Description

Yelp Inc.

Summary

Yelp is a fairly hated name that has more recently been viewed as a broken company with major structural issues, despite strong top-line growth, expanding margins, strong FCF gen, and a cash-rich/debt-free balance sheet. Obviously, there are reasons why shares have sold off and most of the sell-side is negative on the company, but I view the issues as a combination of being overblown, a misunderstanding / incorrect read of the underlying data/KPIs, and missing the bigger, longer-term picture. Ultimately, the significant operating leverage of the business / high incremental margins of adding new advertisers presents a significant earnings growth opportunity over several years even without heroic assumptions on top-line re-acceleration. Combined with a recently more active buyback, an activist making sure shareholder interests are properly represented, and serious takeout potential, YELP could easily generate >20% compounded shareholder returns over a multi-year period with potentially more outsized returns over the next 2 years or so.

Company Overview

YELP is an internet property (website + mobile app) focused on local business reviews and information. YELP is basically the modern-day, digital version of the Yellow Pages. Yelp has built - and continues to build - its database of local businesses and services from a combination of crowd-sourcing (anyone can establish a page for a given business) and “claimed” businesses (wherein the business owner either claims or creates the page for his/her business and uses it to publish information).

YELP is a 2-sided platform, wherein value is provided to both potential consumers visiting the site, as well as business owners who have representation via a YELP page. For the consumer, the vast inventory of user ratings and reviews provide context and useful information on top of general listings for local business/service discovery. Further, the vast suppository of data creates a network effect, wherein the value of Yelp’s data increases as more ratings and reviews are added, providing a more accurate and useful experience for future consumers to continue using the platform for local business discovery. On the flip side, for the business owner who claims or creates a page, - which they can do at no cost - they are afforded the ability to update key information (e.g., hours or operation), “enhance” the page’s content (e.g., menus, prices, pictures/videos, coupons/promotions), advertise (pay-per-click search advertising), communicate (direct messaging, quote/pricing requests, response to reviews), and potentially transact with customers.

Vast majority (>90%) of YELP’s rev comes from advertising. This is largely pay-per-click advertising via both search (e.g., sponsored search) and embedded on business pages (e.g., other Italian restaurants near the Italian restaurant you are currently looking at). Depending on the package that’s purchased, business owners may be able to eliminate competitor ads from their pages, receive varying levels of account support, or add specialized content (like videos, slideshows, etc.). Paying business owners are referred to as paid advertising accounts (PAAs), which are sourced via 3 channels: 1) national salesforce (multi-site businesses or franchises); 2) local salesforce (SMB); and 3) self-serve on the Yelp site.

Additionally, YELP also generates revenue from transactions executed via Yelp. This used to be nearly all from Yelp’s Eat24 business - a GRUB competitor - which was eventually sold to GRUB; post-sale, transaction rev is largely comprised of Yelp’s agreement with GRUB wherein Yelp receives a fee from all food orders within GRUB’s network that originate from Yelp pages (virtually pure profit). There are also other transactions that occur on Yelp, such as ordering flowers, booking a spa, etc. Finally, Yelp has a small, but fast-growing subscription rev channel from other services, such as Reservations (OpenTable competitor – restaurant reservations from Yelp page), NoWait (waiting list for restaurant seating), and Yelp Wi-Fi (Wi-Fi-based marketing for consumers connecting to the Wi-Fi network of the local establishment, e.g., getting a promotion when connecting to the Wi-Fi network at a Starbucks).

Yelp’s CEO is also its founder - Jeremy Stoppelman, who takes a $1 salary and otherwise is totally compensated by stock. Stoppelman owns ~3% of the company. In a buyout scenario (discussed more below), Stoppelman’s payout would actually be pretty low, given it would only consist of equity acceleration, which is ~$2.4M for him, though obviously he’d also get the benefit of cashing out 3% of the company. Regardless, in what is potentially a slight negative, I don’t get the feeling that he’s as motivated by money as one might hope from the perspective of a buyout (i.e., he might have a higher bar for selling because YELP is his baby). All that said, in a very pro-shareholder move, YELP eliminated its dual class structure a few years back, effectively removing the total voting control Stoppelman had over the company. A valid complaint against the compensation structure is the lack of a TSR component or other hard metrics that would better align compensation with performance (both operational and share price)

Why The Opportunity Exists / Variant Perception / What the Bears Are Missing

I’m putting this section upfront, given the core to the thesis is that the evidence for the bear case is thinner than they’d have you believe and, when stripping away many of the bear points, it’s hard to ignore the growth, margin expansion, and capital allocation upside of Yelp at its current valuation.

The opportunity exists because YELP sold off significantly after 3Q18 earnings, where they lowered 2018 rev guidance and showed a flat total PAA metric q/q (see more on this below). This was the coda to what was becoming an increasingly “hated” name.

Judging how “hated” YELP is obviously isn’t an exact science, but 23 of 31 sell-side analysts rate it neutral or sell, including 10 sell-side downgrades (from 9 brokers) since December 2017 (against 2 upgrades, though one of the upgrades was quickly followed by a downgrade by the same analyst 3 months later). Short interest is ~12% of the float and just under a week in days-to-cover. Judging by the reasons given by the plethora of downgrades and non-buy-rated analysts, the bear case includes some combination of the below points. However, several of these items are based on flimsy (in my view) evidence, which (to be fair) doesn’t mean the point is wrong, just that it isn’t well-supported.

Stalling PAAs

Paid advertising accounts (PAAs) are, essentially, YELP’s customer count. Given the open-ended growth story, despite some under-the-surface seasonality to the business, PAAs are expected to grow on a q/q basis. A q/q decline signifies a problem. More to the point, no one argues that Yelp’s business features high churn, but the bull-case depends on the idea that the ability to grow repeat/recurring customers + the significant under-penetration of the addressable market, should yield enough customer additions to more than offset normal churn. If that ceases to be the case, either the platform isn’t valuable enough to keep customers or Yelp has a problem adding them (or both).

Going back to 2014 (just because that’s how far back I built my model), total PAAs have increased every qtr (until 3Q18) anywhere from 30K to 170K, though typically in the 60-80K range. In an effort to further monetize the addressable market, Yelp went to no-term contracts across the board starting in 2018 (previously had tested it in select cases/markets for prior 2 years), figuring that Yelp’s most sophisticated customers pay month to month anyway and recognize the ROI, so removing barriers to trying Yelp (such as a minimum term contract to start) will bring more PAAs onboard to try the product and realize its value.

Indeed, PAA additions spiked in 1H18, with additions of 140K and 170K in 1Q and 2Q, respectively, effectively double normal levels. However, Q3 PAAs were flat q/q and Q4 was actually down 30K. The reaction was predictable - the shift to no-term caused a spike in first-time users, but it was more of a gimmick and didn’t stick. The problem is, this simplistic view ignores the different impact of no-term on - as well as the different characteristics of - Yelp’s core, repeat PAAs, vs. those “just giving Yelp a try”, who may or may not ultimately become a repeat PAA.

Yelp provides data on the % of PAAs in a given qtr that have used Yelp in the prior 12 months (repeat rate), thus providing the split between “real” customers and those still within the customer acq process. Given that repeat PAAs are typically ~75% of the total, less likely to be on promo, and more likely to be a bigger spender, it’s likely that the % of rev represented by repeat PAAs is far greater than its ~75% share of PAAs (though mgmt doesn’t break it out). In fact, when repeat and new PAA growth rates started to diverge in 2018, Yelp’s rev much more closely tracked growth in repeat PAAs vs. total.

To this end, splitting the PAA accounts tells a different story. Going back to 2014 as before, prior to the 2018 change to no-term, typical recurring PAA additions were 40-60K/qtr, as low as 4K (an outlier in 1Q17) and as high as 84K. While the growth rates have decelerated as the base has grown, through 2017 recurring PAAs were still growing at teens rates y/y. The change to no-term spiked the new PAAs per qtr from ~25-35K to >50K in every qtr of 2018. While the new PAAs each qtr became more volatile and choppy, YELP continued converting enough (despite the non-structural issues in 3Q, discussed below), so as to keep the recurring customer count growth on an absolute basis in each qtr, as well as growing at teens rates y/y, through the entire year.

The volatility of the new PAAs shouldn’t be a surprise either - the lack of a gating factor to try YELP introduces significant seasonality into new PAAs (seasonal businesses more likely to try Yelp on a promo with no long-term commitment). This shouldn’t be a problem to the overall business as long as YELP continues to convert enough “new triers” to recurring customers to keep the “real” customer count growing at a health rate (as it is). It also highlights the stark difference between the value of total vs. recurring PAAs, which was never something that mattered as much previously given the lack of meaningful divergence between the 2 metrics.

Bottom line - the PAAs that matter - recurring, higher-paying PAAs comprising vast majority of revenue - are NOT stalling, but in fact growing at still healthy rates. What’s changed is the volatility of the in-qtr acquisition channel, which has a far more muted impact on qtr-to-qtr growth assuming no sustained downward trend in new PAAs.

Declining User Engagement

I’ve seen 2 data-points used as evidence for this, both of which are flimsy: 1) declining unique web browser visitors (data provided by Yelp); and 2) declining app downloads (3rd-party data). What’s conveniently ignored is Yelp’s engagement metric of choice - unique app devices.

Yelp mgmt has - for a while now - strategically pushed user engagement via mobile app as opposed to web browser (whether desktop or mobile-based). Very simply, the app is stickier, better for engagement, and more shielded from competition by Google and FB. In fact, Yelp mgmt has noted on numerous occasions that vast majority (~75%) of total page views come from the app. Further, Yelp has actively and explicitly pushed users in this direction (if you’ve ever gone to Yelp on a mobile browser, you’ll understand - the “open in app” button and the reduced functionality of the web browser vs. the app are not so subtle hints).

Thus, declines in desktop and mobile browser visits is not surprising (a kind of by design). Conversely, unique app devices (i.e., unique devices running the Yelp mobile app) continue to grow at teens rates quite consistently.

Moreover, 3rd-party data around app downloads is quite volatile in a given period, given the dependence new smartphone design and upgrade cycles. In other words, downloads data can fluctuate - or even decline - if fewer smartphones are upgraded (new phone = new download), whereas the amount of unique devices running the app would be a more “real” reflection of ongoing users. As an example, if 30% of Yelp app users upgrade their phones in a given qtr, app downloads data would show a huge spike in Yelp app downloads, but unique devices might stay the same. On the flip side, if the following year only 5% of app users upgraded their phone, unique app device growth could increase 15% but would look like a y/y deceleration or decline in amount of downloads.

Finally, other metrics of user engagement continue to be solid - cumulative reviews continue to grow at ~20% rates y/y and MSD% q/q, while mgmt has noted that internal engagement metrics have been quite healthy. Bottom-line, claiming that user engagement is declining is a significant reach, given the most prominent metric - unique app devices - continues to grow at healthy rates, while browser-based page views are declining by design.

Price Compression

This is a derivative of the PAA issue - as a “proof positive” that Google and FB are compressing CPC prices vs. YELP, some sell-siders have pointed to average rev per PAA, which is dividing advertising revs by PAAs in a given qtr. Of course, given the dichotomy between recurring and new PAAs, in addition to the spike from going to no-term, the declines over the last 4 qtrs (down 4-8% in each qtr; down 6% for the year) are largely (if not entirely) due to the spike in new PAAs from going to no-term. More to the point, its not at all indicative of price compression for standard relationships. To wit, under the assumption that repeat PAAs comprise >80% of rev (maybe >90%), given the similarity in growth rates between recurring PAAs and advertising revs, it looks like underlying pricing is fairly stable.

Declining Visibility / Increasing Volatility

This is a legitimate complaint, but not one that impacts the longer-term view. In fact, this is what effectively creates the time-arbitrage for owning YELP. By eliminating contract terms, revenue is largely recognized in-the-moment (vs. coming off the balance sheet) and the volatility in new PAAs in a given qtr increases qtr-to-qtr volatility as well as visibility, at least until mgmt gets a better handle on appropriate conversion rate assumptions in a no-term world. That said, as noted earlier, as long as - in the aggregate - the new PAAs are converting at high-enough rates to sustain healthy growth in recurring PAAs, qtr-to-qtr volatility in PAA additions won’t make a difference in the long-run. In the meantime, it appears to be causing the market to ignore the long-term story for the near-term qtr-to-qtr beat/miss, which may swing a bit but all in an upward direction in the aggregate.

Competing vs. Google and Facebook

Probably the biggest bear point and one that will never die. Plain and simple, digital advertising is dominated by Google and Facebook and both have somewhat of an interest or a play in local business discovery. Further, there have been complaints by some that Yelp offers inferior ROI or effective cost per click or cost per lead.

Ultimately the anecdotal comments from a few internet sources is difficult to trust, given the multiple factors that can impact the advertising success of ABC local business. Further, as YELP notes, its fastest growing segment and most reliable customers - large, multi-site/franchise businesses and (more recently) ad agencies - are its most sophisticated customers as well, so it’s hard to argue against the ROI of Yelp from that perspective (though I admit, that’s hardly a proof-positive of anything).

More importantly, the pushback I have to the Google/Facebook argument is 1) Yelp’s shift to app-focused engagement, which avoids having to go through GOOG  or FB to get to YELP; 2) the ubiquity of YELP as a must-have page for a business, given the plethora of user reviews/ratings (i.e. even if they don’t like it, every business has a Yelp page; its table-stakes for digital marketing); 3) Yelp’s large existing database of reviews and targeted user data; and 4) YELP users represent a very targeted potential lead, vs. someone doing a google search or seeing something on their FB timeline. Someone on Yelp looking for Italian restaurants is doing something very deliberate and represents a very targeted lead in multiple ways (i.e. it can have several relevant derivatives to other areas of YELP), much more so than what someone may see on Google Maps or on a Facebook timeline.

Ultimately, as noted below, Yelp’s problem isn’t with engagement - it has a tremendous amount of unsold “click inventory” and engagement can run flat and not impact their ability to grow, which is mostly dependent on better monetization of existing click inventory.

Recent History / Background

Yelp’s primary challenge has been, and continues to be, (in)efficient monetization. While many may point to varying focal points of the business, ultimately YELP has more than enough user engagement from the perspective that YELP monetizes a very small % of its ad inventory. Said another way, YELP is only selling a small % of the total amount of clicks happening daily on the platform (to this end, mgmt has noted that if engagement tripled, it would barely impact rev). Even if user engagement went totally flat or down (which its not), the primary growth determinant would be monetizing the existing click inventory, vs. the impact of further growth of that inventory.

This important as context because it is often overlooked that YELP has been in a constant state of experimentation on increasing monetization. Despite this, YELP has built a ~$1B business, is sitting on a pile of cash with no debt, generates >$100M FCF annually, has a significant runway on margin expansion given the high GM% (>90%) and minimal incremental cost of adding PAAs. Of course, therein lies the key - given the tremendous scale already built into the platform and the extremely low marginal cost of adding PAAs, there is almost no downside and tremendous upside from YELP trying new ways to bring advertisers to the platform. With the large excess of ad inventory, additional PAA dollars drop almost entirely to the bottom-line. As an example, YELP’s under-monetization can be viewed in the context of Yelp’s ~16M business listings, of which only ~4M are actively claimed (~25%; actively claimed = claimed page that’s actively managed/kept up to date) and only ~192K are PAAs (~5% of claimed businesses). Only 5% of pages that are actively claimed by the business owner are engaging in some form of advertising on Yelp.

The problem has been that the markets are unforgiving and, at a headline level, YELP has decelerating revs and certain of its key performance indicators (KPIs) in decline (though not the ones of import), on top of calling Google and Facebook competitors. As a result, Yelp hasn’t gotten the benefit of the doubt and has gotten stuck in a very short-termism prism, which doesn’t work well when the company is experimenting on improving monetization. Experiments cause surprises and the market doesn’t like surprises.

This is what’s been in effect over the last 12 months or so. Yelp mgmt’s thesis is that the ROI of using Yelp is clear to those who use it most and in volume. Indeed, mgmt notes that its strongest customer channel in terms of growth is also its largest customers - national customers (i.e. multi-location or franchise businesses), who are the most sophisticated ad spenders and evaluators of advertising ROI. The national channel grew 30% in 2018 (vs. 18% apple-to-apples organic for the total business), accelerating over 2017, accelerating in each qtr of the year, and now accounting for ~20-25% of rev. Thus, if the barriers to using Yelp are reduced and more potential advertisers can just “give it a try”, it will result in more PAAs as exposure to the ROI is more meaningfully exposed. Moreover, with little cost to Yelp from increasing the amount of advertisers “trying” the platform vs. the great benefit from each additional user, the risk/reward is asymmetrical.

Yelp has attacked this from 2 different angles over the past 12-24 months:

1) diversifying PAA acquisition channels away from local salesforce - local sales, though important, is also the least efficient. The targets are not necessarily huge spenders and its very labor intensive. Thus, mgmt has put more investment into national sales channels - both national sales force for higher-spending multi-site businesses as well as partnering with (and providing tools for) ad agencies to get more bang for their buck. As part of this Yelp has been continuously developing greater tools for these sophisticated advertisers that do a better job on attribution and analytics so as to improve the ROI and visibility for advertisers. Additionally - and more importantly - Yelp has invested in making its self-serve channel as powerful and easy-to-use as possible. Self-serve is - unsurprisingly - the most efficient/low-cost way to onboard new PAAs. Here too, mgmt continuous to develop new offerings that prove sticky and easy to purchase for self-serve customers to both onboard and retain them as recurring PAAs; and

2) eliminating hurdles to using Yelp, most notably eliminating minimum contract terms. Previously, Yelp advertising packages came with a minimum term (6 months, a year, etc.). In an effort to remove “try it out” friction, Yelp has gone completely “no-term”. From  Yelp’s perspective, the risk is low given Yelp’s recurring customers - while starting at a minimum-term contract, eventually moved to month-to-month once the original term was up (i.e., minimum terms were typically enforced on new customers; if you stayed on past minimum term you were charged monthly). Thus, regular Yelp customers weren’t being term-enforced anyway and the minimum term was doing more to prevent adding new PAAs that keeping existing ones.

The first item above has helped Yelp start to drive more significant operating leverage as the business is slowly reducing its reliance on local salesforce to drive revenue. The transition is not without speed bumps, but, by and large, Yelp is still growing recurring PAAs at a nice clip, while significantly expanding margins. The second item has also helped to increase recurring PAAs. In fact, the non-recurring PAAs in a given qtr (i.e., first-time tryers) has spiked since Yelp went to no-term - first-time advertisers used to be ~25-35K / qtr; since going no-term, that number has spiked >50K /qtr. While conversion rate from first-timers to recurring is low, it's still early days and the lower rate doesn’t really matter as long as recurring continue to grow at double-digit+ rates given the minimal (if any) cost of having more people trying the platform.

The downside is that 1) the PAA number has now become far more volatile given the non-recurring piece can swing a bit more as the barrier to try Yelp is very low (i.e., likely introduces significant more seasonality into non-recurring PAA metrics); and 2) less visibility beyond a given qtr given the lack of minimum term on first-timers (i.e., Yelp has a good handle on spending trends of recurring customers; the minimum term for first-timers helped alleviate potential visibility concerns  on how long a first-timer would stick around).

In this context, Yelp took a tremendous dive following 3Q18 earnings, where total PAAs were flat q/q (though +25% y/y due to the spike in first-timers and continued, consistent growth in recurring at +16% y/y and +4% q/q). Mgmt lowered 2018 guidance as well, noting several speed bumps related to the transition to no-term, though none of them being real, structural problems. Specifically, 1) timing lag on replacing sales reps who had a hard time transitioning to no-term with productivity ramp of new sales reps; 2) a back-end system issue that was resolved intra-qtr but had an impact on PAA onboarding in the qtr; 3) a promo change that has better LTV characteristics, but impact PAA adds in the qtr; and 4) abnormally low success rate on outbound sales calls. When pushed on this, mgmt noted that sales reps had a hard time getting potential PAAs on the phone in Q3 and that this was a strange anomaly, but also something that will be rectified by getting more aggressive on alternative communication methods (email, text, performance marketing, etc.), as well as the development of the self-serve platform (wherein local sales becomes less of a driver in any given qtr).

While the lowered 2018 guide was certainly a negative surprise and the PAA onboarding issues in the qtr weren’t realistically going to be totally waved away by the market, the magnitude of Yelp’s fall was significant, with the sell-side pile-ons (downgrades, significantly lowered out-year estimates, etc.) overshooting the reality of the situation. Ultimately, Yelp is still growing revenue around double-digit levels, with significant operating leverage over a multi-year period, significant margin potential remaining, solid underlying key performance indicators (at least that ones that matter, such as recurring PAAs still growing nicely, unique app devices growing nicely, etc.) and buying back shares, while sporting no debt and a significant net cash position (~25% of mkt cap).

During Q4, an activist got involved (SQN), pushing for  Yelp to do more on capital return, refresh the board, take action to improve margins and growth, and take a look at strategic alternatives. In response, in conjunction with Q4 earnings (where Yelp beat the qtr), Yelp announced 3 new board members, aggressive share repurchases, and a 5-year plan to expand EBIT% from ~19% in 2018 to 30-35% in 2023, on top of a mid-teens earnings CAGR through the same period, starting with +8-10% in 2019, but accelerating in 2H as investments and growth initiatives take greater hold.

While shares have recovered somewhat year-to-date, the response since earnings has been muted and sell-siders haven’t come on board (indeed a few more downgrades happened). The view is that Yelp’s plan is overly ambitious regarding revenue reacceleration. While that may be true, investors and the market continue to otherwise ignore the tremendous earnings growth that will come from even moderate growth, combined with the margin expansion (which mgmt has a much greater track record of executing and is more in-their-hands than rev reacceleration) and aggressive buyback activity over a multi-year timeframe. What’s more, valuation remains pedestrian (particularly relative to Yelp’s earnings growth, comps, Yelp’s own trading history, and takeout valuations).

Positives

Under-appreciated Growth Profile and Opportunities

YELP is right in the middle of the offline-to-online trend, which many internet companies have been thriving off of (IAC is a good example). The growth opportunity is fairly open-ended given the low penetration of online discovery in a still-largely offline local business market.

At a basic level, as noted above, YELP’s penetration rates of its own business owner users is quite low - only ~5% of active claimed businesses are currently advertising - in some form - on YELP. While unlikely to get to 100%, there is substantial opportunity to increase the penetration of PAAs within YELP’s existing active users.

Secondarily, YELP’s most important KPIs continue to trend positive - from a demand perspective (i.e. advertising demand), claimed local businesses continue to grow ~4-5% every qtr (20% or so y/y), repeat/recurring PAAs (~75% of total PAAs) have never declined q/q and continue to grow double-digits y/y. Paying Ad Locations (new metric meant to capture growth in locations, which is greater than PAAs, given growth in national/multi-site PAAs) are similarly growing both q/q and y/y (double-digits). From a supply perspective (i.e., user engagement), unique app devices (i.e., unique devices with the Yelp mobile app; >75% of page views) continues to grow double-digits, while cumulative reviews continue to increase at a steady pace.

More specifically, beyond “standard” aspects of improving PAA onboarding (improving the product, ease-of-use, more products/features, improving advertiser analytics / attribution data, etc.), YELP has also introduced new facets of monetization, such as request-a-quote (RaQ), where users can request a service quote for a given project (think home-and-local services, like a new bathroom or fixing the roof) and YELP can provide quotes from several other service providers, with the quote requests delivered to those service providers as very targeted leads (early days of monetization here). Another example would be Yelp Verified License, where Yelp charges a low, fixed monthly fee to verify - and prominently display such verification - that he service provider has all the necessary and required professional licenses to perform advertised services.

Beyond simple advertising, transaction revenue is also largely under-monetized, with Yelp’s GRUB relationship dominating the segment currently, but with potential expansion into other areas (spa/health, flowers, etc.). Additionally, Yelp’s subscription services, such as Reservations (think Opentable) and NoWait (now called Waitlist) are growing nicely have significant potential as both a way to keep restaurant PAAs engaged and active while also monetizing its assets in alternative, profitable ways.

These positives are not meant to justify outsized growth rates, but rather to highlight the significant and tangible opportunity to continue growing at well above GDP rates and the low-bar for growing double-digits, consistently, for the foreseeable future. Put another way, the market environment/opportunity should serve as a tailwind, not a headwind for growth. While consensus appears to agree on the surface, the wide range of estimates (high standard deviation) and generally bearish qualitative view (23 of 31 analysts are hold or sell rated), indicate a lack of belief; certainly consensus out-year estimates prior to Q4 earnings were more explicitly indicative of this.

Significant Margin Expansion Potential

As noted above, there is significant operating leverage in the Yelp model, where added PAA revs drop almost entirely to the bottom-line. Further, mgmt’s push toward national and self-serve customer acquisition channels, and somewhat away from local salesforce, only serve to improve the margin mix. Mgmt’s plan to expand margins to 30-35%, from 19% in 2018, by 2023 don’t seem all that far-fetched, given the low R&D intensity, flattish G&A, more efficient sales & marketing usage - which mgmt has proven out multiple times already (excluding 2015 due to the effects of acquiring Eat24 in early 2015, sales/marketing % of rev dropped >300 bps in 1H15, 100 bps in 2016 and 180 bps in 2017; in 2018, due to investments, it only fell 10 bps, effectively re-investing a chunk of the 210 bps of COGS gains from selling Eat24).  

Aggressive Buybacks (at favorable valuations)

Following the post-Q3 sell-off, Yelp got aggressive on buybacks (aggressive at the right time too), repurchasing $115M (~3% of the company) in Q4 alone. Additionally, in November 2018, mgmt authorized a new $250M program (the $115M of buybacks used up prior authorization), followed by the announcement of another $250M added to the autho on Q4 earnings, with half of the $500M total autho expected to be put to use in 1H19. This will amount to ~9-10% of the company - at favorable valuations (at least relative to before 3Q18 earnings) in a 3 qtr period. Given significant FCF gen (likely >$180M in 2019; $115M in 2018 colored by 1-time tax payment related to sale of Eat24 that shaved off ~$29M) and the overcapitalized balance sheet (nearly 25% of mkt cap is cash), continued aggressive capital return can be sustained for the foreseeable future.

Pristine Balance Sheet / Significant FCF Generation

As noted above, Yelp generates FCF well in excess of $100M - strong conversion of EBITDA to FCF given no debt, negative cash conversion cycle, and no cash tax payments for the next several years. In addition Yelp has no debt and $778M in cash and investments, which is ~23% of the market cap

Potential Takeout Target

Since before their IPO, Yelp has been noted as a target for those seeking a differentiated advertising asset, with no shortage of potential suitors (AAPL, AMZN, BKNG, GOOG, FB, IAC, and MSFT, among others). Additionally, Google was rumored as looking at YELP in the ~$90/share range in 2014, with media reports in 2015 (WSJ in particular) indicating a a sales process was underway, with Yelp working with Goldman. Ultimately, media reports suggest that Stoppelman decided against a sale, which killed the possibility of a deal at the time (given the dual class share structure then in existence).

Today, the conditions for sale are much stronger, given the collapse of the dual share structure (minimizing Stoppelman’s ability to single-handedly kill a deal if he wanted), activist investor pushing for potentially looking at a strategic review, and potential synergies with several players (IAC/ANGI at the top of that list). Deal comps would suggest buyout prices well above $50, potentially much higher.

Pedestrian Valuation Relative to the Earnings Growth Outlook

Even if undershooting their targets, Yelp should be increasing EBITDA/FCF at >20% annually for the next several years, probably higher, with substantial contribution from both rev growth and margin expansion. Buybacks add some further gravy.

Yelp is trading at 9x 2020 EBITDA and 15x (11x ex-cash) 2020 P/FCF. TRIP - a somewhat comparable model, but in much worse shape - is trading >3 turns ahead of YELP. They trade at a discount to the broader internet comp group on almost every metric and to almost every individual comp. While some comps are growing faster, some are more profitable, and some better businesses (FB, GOOG), very few have the combination that YELP possess right now. Further, beyond internet names, companies with somewhat comparable financial models (mostly software companies) are valued well above YELP’s level, typically in the 20x range - give-or-take - on FCF.

Catalysts

Re-acceleration in PAAs

Given the continued healthy growth in repeat PAAs and the volatility in new PAAs in a given qtr, consensus/market assumptions of trendlining PAA counts flattish (or down) is a mistake that will become apparent within the next 2 qtrs. I believe shares will react quite positively to re-accelerating PAA adds on a q/q basis.

Margin expansion

Mgmt’s margin expansion plan calls for EBITDA% in 2023 of 30-35%, which implies an average of 270 bps of margin expansion annually to hit the midpoint. Obviously it may not be linear, but given expectations of 200-300 bps of expansion in 2019, it’s not exactly a back-end loaded goal. While consensus is modeling the guidance range on EBITDA% for 2019, consensus implies a deceleration to ~180 bps of expansion in 2020 and ~130 bps in 2021. Between rev and EBITDA%, the margin goal is much more within mgmt’s control and much more likely than re-accelerating revs. In fact, its likely they can hit these goals even without meaningful rev acceleration given the shift away from double-digit increases to local sales headcount. I believe the ability to expand margins meaningfully will become apparent far quicker than revenue reacceleration (probably within the next 2-3 qtrs) and will force up sell-side assumptions in the out years.

Rev re-acceleration

Kind of obvious, but to the extent that mgmt can make good on accelerating revenue up to mid-teens rates over the next few qtrs and years, market disbelief will be pressured. I don’t want to rely too heavily on this, because I think the stock works on HSD/LDD growth, vs. mid-teens, but obviously making good on targets as soon as possible will help re-rate shares.

Continued aggressive buyback activity

YELP has already bought back ~3% of the company and has promised another ~6% or so in the next 6 months, which keeps an underlying bid under the stock at good value. Continued capital returns (i.e. re-upping the program after it runs its course - hopefully in 2H19) will be a positive catalyst.

Sell-side upgrades

Upgrades move the stock, both on higher consensus estimates and better multiples just from the action. There are a lot of sell and hold ratings out there and few buy ratings. From the perspective of the impact of sell-side rating changes, the risk/reward is asymmetrical.

Accelerators

Further agitation from activist

SQN has pushed for changes and potentially a strategic review; Yelp has - in their own way - met many of those demands (refreshed board, new 5-year plan, capital return, etc.). With SQN actively involved, there is potential upside from further agitation (in a good way), including…

Running a sales process

As noted, Yelp has been in play before and there are any number of potential suitors. Certainly, IAC has been brought up and explicitly asked about it; for what its worth, they’ve very specifically not shot down the possibility of getting involved in some way. To this end…

Future GRUB-like partnerships

A Yelp-ANGI tie-up or partnership would make sense in many ways, as would similar tie-ups in multiple other businesses that YELP has (MB would be another interesting fit). Given YELP’s position in the local industry, there are many options for monetizing its “assets” in creative ways with other internet players.

Estimates / Valuation

As a quick note, YELP’s 2018 FCF was artificially low due to a ~$29M tax payment related to the Eat24 sale in 2017, so FCF will look like its spiking in 2019 due to the non-recurrence of the tax payment (indeed, YELP doesn’t pay cash taxes yet and I assume only $5M / year in cash taxes through the projection period. I need to double-check w/ mgmt when they become a cash tax payer, but I don’t believe it will be for a few years still.

To illustrate the power of the model, my projections actually assume mgmt undershoots growth expectations. While obviously falling short of growth targets would - at some point - damage mgmt credibility and impair the multiple, I purposely run the projection at lower revs as an illustration of the earnings growth power, as opposed to any view that they will fall meaningfully short.

I run estimates through 2023, though I’m going to use 2020 FCF as my primary year for valuing the company (see model below). It’s hard to pin down what’s the “right” multiple. To be as fair as possible, just assuming the 1-year forward multiple holds - currently ~18x on market cap / 14x ex-cash - that would yield PT of ~$50-51 (>30% upside) on my 2020 FCF/share of $2.85. Of course, outperformance should yield some multiple lift and YELP routinely traded around 20x FCF prior to the post-3Q18 blow-up. A return to that level would yield a PT in the high-50s/low-60s (>50%) upside. Takeout values on EV/Rev multiples at >4x would put shares into the $60-80 range (more than a double).

Catalyst

• Re-acceleration in PAAs
• Margin expansion
• Rev re-acceleration
• Continued aggressive buyback activity
• Sell-side upgrades
• Further agitation from activist
• Running a sales process
• Future GRUB-like partnerships