[Z – Zillow Group] On lead gen and iBuying
Over the last 5 years, Google has increasingly been crowding out organic results with paid ads and preferencing its own properties in search results, frustrating online businesses that feed under its canopy. TripAdvisor and Angie’s List used to brag that most of their traffic came from unpaid organic search while competitors with weaker brands were forced to pay for leads. But now, with Google pushing organic search results beneath layers of paid links, free search traffic has become a liability and conversely, aggregators like Booking who paid for demand seem to be on sounder footing.
There has always seemed to me a baffling discrepancy between the unique nature of these companies’ assets and cash flows they produce that is in part due to maladaptive revenue models and other missteps and in part due to the Google/Facebook extracting ever more value.Yelp, Angie’s List, LendingTree, TripAdvisor, and Trivago and other online lead gens have been around for anywhere between 15 and 25 years. They have massive addressable markets, carry well-known brands, host hundreds of millions of reviews, intermediate the discovery of millions of businesses/agents/hotels/loan offers every year….and yet, none with the exception of ANGI Homeservices1 is valued at more than $5bn and all produce meager/negative free cash flow margins (after stock comp).
To stem the siphoning of value at the top of the funnel, many lead gens have migrated down funnel, processing transactions, qualifying leads, and generally doing things that a massive horizontal aggregator is not set up to do. HomeAdvisor launched Instant Booking for service providers, TripAdvisor did the same for hotels. Angie’s List offered service quality guarantees for paying members in which it would either intercede to resolve any transaction issues or reimburse the consumer, and as part of its turnaround effort prior to being acquired by IAC, planned to offer a platform for service providers to manage payments, scheduling, and marketing. Yelp lets consumers to reserve seats and join waitlists at restaurants and, through its Request-A-Quote feature, allows homeowners to see how long it typically takes for a service provider to respond to quote inquiries. A few years ago, in an effort to improve lead conversion, LendingTree began playing a bigger role in directing consumers to particular lenders (as I explained in this LendingTree post from 2 years ago).
Zillow has faced similar top-of-funnel competitive challenges. Starting around 2015, evidence that Facebook was delivering better ROIs for real estate agents than Zillow began mounting.
Real estate agents were skeptical about the quality of online leads in general, but per this Inman survey from 2015, those who sourced through online channels found that Google, and especially Facebook, ads delivered higher returns than dedicated portals like Zillow:
“Even though the response rate is low, it is higher than Zillow, Trulia or realtor.com, and the cost is much lower,” said one broker-owner respondent. “It also allows us to market our listings to specific demographics without having a buyer agent’s face on our listings.”…
When their return on investment was rated individually by agents, all four listing portals included in the survey — Zillow, Trulia, realtor.com and Homes.com — received more ratings of “poor” than all their ratings of “mediocre,” “good” and “great” put together.
And yet, despite the ROI disparities, agents at the time were still twice as likely to purchase ads on listing portals than on Facebook.
But two years later, things flipped. A report from ad research firm Borrell Associates suggested that real estate agents and brokers were more likely to buy ads on social media than on any other online channel2. 87% maintained a social media presence; of those, 7/10 paid to promote a post and 6/10 purchased a targeted ad.
From the Inman article:
“Overall, agents are spending $9 of every $10 on digital advertising, and more and more of the spending is going to Facebook and LinkedIn,” Borrell reported. It forecasted a large migration from “pay-for-listings sites” — which is Borrell’s term for listing portals — “to more sophisticated forms of targeted digital advertising, in both display (banners) and particularly in social media.”
Facebook’s competitive impact was ambiguously expressed in the Zillow’s financials – on the one hand, revenue from Premier Agent (the part of Zillow’s business ) was growing ~25%/year; on the other, the company was still EBITDA unprofitable on 90%+ gross margins – but could be clearly deduced in the flurry of strategic and tactical pivots that Zillow has made over the last 4 years.
In 2016, Zillow launched Premier Agent Direct, which allowed Premier Agents to retarget Zillow visitors on Facebook (TripAdvisor is launching a similar off-platform retargeting effort with TripAdvisor Connect), as part of a broader effort to position itself as a sort of operating system for real estate agents, a digital hub where agents could upload and manage their listings, send marketing emails, and buy Google and Facebook ads. Zillow claimed that its intent-based search data made Facebook advertising more effective for agents than buying Facebook ads directly. At the time, agents would typically use Zillow to advertise specific listings and use Facebook to advertise themselves (so on Zillow, a visitor would see a list of houses prominently displayed with an agent’s contact information discretely tucked away, while on Facebook the same user might see an agent’s ad that reads “I put together a list of 1-bedroom homes in Portland listing for under $600k” or something like that).
But the partnership always felt somewhat tenuous as it seemed Facebook could just ingest agent/broker listings directly. And sure enough a year into the partnership, in direct competition with Zillow, Facebook unilaterally launched Dynamic Ads for Real Estate, where it would look at the search activity on a broker’s site, select properties from the broker’s listing inventory that matched a users’ search preferences, and advertise those properties to the user on Facebook and Instagram. Moreover, the solution was geared towards larger brokers with lots of listings, the same sort of brokers that Zillow was also going after at the time.
With Facebook gaining prominence as an acquisition channel and Premier Agents complaining about meager Zillow ROIs, Zillow was forced to experiment. It began directing its highest quality leads to its highest spending agents, forcing smaller agents to churn off in the process. It went through the trouble of qualifying leads, asking potential home buyers a series of questions to make sure they were really in the market for a house and not just casually browsing. It scrapped its subscription model in favor of a more Facebook-like auction pricing mechanism (see my last Zillow post for details), though this led to crazy bidding in some zip codes and higher prices overall (in 2018, the cost per lead spiked by 27%), which forced the company to impose pricing caps to support ROIs and stem agent churn. And finally, Zillow recently introduced a new “Flex” pricing model in select markets for its best performing agents who, rather than pay in advance for predictable lead flow, are given qualified leads and pay only after those leads convert into transactions, better aligning payment with performance. But these measures were modest tweaks to a legacy business model that seemed to be falling out of favor at a time when venture-backed companies like Opendoor and Offerpad were proposing something far more radical.
So in February 2019, steady state CEO Spencer Rascoff – who had grown the company from $55mn of revenue to over $1bn on the back of 15 acquisitions (with little to show in profitability, I might add) – was replaced by the idea/visionary CEO Richard Barton, who, in his own words, had “a particular penchant for and attraction to big swings”.