to the website took off. The team discovered users preferred to read the user-generated reviews rather than the professional reviews the team had curated from the guidebooks and magazines. The increasingly popular website did not begin making money until they inserted text links that allowed users to click off to Expedia to book the hotels. Expedia paid for this referred traffic and the business grew from there.

In 2004, InterActiveCorp (“IAC”) acquired TripAdvisor. Steinert left the company and Kaufer remained as CEO. In 2005, IAC spun out Expedia, which included the TripAdvisor business, into a separate publicly-traded company. In 2011, Expedia spun out TripAdvisor. Today, TripAdvisor’s largest shareholder is Liberty TripAdvisor Holdings, which owns a 21% economic interest and 56% voting interest through its ownership of 18.2 million common shares and all of the 12.8 million super-voting Class B shares.

Financial History

TripAdvisor has grown revenues at a 34.2% annual growth rate since 2006.

The asset light nature of the business allows for phenomenal triple-digit returns on tangible capital.

TripAdvisor’s Evolving Revenue Model

Over the years, TripAdvisor has made major changes to the revenue model of its core hotel business.

Originally, users interested in a particular hotel would be prompted to click on links for Expedia and other online travel agents (the “OTAs”) to check rates. Users would get a pop-up window for each OTA they clicked on and they could check rates and complete a booking on any one of those partners’ websites. The OTAs paid TripAdvisor a “cost-per-click” or CPC every time someone clicked. This was a profitable revenue model, but the large number of pop-ups made it not the best user experience.

In 2013, TripAdvisor added metasearch functionality where users are shown all the available rates offered by the OTAs directly on the TripAdvisor site. Users no longer have to click off to check rates. Those who do are taken directly to the OTA’s hotel booking page where they can complete their booking. The pop-up window model was extinguished. With metasearch, far fewer users click on the links as compared to the “pop-up” version since the rates are already shown, but those who do click are far more likely to complete a booking or “convert” on the OTA site. Essentially, with metasearch TripAdvisor sells fewer, but higher quality leads that convert at a higher rate. Metasearch initially weighed on TripAdvisor’s financial results because it was new and users were not accustomed to it. Within about 12 months, metasearch’s higher CPC pricing fully offset the fewer clicks and TripAdvisor achieved revenue neutrality. I believe metasearch was always intended to be a stepping stone towards adding functionality enabling users to complete a booking on TripAdvisor.

In 2014, TripAdvisor began slowly rolling out an ability to complete bookings directly on TripAdvisor (“Instant Booking”). TripAdvisor collects the user’s payment information and essentially passes the completed booking off to the OTA or hotel. The booking partner processes the transaction and handles all subsequent customer service. The big difference is TripAdvisor earns a percentage of the booking as a commission, much like the OTA revenue model, rather than a CPC for a lead that might convert.

It took longer than management expected to get hotels and OTAs to join the Instant Booking platform. Some hotels initially expressed reluctance because they worried it was another expensive booking channel and they are trying to drive travelers directly to their own sites. Expedia and Priceline were initially reluctant because Instant Booking competes with them for bookings and the initial version of Instant Booking did not sufficiently highlight the OTAs’ brands. Gradually, the hotels began to sign up because they liked paying a lower commission rate than they pay the OTAs and they saw it as a way to potentially gain share on TripAdvisor at the expense of the OTAs. Priceline finally agreed to join last Fall after a long negotiating period after TripAdvisor improved the branding issue. Priceline wanted to not only avoid losing share to hotels, but also wanted its Booking.com to gain share in the U.S., which was made possible by an exclusivity period it negotiated with TripAdvisor that effectively shut out Expedia from the Instant Booking platform for a period of time. Today, the Instant Booking platform has 8 of the 10 largest hotel chains in the world, a growing number of independent hotels, Priceline and its subsidiaries, including Booking.com, in addition to smaller players. Importantly, Expedia is also very likely to join once Priceline’s exclusivity period ends.

Closing the Monetization Gap

TripAdvisor’s evolution towards Instant Booking is driven by a desire to close the monetization gap it has with its industry peers. TripAdvisor’s peers make far more money per visitor than TripAdvisor does because TripAdvisor suffers from extensive “leakage” with metasearch. Leakage refers to the fact that TripAdvisor only monetizes a small fraction of its users.

Leakage occurs because many TripAdvisor users read the site’s content but do not click on the links. When they are ready to book at a later date, they may search the internet for the best price and/or go directly to an OTA or the hotel’s site directly. TripAdvisor does not get paid for these users. Other TripAdvisor users click on metasearch links, but do not book immediately. When they are ready to book, some return directly to the OTA or hotel site directly, bypassing TripAdvisor. While TripAdvisor drove that booking and got paid for the click, it does not always get credit for driving the conversion. Since the amounts the OTAs bid for clicks in the metasearch auction is based largely on how effectively TripAdvisor’s leads convert into bookings, in these cases, those bids are lower than they should be. Essentially, the OTAs and hotels are underpaying TripAdvisor.

The other major impetus for Instant Booking has been the rapid growth of phone traffic. In 2012, TripAdvisor’s phone traffic monetized at just 10% of the rate of desktop traffic because the user experience was so inferior. Given the ongoing mix shift towards phone traffic, TripAdvisor needed to improve the user experience to improve phone monetization. The clear opportunity was to allow users to complete a booking on the TripAdvisor mobile site or app to reduce friction and leakage.

Enter Instant Booking

Instant Booking was designed to get TripAdvisor more credit for the bookings it drives for its booking partners and to improve monetization, especially on the phone. For virtually all hotels, users now see metasearch links as well as an Instant Booking or “Book Now” option.

Instant Booking rolled out to 10% of U.S. phone traffic in the Spring of 2014. Management liked what they saw and expanded it to more phone traffic as well as desktop and tablet users. By September of 2015, Instant Booking was fully rolled out to all U.S. and U.K. users across all devices. It launched in all other English-speaking countries in January of 2016, the rest of Europe’s major markets in February, and the Asia-Pacific and Latin America regions in March. Instant Booking is now live globally on all devices.

The Economics of Instant Booking and Metasearch

In the short-term, Instant Booking has a clear negative impact on TripAdvisor’s financial results. Initial conversion rates are lower than metasearch and TripAdvisor is offering low initial commission rates to incentivize hotels and OTAs to join the Instant Booking platform.

This negative impact is best illustrated by examining revenue per hotel shopper – TripAdvisor’s key monetization metric. This metric has fallen from as high as $0.60 in mid-2014 to a low of $0.46 this past winter before improving slightly to $0.48 in the second quarter of 2016.

Lower revenue per hotel shopper has had an adverse impact on Hotel revenue growth and margins.

[Note: Hotel Adjusted EBITDA has been adjusted to exclude non-recurring television advertising spending in 2014 and 2015.]

This has occurred because Instant Booking is taking share from metasearch, which despite its shortcomings, was a relatively efficient model for TripAdvisor. Many of the OTAs “bid to breakeven” in the metasearch auction, meaning the OTAs’ immediate profit from the converting leads simply pays for the cost of acquiring those leads. The OTAs are fine with that because they focus on the higher lifetime value of a user, not simply the immediate value.

Exhibit 6 shows the key variables in play as it relates to the OTAs bidding in the metasearch auction. TripAdvisor and the OTAs do not explicitly disclose conversion rates and commission rates, so these numbers are estimates and meant for illustrative purposes.

Essentially, the OTA would generate $255 of revenue assuming it converts 5% of metasearch clicks to completed bookings and captures 17% of the $300 average booking amount. Given the competitive nature of the metasearch auction and the OTAs’ willingness to bid to breakeven, they might spend all that revenue on acquiring this traffic, so they might pay TripAdvisor up to $2.55 per click.

Instant Booking’s initial economics are much different. Conversion rates are lower than metasearch because it is a new user experience and TripAdvisor’s commission rates are lower than the OTAs earn because it is being promotional to incentivize booking partners to join the platform. Exhibit 7 shows how the lower conversion and commission rates result in lower “effective CPCs” for TripAdvisor than it earns through metasearch.

As you can see, the lower conversion and commission rates significantly reduce the effective CPC that TripAdvisor makes. Importantly, the economics of Instant Booking should improve materially over time.

Hotels to Gain Share As Auction Dynamic Develops

Instant Booking should evolve into an auction dynamic as the supply of viable Instant Booking partners for every hotel increases. Hotels are joining the platform because Instant Booking significantly increases their ability to gain share on TripAdvisor at the expense of the OTAs.

With metasearch, hotels could not effectively acquire leads because they were usually outbid by the OTAs for the top placements. This occurred because the OTAs convert metasearch leads better because they offer a wide selection of alternative hotel choices should the user pass on the original hotel choice; the hotels by definition only offer one choice.

Instant Booking changes those dynamics. With Instant Booking, TripAdvisor is selling fully-converted bookings, not leads. TripAdvisor chooses its Instant Booking partner for a given hotel based on two variables: 1) the commission rate paid to TripAdvisor, and 2) the conversion rate.

Hotels are currently signing up for Instant Booking and agreeing to pay 12%–15% commissions to TripAdvisor. Priceline and TripAdvisor have not disclosed the commissions Priceline is paying but I believe it is about 10% on average. So the participating hotels are more likely to “win” the Instant Booking button over Priceline or other OTAs because TripAdvisor will partner on a given hotel with the partner who pays the highest commission rate, provided that the hotel’s conversion rate is sufficient.

Imagine an independent hotel pays a 15% commission but only converts 3% of hotel shoppers, while Priceline’s Booking.com pays a 10% commission but converts 8% of hotel shoppers. Generally, TripAdvisor will partner with Booking.com because the higher conversion more than offsets the lower commission and maximizes TripAdvisor’s revenue.

[Note: These figures are for illustrative purposes only.]

Clearly, the hotels are incentivized to improve their conversion rates so they can be chosen as the Instant Booking partner more often. TripAdvisor is also incentivized to help the hotels improve conversion rates because it increases the supply of viable booking partners, making the auction more competitive. That explains why TripAdvisor is working with the independent hotels to improve their content, including room descriptions, room choices, and pictures. Better content leads to higher conversion rates.

As hotels improve their content and conversion rates, they will increasingly be on comparable footing with the big OTAs as far as conversion. At that point, the booking partner that pays the highest commission rate to TripAdvisor will be selected as the booking partner. All of a sudden, Instant Booking becomes a competitive auction like metasearch.

The auctions will continue to get more competitive as more hotels and OTAs join the Instant Booking platform. Expedia is expected to join in the coming quarters after the exclusivity period that Priceline and TripAdvisor negotiated ends. On Expedia’s second-quarter conference call in July, CEO Dara Khosrowshahi said the following:

“Obviously, we are not in Instant Book at this point. We think the product has improved as far as the brand representation of players who are working with Instant Book. And so we will be in dialogue with TripAdvisor when we can, although I wouldn't expect anything to happen soon.”

To understand how the Instant Booking auction will evolve over time, one must consider the perspectives of the hotels and the OTAs. Again, the hotels are currently paying 12%–15% commission rates through Instant Booking, which is a bargain compared to the “upper teen” rates they pay the OTAs. So while the hotels like paying lower rates, they are willing to pay more if they must.

Priceline and Booking.com seem to be paying about 10% commissions to TripAdvisor. TripAdvisor was willing to be so aggressive on the commissions due to the strategic importance of getting Priceline’s hotel content on Instant Booking. Therefore, Priceline appears to be earning an unusually high spread between what it pays TripAdvisor for completed bookings (~10%) and what it turns around and charges the hotels for the same bookings (“upper teens”).

However, as more hotels join the platform, and improve their conversion rates, and Expedia joins the platform after Priceline’s exclusivity period ends, the additional competition should drive up the “commission rate bids.” After all, the independent hotels are used to paying rates in the upper teens to the big OTAs and the OTAs are comfortable breaking even on their traffic acquisition channels. Therefore, the commission structure should evolve from the left-side diagram in Exhibit 7 to the right-side diagram as more booking partners join the platform and the competitive auction dynamic develops.

I think conversion rates are the bigger wildcard. If conversion rates on Instant Booking surpass metasearch conversion rates, which already seems to be occurring in Instant Booking’s most mature markets, Instant Booking’s economics will blow away that of metasearch. Exhibit 10 shows my rough approximation of how TripAdvisor’s economics differ between metasearch, Instant Booking at launch, and Instant Booking at maturity.

Exhibit 10 is intended to be a high level view of the transition TripAdvisor is undergoing and the key variables are only rough approximations. The key question is “How high can Instant Booking conversion rates and commission rates go over time?”

Instant Booking Progress

Despite the weak headline financial results, TripAdvisor is making excellent progress with Instant Booking in the markets it has been in the longest. Instant Booking first launched to a portion of U.S. phone traffic and the company steadily increased the rollout as it liked the results. Next, it was rolled out to tablet and desktop users in the U.S. and U.K., then other English-speaking countries, then the rest of Europe, and finally in Asia-Pacific and Latin America earlier in 2016.

On November 11, 2015, Kaufer said the following:

“And we have 18 months of history where we have improved our conversion quarter after quarter after quarter so that we are confident we're always going to be making progress. It may not always be as fast, or it may not always be instantly, but we'll get there, and that's what gives us confidence to do a quick rollout, even if it's not instantly accretive.

The reason for my optimism on where this story ends, is because we’ve seen from the instant-book rollout, we’e seen the improved conversion — slow but very steady. We know what we need to do to improve conversion going forward, with Priceline and the room content management system and all the rest of getting those economics in shape, and over time, training the user. In mobile, we’ve seen better monetization over time. So the story looks really nice when you play it out over time. And we’re in it for the long term. So if there’s another revenue trough, so be it. We know where we want to get to.”

On August 4, 2016, Kaufer wrote the following:

"Our efforts are yielding nice gains in user conversion rates, repeat rates and monetization rates. Looking at our longest standing market, the U.S., desktop instant booking conversion rates have grown approximately 20% year-over-year. U.S hotel shoppers that click on instant booking are increasingly repeating on our site, coming back to shop, input dates, and click, when compared to users that click off to a partner in metasearch.

Travelers are increasingly storing their credit cards to make future purchases fast and easy, and the percentage of bookings using a stored card continues to grow. A substantially higher percentage of first time bookers are coming back to book again compared to last year. Prior instant bookers are converting at better rates than first-time instant bookers and prior meta bookers. A greater percentage of prior bookers are also visiting TripAdvisor through direct channels, which have much lower overall acquisition costs.

Mobile is playing, and will continue to play, a growing role. TripAdvisor-branded click-based and transaction revenue grew in excess of 30% on phone in the second quarter. One third of hotel shoppers in the second quarter visited on the phone, and on some days phone represents more than 40% of shoppers. We are most pleased with our growth on our app, where hotel shoppers and TripAdvisor-branded click-based and transaction revenue each grew by approximately 50% in the second quarter. Looking again at the U.S., we are pleased to see our most engaged shoppers on our app, as users are more likely to convert in our instant booking flow than if we were to send a user off to a metasearch partner. High repeat usage rates on app exhibits our brand strength and engagement potential. On the monetization side, this year we have seen that repeat revenue from U.S. mobile app bookers is approximately double the repeat revenue from other devices.”

The fact that Instant Booking is clearly working in the markets it has been in the longest gives me confidence that the more recently launched geographies will follow the same path. Revenue per hotel shopper has significant upside just getting back to where it was two years ago, but the purpose of Instant Booking is to improve that metric far beyond previous levels. This strategy will take time, but it does appear to be working.

The Non-Hotel Segment

About 50% of TripAdvisor’s traffic visits the site for the attractions, restaurants, and vacation rentals content. TripAdvisor did not begin to monetize this traffic in a meaningful way until it acquired La Fourchette, its restaurant reservation business in Europe, and Viator, the leader in attractions booking, in 2014. The company is investing heavily in expanding these businesses around the world, which explains why this segment is currently unprofitable.

Over the long term, this segment should continue to grow at a rapid clip and its margins should steadily increase. As evidence of this segment’s margin potential, comparable companies at scale are nicely profitable.

• La Fourchette, since renamed TheFork, is essentially the same business as OpenTable, which had 41% EBITDA margins in its mature North America segment before it was acquired by Priceline.

• Viator is basically the OTA for attractions and charges healthy 20% commissions to its activities and tour partners. There is no reason it cannot have OTA-like margins in the 30%–40% range at scale.

• The vacation rentals business should also earn significant margins over time. HomeAway, before it was acquired by Expedia, had EBITDA margins around 16%, but it was investing heavily for growth. HomeAway’s CEO used to say it could have eBay-like margins eventually. eBay has EBITDA margins near 35% and its auction business has EBITDA margins near 50%.

TripAdvisor could generate profits in this segment right now, but it would require them to stop investing for future growth. Kaufer made the following comments in regards to the attractions business, which he believes is going to be a $1 billion revenue business by itself over time:

“How would we as stewards of this company look at the opportunity in the attractions category? And we say, wow, fantastic, could turn it profitable tomorrow by a lot if we wanted to but why would we choose to do that when the growth opportunity is so tremendous? We feel so optimistic about this category because it's so big and we're the ones that can bring the demand and the supply together, we just need more inventory. With more inventory, we can both grow our organic demand and buy more demand, and then we become known as the aggregator of, hey, what do I want to do when I get there. And that feeds upon itself, we get more supply, we get more demand, bingo, and we not only retain our current leadership position in that category, but we become unassailable over years. And then, as with any business like this, margins flow really nicely out of it because as traffic builds, especially if it's repeat traffic or organically grown traffic or app traffic, customer acquisition is nil and you've got a beautiful margin machine here.”

Valuation

I value TripAdvisor using a discounted cash flow model. Some of the key assumptions are:

• Hotel segment revenue grows at mid-teen rates from 2017 through 2022 before downshifting to 10% for a few years and then a low-single digit rate thereafter. I think this is conservative because it assumes a slow improvement in revenue per hotel shopper and modest 8% hotel shopper growth. Hotel shoppers grew 15% in 2015 before the steep fall in revenue per hotel shopper, which caused the company to acquire less traffic. As revenue per hotel shopper improves, the company will acquire more traffic. 8% hotel shopper growth does not account for this.

• Hotel segment adjusted EBITDA margins gradually recover to the low 40%s over the next five years where they plateau. I believe that is conservative because Hotel margins were 47% in 2012 when the company had the “pop up” revenue model, which should prove inferior to Instant Booking, and was incurring investment losses in China. Further, margins should not plateau anytime soon because incremental margins are much higher.

• Non-Hotel segment revenue, which has grown 18% year-to-date, grows at a low double-digit rate through 2022 before growing at 10% for a few years and finally downshifting to low single-digits. I think this is conservative because this segment has a very large market opportunity, especially the attractions business.

• Non-Hotel segment adjusted EBITDA margins gradually improve and plateau at 25% by 2021. I think that is conservative because the economic characteristics of these businesses, as evidenced by the margins of comparable companies, should lead to at least 30%–35% segment margins at scale.

• I deduct stock-based compensation, depreciation and amortization, and taxes and assume depreciation equals capital expenditures over the long term. I use a 35% tax rate to be conservative, which is higher than the effective tax rate in the mid-20%s.

• I capitalize unlevered free cash flow in the terminal year at a 4% rate and apply a 10% discount rate.

This DCF indicates TripAdvisor is worth about $14.8 billion today and the equity is worth about $16.0 billion after adding the net cash and investments on the balance sheet and assuming all options are exercised. That is a per-share valuation of $103, which is about 73% higher than the current $62 per share stock price. Between long-term business growth and the closing of the discount, I expect TRIP to compound at a mid-teens annual growth rate or better over the long-term.

Catalysts and Other Considerations

Last quarter, the second-quarter of 2016, is likely to be the “peak headwind” period for both the Hotel and Non-Hotel segments. Management expects both segments to show improving results in the second quarter of 2016 and beyond.

• For Hotel, the company began lapping the Instant Booking launch in the U.S. and U.K. last Fall. Second-quarter 2016 revenue growth in North America accelerated to 6.6%, which supports the point that Instant Booking is starting to work in its oldest markets. Conversion rates, repeat booking rates, and vaulted credit cards are increasing. Accelerating growth in the oldest markets should more than offset the initial headwinds from the more recently launched smaller markets. In addition, revenue growth should accelerate in the second half due to revenue recognition timing. The company recognizes revenue upon the completion of the stay with Instant Booking whereas it recognizes it upon the click with metasearch. So some Instant Booking bookings that occurred in the first half of this year were not recognized as revenue and instead were deferred until the second half, and specifically the third quarter, when many people go on vacation.

• For Non-Hotel, investment spending should begin to lighten up in the second half of this year. In addition, there is a similar revenue recognition issue at play. The company’s vacation rental business recently transitioned from a subscription model to a free-to-list model. As a result, vacation rental revenue is now largely recognized on stay instead of when the booking is made. That means first-half bookings for second-half stays were not recognized as revenue in the first half and were instead deferred to the second half. Again, much of that revenue will show up in the third quarter when many go on vacation, which will cause an accelerating growth rate.

• Since TripAdvisor’s revenue per hotel shopper has fallen so much over the last 18 months, the company does not acquire as much traffic from Google and other traffic acquisition channels. This is because traffic acquisition spending is based on an ROI calculation driven by revenue per hotel shopper. Lower revenue per hotel shopper means lower ROIs, which means lower traffic acquisition spending. This has manifested itself in hotel shopper growth, which slowed to just 3.0% in the second-quarter of 2016 from 15.0% growth in 2015. However, this works both ways. As revenue per hotel shopper recovers, the company will increasingly acquire more hotel shopper traffic. I think about this in terms of price and volume. In TripAdvisor’s case, volume and price fall together, so the negative impact is acute right now. When price recovers, volumes will recover too, leading to a super-charged recovery. This phenomenon will act like a coiled spring.

• TripAdvisor has historically donated 2.0% of its prior year’s OIBA (a measure of earnings) to the TripAdvisor Charitable Foundation. In December of 2015, TripAdvisor changed course and decided to settle its obligations to the foundation in one lump sum rather than continue the annual donation. TripAdvisor paid $67 million in stock to the foundation “in full satisfaction of all future annual contribution obligations under the Pledge Agreement between the Company and the Foundation.” The Pledge Agreement allowed this based on paying a multiple of 8 times the prior year’s contribution. Why would TripAdvisor do this? If the company thought its earnings were likely to increase substantially in future years, it would be rational to settle the obligation up front and pay an amount that is calculated off depressed earnings. I believe the company paid in stock rather than cash because the stock was in the mid-$80’s at the time, the foundation would benefit from TRIP’s long-term appreciation, and the company could always buy back the 801,042 issues shares in the open market later when the stock price was lower. Management almost certainly knew the stock price would trade lower given they were accelerating the global rollout of Instant Booking, which has a sharply negative impact on results in the near term. Sure enough, TripAdvisor started repurchasing shares in the first half of 2016 at depressed prices after the stock fell sharply. The company has bought back 208,137 shares at an average price of $59.79 through June.

Bearish Points

When investors first look at TripAdvisor, they see:

• An expensive-looking stock trading at about 37x trailing earnings and about 39x consensus 2016 earnings.

• A company with rapidly falling revenue growth. Revenues grew 31.9% in 2014, 19.7% in 2015, and actually fell in 2Q by 3.5%. Revenue is expected to increase by only 2.2% this year.

• A company with EBITDA margins that have consistently fallen over the last several years from over 50% in 2010 to just 18.7% last quarter.

Clearly, this does not look like an interesting investment opportunity. One would have to spend a lot of time understanding the business and why revenues and margins have been behaving this way to understand why the current weak patch is not only temporary, but has set up a great investment opportunity. Few are motivated to do that when the initial metrics look so unpromising.

More informed investors who understand TripAdvisor well may have one or more of the following concerns:

• TripAdvisor had to give Priceline enormous concessions to entice it to join the Instant Booking platform. As a result, TripAdvisor gave up all the potential upside of Instant Booking to Priceline.

That is almost certainly not true. TripAdvisor agreed to let Priceline pay it an unusually low commission rate, which seems to be about 10%, but this is a temporary promotional rate. Industry players have made comments suggesting it is only a matter of time before Expedia joins the platform. Once Expedia joins along with more and more independent hotels, Priceline is unlikely to continue paying a very low promotional rate. They could bid 10% if they prefer, but they will be outbid by Expedia or the hotels and will lose market share on TripAdvisor. Big OTAs like Booking.com are comfortable bidding to breakeven in these channels. Given they are making “upper teens” from the hotels, so there is no reason they would not pay the vast majority of that to acquire those bookings through Instant Booking.

• Instant Booking will not be an improvement over metasearch because Instant Booking conversion rates are worse and will not improve.

This point of view stands in contrast to the evidence to date. Initial Instant Booking conversion rates are lower than metasearch, but they are improving every quarter in the most mature markets. In some use cases, Instant Booking conversion is already higher than metasearch. The metrics are all trending in the right direction.

•  Instant Booking will evolve into an auction dynamic, but commission rates will trend lower, not higher.

In any competitive auction, the highest bidder wins. In this case, the highest bidder will “win” the Instant Booking partnership for a given hotel. Metasearch is a competitive auction and there is no reason Instant Booking will not move in that direction. Metasearch leads and Instant Booking completed bookings remain valuable revenue sources for the hotels and OTAs. Commission rates are likely to drift higher in a competitive auction when the auction participants generally pay higher commissions in comparable marketing channels. The hotels and OTAs view TripAdvisor as a sales channel and will act accordingly.

• Users will always consider TripAdvisor a research site, not a booking site, and the company will be unsuccessful changing users’ behavior.

It is true TripAdvisor is primarily considered a research site today. But Instant Booking is clearly gaining traction. This is happening because it is a lower friction user experience than metasearch, where users click off to one or more alternative booking sites. The data on conversion rates, repeat booking rates, and vaulted credit cards suggests users like the Instant Booking experience.

Conclusion

TripAdvisor’s stock is an excellent long-term investment at $62 per share because the company’s profitability is temporarily depressed. As Instant Booking and the Non-Hotel investments pay off over time, the company’s earnings and cash flows will improve. The current stock price does not appear to reflect the extent of these improvements. Most importantly, it should be very difficult to lose money in this investment given TripAdvisor’s powerful content-driven brand and enormous site traffic, which makes it an important revenue source for its partners, and the big discount at which the stock is trading.

Appendix A: Revenue Per Hotel Shopper by Device

TripAdvisor does not explicitly disclose much information about how revenue per hotel shopper differs on the phone versus on tablet and desktop. However, management does disclose interesting facts on an ad hoc basis. Two recent facts were:

• In 2Q 2016, phone hotel shoppers made up 33% of total hotel shoppers. Therefore, tablet/desktop hotel shoppers made up 67%.

• Phone hotel shoppers monetized at roughly 33% of the rate of tablet/desktop users. This is an improvement from 10% just a few years ago.

With these two data points, I backed into revenue per hotel shopper on phone versus tablet/desktop. We already know overall revenue per hotel shopper was $0.48 last quarter. So the algebra is simply: (33% hotel shopper mix on phone * X) + (67% hotel shopper mix on tablet/desktop * 3X) = $0.48. After solving for X, it is clear that revenue per hotel shopper was approximately $0.205 on the phone and approximately $0.615 on tablet/desktop last quarter.

It is also interesting to look at how that has been evolving. In just the prior quarter, phone made up 30% of hotel shoppers and management said phone monetized at about 30% the rate of tablet/desktop users. Using the same algebra, revenue per hotel shopper was $0.17 on phone and $0.58 on tablet/desktop. Revenue per hotel shopper on the phone improved from $0.17 to almost $0.21 – about 16% – in only three months. Tablet/desktop has also improved from $0.58 to $0.62 – about 7% in those three months.

If I extrapolate the recent trends in phone mix, phone monetization, and tablet/desktop monetization, it seems like overall revenue per hotel shopper will continue to improve at a good pace. $0.60 overall revenue per hotel shopper seems attainable in 2017 and $0.85 seems possible by 2019. However, this is only an extrapolation of recent trends, and I incorporate much more conservative assumptions into my DCF valuation.

Disclosure: Long TRIP.

The author's fund is long TRIP. This is not a recommendation to buy or sell any security. All information in this write-up may be completely wrong. Please rely on your own research.

Catalyst

The worst of the Instant Booking rollout is now in the past, and 2H 2016 and beyond are likely to show improving revenue and margins in both segments.

Expedia joins the Instant Booking platform. 

Revenue per hotel shopper exceeds expectations, which allows TripAdvisor to spend more acquiring traffic, which accelerates hotel shopper growth rates.