Redfin’s Q3 results showed a solid rebound from a rather disappointing Q2. It looks great all around, with revenues, profit, market share, etc. all up and up very healthy. Agent productivity is up, traffic is up, and Redfin Now is a $80 million business in Q3. So all in all, a very nice rebound.

There is, however, a lingering shadow over all the good news, and there are a couple of details from the earnings call that are worth calling out and examining a bit. Let’s do that, shall we?

The Numbers

Let’s start with the numbers, as we always do.

Those are solid numbers to be posting. The 70% YOY increase in revenues is the headliner, but the more important is the 23% increase in Service revenues, since most of the Product segment comes from Redfin Now, the iBuyer program at Redfin, which tends to inflate the topline revenue numbers.

The growth in gross profit from the core brokerage business is also quite nice at 27% YOY, and by virtue of holding the line on costs, Net Income almost doubled YOY for the quarter. Really nice work there all around.

Speaking of the core business, it’s good to see gross margins for the brokerage business up 110 bps from 34.0% in 2018 to 35.1%. And most of the key metrics look great:

25% increase in transaction sides, to over 16,000 of them; Redfin is already over 40K transactions for the year and probably should end 2019 with around 50,000 transactions. That’s quite a lot of growth, especially since all of it came without any acquisitions. Transactions, revenue and gross profit per 1,000 Unique Monthly Visitor are all up, which is encouraging as well.

All of that is very, very good. Like I said, a solid rebound quarter.

The Lingering Shadow

Having said all that, I’m still puzzled by and concerned by Redfin Now.

For starters, I’m a bit nonplussed that Redfin still won’t release any unit economics for its Redfin Now business, especially now that it is an $80 million in quarterly revenues business. That’s over a third of total revenues, and we have very little transparency or visibility into how Redfin Now is working.

It is especially frustrating since Glenn Kelman keeps talking about Redfin’s competence in being careful, not losing money, and in execution. From the earnings call:

And then the levers on profit are really the ones that we already discussed. Operational excellence and just grinding on margin are something that we really pride ourselves on because we’ve been in a cost-sensitive business for so long. We’ve been a value leader for so long. We view the properties business as a race to the bottom nobody cares whether the check came from Redfin or somebody else. They care about the amount. And so this is an economic war that we feel well equipped to fight. Having said that pricing is just so important if you misprice a house by $5000 or $10000 no matter how many efficiency gains you get on mowing the lawn for $55 instead of $60 a week it just doesn’t matter. So that is a software problem. It’s a software problem we’ve been working on for a long time. The fact that our machine learning team just came up with another astounding increase in accuracy for our estimate we feel good about it.

Elsewhere in the call, Glenn explains Redfin’s focus on profitability over growth:

So growth is the easy part in the properties business because it’s such a novel product. And because it’s so easy to sell dollar bills for $0.95. So I think we just have to be really careful when we look at the properties business which has grown 600% year-over-year that we grow that business in a responsible way. So if you are looking for a shift in tone it probably came there but almost everything else is the way that we’ve always been. And even with Redfin Now I feel like that’s the way we’ve always been but the market has just had different view that other players in the space have been more growth oriented so we wanted to be clear about where we stand.

Okay, well if profitability is the focus of Redfin Now — versus competition (read, Zillow) whose focus might be growth — then Redfin Now has to not lose money on a gross basis. Yes, going from negative 2.7% gross margins in the Properties segment to negative 0.9% is an improvement, and a big one, but when Opendoor is showing 10% gross margins and Zillow is showing 7.5% gross margins, losing money doesn’t strike as acceptable.

When Glenn says other companies are selling a dollar for $0.95, I just have to wonder from where he’s getting that number.

I fear we may be getting a bit too deep into accounting choices and so on, since Glenn talked about how conservative Redfin has been in accounting: “counting every possible field expense against our gross profit.” But that just isn’t what my numbers are saying. Opendoor, Offerpad, and Zillow all make money on a gross basis after counting their field expenses. They all lose money after you take the overhead of growth (new offices, new employees, etc.) into account, but since Redfin is using its own Redfin agents to sell the homes it owns, unlike its competitors, shouldn’t the gross margin be a whole lot better than a negative number? Especially when Opendoor and Zillow are showing 10% and 7.5% gross margins, respectively?

Put another way, I would love for Redfin to release some unit economics numbers, and do a special release showing what Redfin Now’s gross margins would be if they did the same accounting treatment that Zillow does. (Opendoor being a private company, we have no idea.) If that number isn’t more than 7.5%, then we have a problem.

Because as I said in the Q2 Redfin writeup, Redfin can lose to Zillow in website traffic, lead routing, referral business, even mortgage and title. It simply cannot lose to Zillow in buying and selling homes, not with its 14 years of history and boots on the ground and local data and local knowledge. It just can’t. Just telling us that it’s winning on profitability without actually showing us some data, showing us some work, is starting to get rather tiresome.

The gross margins of Redfin Now has been and still remains a lingering shadow on Redfin. It’s high time that Redfin start disclosing some data on that so we all (but particularly investors) can start to compare apples to apples between the two public companies that do instant offers. And if Redfin is focused on profitability, on not making stupid purchases, and really does know how to make money, its refusal to boast about how much more cash they’re banking compared to Zillow is really quite a headscratcher.

To quote the poet philosophers from Long Island, De La Soul, “Don’t flaunt that the candy is good, unless it can be plenty.

Some Interesting Details

Finally, there are some intriguing details in the earnings call that are… well, intriguing.

Mortgage Attach Rates

One of the more interesting items was Glenn talking about how Redfin Mortgage was going strong, and that they almost hit their goal of 25% attach rates. I don’t know if that’s good or not, but I’ll assume good since Glenn set 25% as the goal.

Thing is, Zillow is out there talking about new home builder type of attach rates, which could be 75%, because as the actual owner of the home, Zillow can offer incentives to use Zillow Home Loans for the ultimate buyer as new home builders do. Redfin can do the same with Redfin Now homes, of course, but we just have very little clarity on that front.

Redfin Converging Towards the Traditional?

There was quite a bit of conversation in the call about agent productivity. The jump from 9.2 to 10.2 YOY is quite impressive, as is the jump from 9.7 to 10.2 Q/Q. But Redfin also cut agents (at least Lead Agents) from Q2 to Q3, so that makes it even more so. Whatever they’re doing on that front is working, so good on that.

And according to Chris Nielsen, CFO, most of the gross margin gains are from controlling personnel costs:

Sure. So let me start with the question on real estate services gross margin. Those did primarily improve related to personnel expenses as a percentage of revenue. And not only did we get improvement in agent productivity that supports staff productivity in the third quarter of 2019 as compared to third quarter of 2018.

That’s a nice sign. Combine increased productivity back up to say 2017 levels (34 transactions per agent) from what they were in 2018 (31 or so) while controlling costs, and Redfin should see real sustained margin expansion.

The interesting part of that equation, however, is the different approach Redfin is taking with its agents. Basically, Redfin is embracing a lot of the concepts from the traditional brokerage models that it has spent 14 years upsetting. Basically, the new performance standards for Redfin agents is that they add lead conversion to the mix.

Here’s Glenn:

And so I think we take a closed sale as the most sincere expression of confidence in the agent’s service and telling the agents that if you’re going to meet a bunch of customers who have been sourced from our website you got to keep up your end of the bargain and really knock their socks off.

The rationale is that Redfin agents used to be “too passive” when dealing with the customer. Redfin would like its agents to be more proactive in trying to sell the customer on a neighborhood, or on a house. Okay, that’s… good, I guess. But then we get this:

And so having a sales force that is more proactive but still driven by the values of the company to be different than every other brokerage and to take care of the customer and to tell that customer to walk away from the wrong deal is the balance we’re trying to strike. It’s a huge cultural shift. And I think it’s absolutely crucial because for the longest time the engine of our growth has been our website and most employees at the company real estate agents have not had a role in driving growth. They basically have been relegated to some kind of fulfillment role. And they want to be those partners. [Emphasis added]

But there’s more! At one point, Glenn Kelman talks about a new level of Redfin Agent, the “principal agent”:

There is a new level called principal agent for some of our best-performing agents. And that rewards not just productivity or customer success rate but also customer loyalty. If you’ve been at Redfin for five, six, eight years as an agent you often can be off the website for six, nine months out of the year because so many customers are asking for you by name, they’re referrals, it’s repeat business and all the rest.

So we want those agents to earn a premium because they’re so profitable for us and recognizing them as principal agents is probably the harbinger of other changes where we continue to encourage our agents to use the website to build a book of business but then also invest in that book of business over time where they check in on clients, or sending holiday cards because obviously online loyalty programs can be effective but it makes a difference when your agent’s in the neighborhood and stops by to say hi. [Emphasis added]

So, in other words, Redfin agents used to just work the leads that Redfin’s website generated. They were in the fulfillment role, on purpose and by design. Because Redfin wanted the relationship to be between the consumer and Redfin, not between the consumer and an individual agent. Down that path lies 1099 independent contractors, traditional brokerage, and margin compression.

But now, Redfin agents now have to convert web leads at some level acceptable to Redfin, or get fired, and Redfin will not only reward some of the top agents, but will encourage Redfin agents to build a book of business. That’s a pretty major shift.

Convert leads at a satisfactory level, and build a book of business, including sending holiday cards and stopping by to say hi, and… generate referrals and repeat business and all the rest…. That sounds a whole lot like what every traditional brokerage in America wants its agents to do: build a sphere of influence, or a geographic farm, or a base of referrals and repeat business….

That’s not quite the same thing as say Keller Williams or Realogy, but it’s no longer in a different universe either. It’s at least in the same ballpark as traditional sphere-and-farm lead generation deal that traditional brokers and agents have been doing for decades.

It might be a wonderful boost for Redfin, and it might be a total fail. But it does strike me as Redfin converging towards traditional models. It’s at least taking a step in that direction, as opposed to a step in say Zillow’s direction.

Redfin Direct to Non-Redfin Agents

Finally, this is one of the most interesting exchanges of the call:

John Campbell — Stephens — Analyst

Hey, guys, good afternoon and congrats on the continued success. But Glenn I’ve got one more on direct offers and don’t worry we’re going to stay patient there. But you guys have rolled that out in I guess a handful of markets now it’s only on Redfin for-sale properties. But I think recently maybe a couple of months ago you guys started to solicit feedback around basically open that up outside of Redfin and kind of into the broader industry. But I’m curious about how that’s gone so far? And what if any pushback you might have received?

Glenn Kelman — Chief Executive Officer

John not a sparrow falls without you noticing it. We’re trying to be sneaky about that. Just because our diabolical plans haven’t been announced to the world… so at some point it makes sense for us to let other agents including the agents who have complained about Redfin Direct put that button on their own listings because it’s good for their clients and it’s actually good for their own business. I don’t think we’re quite ready to do it yet but the results of the survey you asked about where we asked some agents would you be willing to do this were good. Plenty of agents would be interested in doing it.

John Campbell is very good indeed, but it’s even better that Glenn gave a straight answer. Yes, Glenn said they’re not quite ready to offer Redfin Direct (unrepresented buyer making an offer to save some money) to non-Redfin agents, but… I think we can safely say that this earnings call might as well be the announcement to the world of their “diabolical plans.”

That will be an interesting development indeed, if Redfin really does start to offer Redfin Direct to non-Redfin agents and non-Redfin agents take them up on it. Sure, it’s another step in the erosion of cooperation and compensation, but… if that’s what the consumer wants, then that’s the way the industry will go.

Conclusion

Let’s leave it there for now. I may need to come back to Redfin after Zillow and Realogy report.

On the whole, Redfin had a great Q3. It really was a nice rebound from the past few quarters. Great performances on the core brokerage business, on revenues, on margins, on agent productivity — it’s all good.

My one reservation remains the mysterious lack of transparency on Redfin Now, which can no longer be classified as a little experiment, since it makes up more than a third of Redfin’s revenues. I hope we’ll see Redfin be more forthcoming with data and information on this important business segment.

And there are some hints that Redfin is converging or at least taking a couple of tentative steps towards the more traditional model of brokerage, where the agent is not merely a service delivery channel but also responsible for converting leads and sourcing leads, particularly the kind of sphere, farming, and past clients type of leads that have long been in the wheelhouse of traditional brokers. If Redfin can show that kind of individual agent brand marketing (because that’s what most of those are) working within the tightly controlled W2 model that it has, that’s a pretty significant win for Redfin. If, on the other hand, we see the kind of commission pressure that other brokerages face because of their reliance on those individual agents and their spheres/farms/referrals, then it could be disastrous for Redfin as it has been disastrous for traditional brokerages.

We shall see.

-rsh