Redfin Corporation (RDFN) Q4 2022 Earnings Call Transcript

Redfin Corporation (NASDAQ:RDFN) Q4 2022 Results Conference Call February 16, 2023 4:30 PM ET

Company Participants

Meg Nunnally - Head of IR

Glenn Kelman - President, CEO and Director

Chris Nielsen - CFO

Conference Call Participants

Ryan McKeveny - Zelman & Associates

Tom White - D.A. Davidson

Ygal Arounian - Citigroup

John Campbell - Stephens

Michael Ng - Goldman Sachs

Jay McCanless - Wedbush

Chris Suchecki - Jefferies

Operator

Greetings, and welcome to the Redfin Corporation Fourth Quarter 2022 Earnings Conference Call. At this time, all participants are in a listen-only mode. A question and answer session will follow the formal presentation. [Operator Instructions] Please note that this conference is being recorded.

I will now turn the conference over to our host, Meg Nunnally, Head of Investor Relations. Thank you. You may begin.

Meg Nunnally

Good afternoon, and welcome to Redfin's financial results conference call for the fourth quarter and full year ended December 31, 2022. I'm Meg Nunnally, Redfin's Head of Investor Relations. Joining me on the call today is Glenn Kelman, our CEO; and Chris Nielsen, our CFO.

Before we start, note that some of our statements on today's call are forward-looking. We believe our assumptions and expectations related to these forward-looking statements are reasonable, but our actual results may turn out to be materially different. Please read and consider the risk factors in our SEC filings together with the content of today's call.

Any forward-looking statements are based on our assumptions today, and we don't undertake to update these statements in light of new information or future events. On this call, we will present non-GAAP measures when discussing our financial results. We encourage you to review today's earnings release, which is available on our website at investors.redfin.com for more information related to our non-GAAP measures, including the most directly comparable GAAP financial measures and related reconciliations.

All comparisons made in the course of this call are against the same period in the prior year, unless otherwise stated. Lastly, we will be providing a copy of our prepared remarks on our website by the conclusion of today's call, and a full transcript and audio replay will also be available soon after the call.

With that, I'll time turn the call over to Glenn.

Glenn Kelman

Thanks, Meg, and hi, everyone. Redfin generated $480 million of fourth quarter revenue, exceeding our projected range of $430 million to $459 million in revenue, mostly on the strength of Redfin now sales. Our net loss of $62 million included a $57 million gain from repurchasing at a discount $143 million of debt due in 2025. The adjusted EBITDA loss of $63 million was near the favorable end of our $58 million to $71 million guidance.

Earnings mostly exceeded expectations, but comparing the fourth quarter of '21 and 2022, Redfin lost 2 basis points of market share, in part due to layoffs and the loss of RedfinNow driven demand. As we compete better for online real estate traffic and improved sales execution, we expect share gains to accelerate in the second half, especially when we're no longer comparing our sales to a period of aggressive spending on agent hiring and home purchases.

It will be a major achievement to take share in a year where we're also improving annual profits by nearly $200 million, driven by higher gross margins, lower spending and the closure of money-losing businesses. We couldn't be more excited about the year ahead. Two months into 2023, we're still on course to earn an adjusted EBITDA profit for the full year and on schedule to sell our last RedfinNow home in the second quarter.

For the Property segment that includes RedfinNow, the gross profit losses in the fourth quarter were at the favorable end of our range. The full year gross profit losses for this segment were $23 million, and the 2023 gross profit losses should be a few million dollars -- excuse me, should be a few million dollars or less. Only 19 homes originally purchased by RedfinNow for $12.2 million neither been sold nor accepted an offer to be sold. The closure of RedfinNow is part of a larger shift to higher margin, less cyclical revenues.

The percentage of homebuyers served by our partner agents instead of our employees will increase from 37% in 2022 to projected 42% in 2023. We decided to sell more demand to partners after accounting for costs that aren't directly tied to a sale, but that still grow with the number of agents we employ, like the cost of human resources support and training for agents.

This decision will not only increase 2023 profits but also limit layoffs and losses in future downturns. Real estate services will drive more digital revenues as well rent, our online marketplace for promoting rental homes. Once the liability Rent has accelerated its growth in each of the last four months and expects to earn an adjusted EBITDA profit in the fourth quarter of 2023. Beyond rent, we launched two other digital businesses in the second quarter of 2022, our own mortgage marketplace and ads on redfin.com, both of which are now growing faster than any other Redfin business.

The mortgage marketplace offers our visitors a choice of lenders beyond the lender we acquired in April 2022, Bay Equity Home Loans. Bay Equity has a retail sales force to work with homebuyers already engaged with an agent, but doesn't have the call center to handle online inquiries at all hours of the day and night. Over the next two years, we expect to launch additional digital businesses, with the goal of running more money from an online visit than any other operator of a major real estate site.

Those are the structural changes we're making to improve Redfin's margins even at a onetime cost to our growth. More of every dollar of revenue should fall to the bottom line, but now we need more dollars to. This is why we'll spend the rest of this call on growth. The primary way we've grown is by reaching more people through our sites and mobile applications. Comparing the fourth quarter of 2021 and 2022, the average monthly visitors to Redfin's website and mobile applications declined by 2%. But over that same period, searches on Google for homes for sale declined 33%. The difference between these two numbers indicates why we likely increased Redfin's share of online real estate traffic. ComScore, which lets us compare ourselves to other sites reported a 6% fourth quarter decline for Redfin compared to 22% for realtor.com and 2% for Zillow.

According to comScore, we started keeping pace with Zillow in December despite a second half budget for TV ads, that was a quarter of the size of the Zillows. To improve our long-term competitive position, we know we have to draw visitors away from all our major rivals, not just one, and we believe that we can.

For Google searches on a home address and our 10 longest established markets, Redfin is now most likely to appear as the first result across the U.S. for these searches, we're now likely to appear for us more than realtor.com, which has nearly doubled our traffic. And we can still grow by expanding the parts of the U.S., our competitors already cover and by improving the machine learning software we use to recommend listings.

Drawing more visitors to Redfin is the first step in our growth. but we also want a higher proportion of those visitors to hire our agents. Because we look like other real estate sites, consumers often assume we're a marketplace for promoting the agents who paid us the highest fee. In fact, the whole reason we employed our own agents has been to deliver faster service at a lower fee from top producers. Almost no one knows that in 2022, Redfin agents had the highest average sales volume of any major brokerage beating our closest competitor by almost 20%.

Our agents' experience is one reason why our service is better. Now the site is telling that story, a process that started yesterday when we launched a redesigned redfin.com to promoter agents to luxury customers. We now route those customers to the Redfin agents with the most luxury experience. The week before, we launched an ad campaign that explains why for seven years straight. We've sold homes for more money than traditional agents because of our top producers but also due to the extra exposure each listing gets on redfin.com.

In 2023, we expect to gain share, not just due to more traffic and more customers from that traffic but also by getting more sales from each customer through mortgage and title sales and, over time, repeat and referral transactions. Our goal is to develop the brokerage into a second engine of Redfin's growth above and beyond our online presence.

We have more room to grow as a brokerage on as a website, nearly 20 years since online real estate portals first launch, about 95% of homebuyers search for listings online, just based on the sales claimed by Zillow and Redfin, it seems likely that only about 5% of home sales originate with people asking the agents on major real estate sites for service.

These sites can always squeeze a bit more revenue from traffic games or from traffic gains, not games or by claiming a higher share of the commission from each preferred sale. But if Redfin gives customers a reason to choose our own agents, building a brand for better service and value, that 5% of sales that start online can one day become 50%.

Employing our own agents could increase consumer affinity for our service but can also increase close rates. Our data indicates that the customers who asked us for service in the third quarter were less likely to go through with the purchase whether with Redfin or with a competing broker. But are the Redfin customers we met in the third quarter of 2022 who ended up buying a home, a projected 35% will have stuck with the Redfin agent for the purchase compared to 26% in the third quarter of 2021.

This tells us that even though the market is down, our sales execution is up. Beyond better service for customers who come to us via redfin.com, our agents are also generating their own sales. 34% of fourth quarter sales came from repeat and referral customers compared to 32% a year earlier. Even here, a Redfin agent has a massive advantage over traditional agents, having about 100 to 200 customers via redfin.com each year with each customer's contact information and online search activity tracked in our database.

Whereas many traditional agents are canvassing every Tom, Dick and Sally for sale, our agent's customer network typically starts with the hundreds of buyers and sellers who those agents met over the years through Redfin. More than ever, the agent's Redfin employees today are capable of driving loyalty sales. One reason for our improved sales execution is an improvement in the quality of our sales force, which ended 2022 30% smaller than it was 9 months earlier. Back in March, 37% of our agents had less than one year of Redfin tenure and only 29% had more than 3 years.

Today, only 13% have less than a year of tenure, whereas 42% have three or more years under their belts. That sales team now includes not just agents but also loan officers. After all, 1 rationale for serving the customer ourselves is more follow-on mortgage and title sales with all of our services working together to make a customers move easier and less expensive.

As in the third quarter, 17% of our brokerages fourth quarter homebuying customers borrowed money from Bay Equity Home Loan. The pre-acquisition high was 8%. After our January company kickoff got each region's agents and lenders in the same room for the first time, that number surged to a projected 21% for this February. From the fourth quarters of 2021 to 2022, the percentage of eligible brokerage customers who used our title service, Title Forward, also increased from 12% to 44% and lending and title margins improved in the fourth quarter, a trend likely to continue through 2023 as the lending industry completes its adjustment to lower volume, price competition may ease.

And as the housing market recovers, our brokerage lending and title businesses will be well positioned for growth. Our rent business, by contrast, has already benefited from rising apartment vacancies in the second half of 2022 and it's growing revenue now. This is a dazzling turnaround for a business acquired out of bankruptcy in April 2021 whose new CEO didn't start until August of that year. Net bookings, a measure of the annualized revenue rent added through sales to new customers less the annualized revenue loss from departing customers nearly doubled quarter-to-quarter from $5 million in the third quarter to $10 million in the fourth quarter.

In every quarter of 2021, net bookings have been negative $4 million to $5 million, and we're barely positive in the first half of 2022. Rent's fourth quarter revenue grew year-over-year for the first time since 2017 by 5% compared to the third quarter of 2022, fourth quarter revenue grew 6%. We expect Rent's revenue gains to accelerate on the strength of new products, a January 2022 price increase January 2023 price increase and improved sales productivity, which has more than tripled since the summer of 2021.

In support of the price increase, redfin.com's integration of Rent listings has generated more demand for Rent's property management customers. In June, redfin.com added 12% of rental visits above and beyond the visits to rent site. In the fourth quarter, this contribution grew to 18%. Rent also launched a $3 million fourth quarter mass media campaign offsetting some of the money we saved from the departure of rent employees.

We felt careful about investing in a second brand, but the appeal of a self-explanatory for large domain was so powerful that we had to give it a shot. We'll carefully evaluate the return on this advertising investment and run the business to generate adjusted EBITDA in the fourth quarter under a variety of market conditions. The success we've had so far is a tribute to the leadership of Rents CEO, Jon Ziglar and his whole team.

Now before turning the call over to Chris, let's discuss the housing market. On March 15, 2022, we were one of the first to state publicly that the market was cresting and "it was crazy for demand to be so strong in the midst of a market volatility and inflation. By May, the market began its first sustained decline since the great financial crisis. Ten months later, on January 25 of this year, we said that housing in January had been stronger than anyone could have hoped and that the market will "fragile" was "recovering." We caution that the recovery could be cut short by a rate hike.

By the end of that week, the National Association of Realtors reported that a seasonally adjusted index of pending home sales had improved 2.5% in December, a result surprising to many after six straight months of declines.

But in February, the market got another gold, first from last week's record low unemployment data and then from this week's report a persistent inflation and strong retail sales. Mortgage rates had fallen from 7.3% in November to 6% in early February, then climbed above 6.7% yesterday on the news. Unsurprisingly, Redfin's February demand is still better than it was in November even after accounting for the season, but worse than it was in January.

We still believe that our 2023 budgeting assumption of 4.3 million existing U.S. home sales is reasonable. Rate volatility and buyers' geriness about rates just make that market unusually hard to predict. We're running Redfin out of the cash register in 2023. So the existing home sales seem likely to follow $4.3 million, we'll reduce our spending.

Regardless of what happens to rates in 2023 and beyond, inventory will likely stay low. What's most remarkable about this housing downturn is that the number of homes for sale hasn't meaningfully increased from the calamitous lows of the pandemic. Sure, the number of homes on the market at the end of January 2023 was up 40% since January 2022, but it was still at roughly half the pre-pandemic level it was from 2016 to 2019 during a strong seller's market.

Our agents report that would be sellers with 30-year mortgages at a rate below 3% are choosing to keep their homes instead of selling either to live in or to rent out. This is why from May 2020 to May 2022, home prices increased 40%, but have fallen only 3% since. The millennial generation that mostly came of home buying age just after home prices and mortgage rates shot up still faces an affordability crisis with no real relief in sight.

Because of low inventory, we continue to believe that sales volume will be more volatile than home prices. Regardless of market conditions, Redfin will generate adjusted EBITDA in 2023 and net income in 2024. Once we recover from restructuring our business to be more profitable, our share gains will resume and accelerate. And if we can make money in a housing downturn, we'll be in a good position to make a lot of money when the market recovers.

Now let's hear from Chris on our financial performance and guidance.

Chris Nielsen

Thanks, Glenn. 2022 was a challenging year, but we've taken the right actions to position Redfin for long-term profitable growth. We're entering 2023 with appropriately conservative plans and the knowledge of the recovery may be touch and go, but it's comforting to see the start of the year tracking in line to slightly better than our expectations.

Since announcing our decision to wind down RedfinNow last November, we've moved quickly to reduce inventories. We applied the influx of cash to fully pay down the credit facility that supported home purchases and extinguished more than $140 million of convertible notes. We've also taken action to reduce costs and expand margins, including the elimination of our buy-side refund in December, which should set the business up for long-term margin expansion as we move through the year.

Fourth quarter revenue was $480 million and decreased 25% from a year ago. Real estate services revenue, which includes our brokerage and partner business generated $146 million debt in revenue, down 35% year-over-year. Brokerage revenue or revenue from home sales closed by our own agents decreased 34% on a 34% decrease in brokerage transactions.

Revenue from our partners decreased 45% on a 40% decrease in transactions and mix shift to lower value houses. Real estate services revenue per transaction increased 1% year-over-year. The property segment, which consists primarily of homes sold through RedfinNow generated $261 million in revenue, down 31% year-over-year. Revenue for this segment will be de minimis once the wind down is complete. Strong fourth quarter revenue for this segment is indicative of a quick action we've taken to reduce inventory.

Our rentals business generated $41 million in revenue, up 5% from a year ago. As Glenn highlighted, this is the first quarter of year-over-year revenue growth for the business since 2017. Our mortgage segment generated $28 million in revenue compared to $4 million in revenue in the prior year. The decrease was due to the acquisition -- increase -- sorry, it was due to the acquisition of Bay Equity, which occurred last April. This result was slightly below our guidance range of $29 million to $32 million with lower-than-expected volume over the last few weeks of the year impacting locked loans.

Finally, our Other segment, which includes title and other services contributed revenue of $6 million, an increase of 89% or $3 million year-over-year. The increase primarily was attributable to a $2 million increase from our title business and a $1 million increase from display ad revenue. As Glenn mentioned, growing high-margin digital revenue is an important priority, and we're just getting started.

Total gross profit was $37 million, down 65% year-over-year with total gross margin of 7.8%. Total operating expenses were $156 million, up $23 million year-over-year. Restructuring expenses contributed $22 million in Bay Equity, which we acquired in April contributed $9 million. Excluding these increases, operating expenses decreased by $8 million year-over-year.

Technology and development expenses increased by $3 million as compared with the same period in 2021. The increase was primarily attributable to the $2 million increase in online services, removing more of our technology infrastructure to cloud services. Marketing expenses increased by $2 million as compared with the same period in 2021. The increase was primarily attributable to a $1 million increase from Bay Equity and higher marketing expenses for our Rentals business.

General and administrative expenses decreased by $4 million as compared to the same period in 2021. Bay Equity added $8 million to these costs. Excluding Bay Equity, general and administrative expenses declined by $12 million, driven by a $7 million decrease in personnel costs.

Turning to segment level profitability. Real Estate Services gross margin was 18.0%, down 1,550 basis points year-over-year. This was driven by a 1,170 basis point increase in personnel costs and transaction bonuses, Total net loss for real estate services was $28 million, down from a net income of $19 million in the prior year. The decrease is primarily attributable to lower revenue and gross margins as the housing market slowed, partially offset by a $3 million year-over-year decrease in operating expenses. Properties gross margin was negative 7.0%, down from a positive 1.1% in the prior year.

Margin compression was driven by an 820 basis point increase in home purchase costs and related capitalized improvements as we sold through homes purchased earlier in the year. Gross profit losses were $18 million towards the better end of our $21 million to $17 million loss guidance range. Total net loss for properties was $26 million. Rentals gross margin was 76.4%, down 620 basis points year-over-year. Margin compression was driven by a 460 basis point increase in marketing as well as product mix shifts.

Total net loss for rentals is $22 million, down from a net loss of $14 million in the prior year. The increased loss is primarily attributable to lower gross margins and higher marketing operating expense, including the mass media campaign Glenn mentioned earlier. Mortgage gross margin was negative 8.9%, down from a positive 9.7% in the third quarter. Margin compression was driven by price and competition across the mortgage industry as lenders grappled with rising interest rates and excess capacity.

Total net loss for mortgage was $12.3 million. Other segment gross margin was 7.4%, an improvement from the negative 17.7% one year ago. Total net loss was $1 million compared to a net loss of $2.5 million in the prior year. Net loss was $62 million compared to a loss of $27 million in the prior year. While the year-over-year decrease in operational losses was wider, the drop in net income was compared to a smaller due to a $57 million gain on extinguishment of notes as well as $4.5 million higher interest income that benefited the quarter. Diluted loss per share attributable to common stock was $0.57 compared with diluted loss per share attributable to common stock of $0.27 per share one year ago.

Now turning to our financial expectations for the full year and the first quarter of 2023. We're not providing formal guidance for 2023, but as Glenn and I have both discussed on prior calls and reaffirm today, we're running the business to generate positive adjusted EBITDA for the fiscal year 2023. We've included a slide on the larger drivers behind this target in our earnings release presentation. Compared to 2022, we've already eliminated our refund to home buyers. This will add $1,000 to the revenue and gross profit of each brokerage transaction.

We'll get another 500 basis points of gross margin improvement from the staffing changes that we made in 2022. Closing RedfinNow saves $20 million in gross profit losses. Even with adding Bay Equity for a full year, the actions we've already taken will lower full year operating expenses by $25 million. That work is all done, and we expect more than $40 million in adjusted EBITDA improvements from momentum in our rentals and mortgage businesses.

Turning to our expectations for the first quarter of 2023. We expect consolidated revenue between $307 million and $324 million, representing a year-over-year decline between 49% and 46%. We expect our real estate services segment to account for $122 million to $130 million of that revenue, representing a year-over-year decline between 31% and 27%. Properties revenue is expected to be between $108 million and $113 million as we sell through most of the remaining inventory. Rentals revenue is expected to be between $41 million and $42 million, representing a year-over-year increase between 9% and 12%.

Mortgage revenue is expected to be between $29 million and $32 million. Turning segment gross profit. We expect real estate services gross margins to be flat to slightly down year-over-year. Headwinds to real estate services gross margins in the first quarter included about 100 basis points from an in-person company kickoff event that we did not have in the prior year due to COVID and will not repeat in 2024. That event also added about $8 million to our first quarter G&A expenses.

Turning to properties. We expect gross profit to be slightly negative to near breakeven for the first quarter. For mortgage, we expect first quarter gross margins to return to positive territory. Total net loss is expected to be between $160 million and $105 million. This net loss assumption includes a gain on extinguished lines notes of $7 million, reflecting the repurchase of $19 million in convertible notes already completed in the first quarter. These gains are excluded from adjusted EBITDA guidance. Adjusted EBITDA loss is expected to be between $84 million and $73 million.

As a reminder, the first quarter of the year is typically our lowest volume quarter, and even accounting for these losses, we still believe we are on track to generate positive adjusted EBITDA for the full year 2023. On a consolidated basis, this guidance includes approximately $45 million in total company marketing expense; $20 million in stock-based compensation, $17 million of depreciation and amortization and flat net interest expense associated with our convertible senior notes and other credit obligations.

In addition, we expect to pay a quarterly dividend of 30,640 shares of common stock to our preferred stockholder. This guidance assumes, among other things, that no additional business acquisitions, convertible note or stock repurchases, investments, restructurings or legal settlements are concluded and that there are no further revisions to stock-based compensation estimates.

And now let's take your questions.

Question-and-Answer Session

Operator

[Operator Instructions] Our first question comes from Ryan McKeveny with Zelman & Associates. Please state your question.

Ryan McKeveny

Chris, I think you might have hit on this a bit within the guidance commentary. But I just wanted to touch base on it. The brokerage revenue per transaction and kind of squaring that with the gross margin. So I guess, as we think about the moving pieces to the revenue per transaction, I guess it's home prices mix. But then on the other side, you have the buy-side refund coming out, which I guess all else equal would benefit that.

So can you maybe just give a little more detail on how those various factors are flowing into the revenue per transaction? And as we look forward, maybe how you would expect that to trend into 2023.

Chris Nielsen

Sure. So the note that we provided here is that we're expecting about $1,000 revenue per brokerage transaction increase as compared with 2022. That takes into account the factors that you just mentioned where we will be and have eliminated the homebuyer refund. It also takes into account the relative mix of homebuyer and home seller transactions, plus various other market mix things, including the geographies across the U.S. So that's a round figure, obviously, but that's the right perspective on how to think about that combination of things as it relates to revenue per brokerage transaction.

Ryan McKeveny

And Glenn, I guess just from a data and kind of market trends perspective, obviously, you always have really good insights. And I guess if we think about the West Coast, California, Seattle, some of the markets where market share-wise, you guys are fairly large. Any updates? I understand the general commentary you provided on January and February, but I guess specific to the West Coast, anything you would call out about what you're seeing within those markets?

And I ask because I assume data wise for the industry, it would certainly seem that the declines in transaction activity was much steeper, especially in the back half of '22, which, given your footprint, I think, was probably an incremental headwind that you guys faced. So yes, any thoughts you can share on kind of current thinking within the West Coast markets? And would you expect some of the industry headwind from the mix to eventually potentially become a tailwind if those markets start to improve?

Glenn Kelman

Well, it's a tale of two markets. So Seattle is actually doing significantly better. It may be related to the fact that Seattle doesn't have an income tax. So people are still migrating to Washington State. But California continues to be a place where people are leaving. And so we are facilitating their moves to other states. The same is true of Oregon. Our geographic concentration has been an issue that we've worked on, both through the two layoffs that we had last year and on a go-forward basis, when we do hire someone, it's going to be in states like Texas, Georgia and Florida.

Operator

Our next question comes from Tom White with D.A. Davidson. Please state your question.

Tom White

Glenn, your commentary about sort of leaning more into your partner network and brokerages, would you characterize that as sort of like a temporary shift as you guys kind of trying down the hatches and move towards profitability? Or is it more of a kind of permanent change in your thinking?

And if it's temporary, just curious whether you guys are kind of taking any added steps to make sure that these partners are providing the service levels that you want or -- and I don't know maybe there's anything you could do to get them to offer Redfin like pricing and how you think about the trade-off between this partner experience for the customers and kind of the Redfin brand?

Glenn Kelman

Great question. So it is a durable, but modest shift. So going through this near-death experience of trading at $3 or $4 a share may us examine every cost associated with employing agents and comparing that to the gross profits we get from partner agents. And we do think that over time, modestly shifting demand to partner agents will lead to more gross profit and overall profit.

But having said that, Ryan Sudeck is a new executive at Redfin. He's been at the company a long time previously in business development. He's running our partner network. And we are different than most of the lead generation sites. We don't sell leads to real estate agents and are indifferent financially to the outcome.

We depend on those agents to close sales. So calling the herd, making sure the partner agents deliver fantastic service is not only important to the mission of the company, but it's very financially important to us. One headwind on taking share in 2023 is just that partners have lower close rates than our employees, and that is so significant that it's hard sometimes for us to give a customer to a partner the partner will generate more gross profit. So even though we are responding to that financial reality. Really, it's a customer service challenge to make sure that the partner is delivering service that we're proud of.

So we're going to be spending more demand to the best partners asking the worst partners to leave our site and also building technology to make the handoff between the customer and the partner more seamless. It will never be as good as it is with our employees because the employees are augmented by a network of contractors to offer on-demand service 24/7, but it can get better with technology and through better curation of the agents.

Tom White

A quick follow-up. Is there a way for you guys to participate in ancillary kind of revenue streams on those partner transactions? I mean can you guys attach some sort of condition to the lead to the partner agent that they've got to promote maybe your mortgage or title product?

Glenn Kelman

Well, we have to be careful about that because we don't want to break the law, but we can certainly encourage our partners to recommend Bay Equity because it offers the best service at an incredible rate. We can also take some of the real estate agents who have been served by Bay Equity, traditional agents at RE/MAX or Keller Williams or some other brokerage and bring them into our partner program so that their partners not only with our mortgage business, but with our website, and that will naturally drive attach rate.

We've certainly seen that other websites don't employ agents, so really trying to drive attach rate even when those agents don't work for them. And we're obviously limited by the margin we want to make on the mortgage, but we think we can do better there. Our first priority is going to be doing really well with the employees, but our second is driving it through the partners. It's a great strategic front.

Operator

Our next question comes from Ygal Arounian with Citigroup. Please state your question.

Ygal Arounian

I want to expand on that last point for a little bit more because it feels like a pretty big shift in your view and philosophy. And Glenn, I think in the past, what you've said is that even if it's at a lower gross margin percentage to do it yourself within your brokerage the overall gross profit dollars are better and higher. And it sounds like you're at least changing your view on -- so maybe as you kind of dug into a $3 when your stock is at $3, you dug in all this. What changed in your view here? And what did you uncover that made you kind of push in this direction a little bit more?

Glenn Kelman

Well, first of all, I don't want to be defensive at all equal, but we haven't shifted -- sorry, there's an echo on the phone, but we haven't shifted our view that we're trying to maximize gross profit dollars. If we were trying to maximize gross margin, we would be just a pure website, but we're running the business to generate as many dollars of profit as we possibly can. And what changed was just the accounting of every possible expense.

So last year, we just asked ourselves, is it more expensive to employ agents than we realize? Are there costs that even though they're not cost of revenue, are still cost of employing agents. And when we added those costs, it modestly shifted the balance between partners and employees. Sometimes a theme on the scale in favor of the employees because when we build a customer relationship with someone, we get follow-on sales through repeat and referral transactions but also higher mortgage and title attach rates. And at this point, because we're running for the roses in 2023 to make sure we generate adjusted EBITDA profit, we just had to have the discipline to say the transaction that will generate the most gross profit, whether it's through the partner or the employee that will rule the day.

And so it's only a 5-point shift. We still believe it was a hybrid model. if we try to serve everybody from employees, we'd go crazy hiring up and laying people off. We could never keep people busy through the winter, and we could never serve customers well through the summer. And conversely, if we referred all the demand on our website to partners, we could never build a brand for service. And so the only way we could grow is really by driving more online traffic and conversion, and there's just a limit to that, too. So we believe in the combo more than ever, but we're going to adjust the balance between these two businesses. Every time we see close rates go up with partners or we see costs go up with employees or vice versa.

Ygal Arounian

Yes. And it certainly wasn't trying to imply that you weren't trying to maximize gross profit dollars. I understood that, but it feels like some of the dynamics. Yes, what you explained here. So -- and then -- so my follow-up question is on this -- the digital margin -- high-margin digital revenue still pretty -- maybe you could just expand on, I don't know how much it's contributing and what the long-term opportunities are, where else you can go with it. Whatever else you could share understanding that it's early days.

Glenn Kelman

Sure. Well, I just think our website has been undermonetized and the contribution may seem small when you look at the revenue, especially compared to a near zero margin business or a negative margin business like our properties business, but the contribution to direct profits that is high-octane fuel for our business just because it runs at such an insane margin. So display ads in the mortgage marketplace were what we launched in the second quarter, that's under great leadership at Redfin.

We also believe that there's all sorts of other constituents who want access to consumers. So for example, the builders want to create a better consumer experience on redfin.com. Right now, we show all the homes that are currently for sale, but we don't show homes that haven't been finished. We don't show floor plans and other things that the builders want to market to those consumers. So giving the builders an opportunity to reach that audience is something that will make money for us, but it will also improve the consumer experience. And there's just so much money in motion when people move and such a small fraction of people who decide to use Redfin agents that we really think it's a big opportunity. So it will take us time to build that. But in some ways, Redfin has been getting smaller and it may feel like we're diminished by that, but in some ways, we're getting bigger.

And where we're getting smaller is in really capital-intensive low-margin businesses and where we're getting bigger is in really high-margin businesses that can monetize an already very large audience.

Chris, do you have anything you want to add to that?

Chris Nielsen

No, just from the financial reporting standpoint, you'll see that reflected in our other segment that it primarily includes these kinds of businesses plus our title business. So we're talking about a few million dollars here for the fourth quarter, but as we both indicated, so we're encouraged with what we've seen so far here.

Operator

Your next question comes from John Campbell with Stephens. Please state your question.

John Campbell

On the EBITDA inflection goal for this year, I appreciate you guys providing that bridge in the investor deck. That's super helpful. It's actually a question I had kind of queued up for you guys punch, but it seems like a bulk of those kind of profit drivers you've outlined are basically a given -- it sounds like you've already actioned most of that, obviously. Could you maybe talk to what you're foreseeing as the variable or kind of macro influence components? I'm thinking the $40 million of business momentum and then maybe to an extent the 500 bps of real estate gross margin expansion?

Chris Nielsen

Sure. So you're right, most of these programs, most of these changes are already underway. Just using the example, the 500 basis point improvement in real estate services gross margin. That's because we believe we've stopped the business to be consistent with the kind of demand that we'll see during the course of the year. If we see more demand than that or less demand than that, we may need to adjust our staffing levels, but we believe that we're set up to be a good match with the kinds of volumes that we're going to expect through the course of the year.

It's really the same thing on the additional adjusted EBITDA from the rentals business and the mortgage business. But I think just the one extra comment that I'd provide here is that while many of these actions have already been taken, we also will continue to pay attention to the demand side of things. If it turns out that demand is different from what we've expected, we know that we can make changes elsewhere in the business to get back to that goal.

John Campbell

And then on the loyalty or the repeat sales, I feel like that's a metric that often gets overlooked. Average homeowner duration, I think, has grown to 13 years. And then you guys obviously wind back the clock that far. You guys were doing a fraction of the transactions back then versus what you're doing now. So to get I think, Glenn, you said the 38% mix of transactions being loyalty sales, I've got to think that the vast majority of past customers are using you guys again. So I mean, the back of the tell me that's pretty big deals. Am I thinking about that right?

Glenn Kelman

Well, first of all, it's 34, not 38, but we'll aspire to 38, John. Some of this is repeat business, where we are capturing a really high percentage for a long time. We made the claim. I don't think we comfort anymore that if you look at how likely a customer is to stick with the brokerage, we were number one and some of that is because the brokerage just stands for something. It's not just that we happen to have the same agent there with a person relationship, it's a 1% fee in on-demand service and great technology through the whole transaction.

But some of this is just referral. At some level, the goal of the website is to light the log, but not to be the log itself. We wanted to put map-based search online so that we could tell the world there's a better way to buy and sell houses. But if we only grow in line with traffic, that means the rest of the world doesn't believe that. And if you look at where Zillow Redfin and infer where Realtor is, you are at about 5% of all sales originating from portals and yet 95% of people are using those websites. So clearly, we've done well to put listings online, but technology hasn't made the rest of the experience obviously better for consumers, and that's where we want to break through. So that's why long term, we want to see half of our sales come from repeat and referral sales so that our growth is untethered from the website are compounded by the websites growth.

By the way, as John asked questions before. I love the housing stuff, John. It's really good.

Operator

And our next question comes from Michael Ng with Goldman Sachs. Please state your question.

Michael Ng

My first one is just on the trajectory of OpEx for the rest of the year. And also, if you could just give a little bit more color around your marketing plans. I was a little bit surprised to see the $45 million outlook because I thought in the year ago period, there were some timing as it relates -- related to like mass marketing campaigns that made 1Q kind of unusually high, but would just love to hear a little bit more about those plans for this year.

Chris Nielsen

Sure. So on the operating expense front, maybe just let me start with marketing there. We expect to continue to run a mass media campaign that will go mostly in the first quarter and second quarters of the year. And so there'll be extra weight on operating expenses in the first half related to that. We just found that those campaigns are much more successful as we're speaking with customers leading into the main part of the home buying and selling season than they are in the second half of the year.

And so that's why we've put the weight in that direction. I also mentioned in terms of our G&A expenses. We had a little bit of extra G&A expense in the first quarter of this year, about $8 million associated with our in-person company kickoff of that. That's an example of something that will fall off as the year goes on. But it's probably those two dynamics that are most important in terms of the trajectory of operating expenses during the course of the year that once we get into the second half, things should be pretty much level quarter-to-quarter because we will have gotten past those big marketing expenses and some of the unusual first quarter this.

Glenn Kelman

Can I just speak to the higher level thinking around our allocation of capital? We sucked it up and cut costs on software for our real estate agents because that work is principally done. But the goose that lays the golden egg investing in the listing search capability to drive more traffic and then in media to build our brand. That's the one place where we really suffered because we want to keep the top line moving. And so we definitely have heard that other companies are leaning in and spending more. We're trying to have the best of both worlds where we spend more in the areas that will really drive growth and we spend less everywhere else.

Operator

[Operator Instructions] Our next question comes from Jay McCanless with Wedbush. Please state your question.

Jay McCanless

The first one I had and what you're talking about with building out the call center for the mortgage operation, that's typically higher cost -- ongoing cost type of operation, I guess, could you talk about the development of that with the goal also of being EBITDA positive by year-end?

Glenn Kelman

Sure. Well, I think I misspoke because in the call in the prepared remarks, I was trying to say that we don't have a call center and instead, we're going to refer people who come straight from our website, typically looking to refinance a loan to a lender that does have a call center. So rather than building out our own call center for Bay Equity home loans, we're going to build a digital mortgage marketplace. In fact, we've already built that to refer those consumers to other lenders. I was just explaining the rationale. There are some customers who want to serve directly. Mostly those are people working with our real estate agents who are well qualified, who want a local lender employed by Bay Equity. And then there's these other folks on our website, who at 1 in the morning, say, "I want to quote and it's better to send them to a lender like rocket, but we're not going to build a call center in 2023 for lending.

Jay McCanless

The second question I had, the EBITDA loss that you guys are projecting was about double what I think consensus and we were looking for -- is that going to be more gross margin or cost of goods sold weighted this quarter because of selling through the remainder of the homes and properties or just a little heavier OpEx, what you're talking about earlier with the in-person event and the higher marketing spend, how should we balance that out?

Chris Nielsen

Yes. We do expect to have some gross profit losses associated with selling through the remaining inventory. But I think mostly what you're seeing and the information we were trying to provide just to give you a little bit of a shape of the year in terms of operating expenses and where we expect to see gross margin improvement during the course of the year.

And so I think that's the best way to think through it. Maybe just one other comment, which is from our standpoint, this is pretty typical in terms of the way the year plays out, which is we incur the highest expenses as we're advertising to bring customers into the market for the year and have the lowest closes in the first quarter. And then that pays off as we get into the heart of the buying and selling season.

Jay McCanless

And just one other, if I could sneak it in. Maybe any quantification of business trends. I believe you said it was down in February versus January. Any quantification or additional color you could give on what type of declines you're seeing?

Glenn Kelman

Well, I think it's fair to say that we were down 30-plus percent year-on-year in November in terms of people asking to tour homes coming forward to write offers. And then in January, that year-over-year drop narrowed to high teens, low 20s. And in February, it's solidly in the 20s. So it's not calamitous. And actually, if you look at booked offers, where you're just focusing on how many people wrote an offer and got it accepted, that year-over-year drop has just narrowed steadily from November to now, but that's usually a byproduct of rate increases.

So the people who made an offer in the last couple of weeks, who locked the loan three or four weeks ago. There's no way they're not going to go through with it because they got access to a 6% loan when rates pop back up to 6.75%. I think tours is a better leading indicator.

And so going from the 30s down to the high teens into the low 20s year-over-year drop is what we're seeing. And I wouldn't be worried that it's a drop just generally because those pandemic codes were unsustainable. What we're really looking to do is get better than we were in November.

Operator

Our next question comes from John Colantuoni with Jefferies. Please state your question.

Chris Suchecki

This is Chris Suchecki on for John. Can you just walk us through some incremental detail on how the Bay Equity integration has kind of progressed throughout the quarter? And then another 1 on kind of mortgage attach rates. I'm just kind of curious what drove the acceleration between the end of the quarter and what you were seeing in February and kind of any leverage you think you have to drive that higher in '23?

Glenn Kelman

Why don't I start, Chris, if you have anything to add, you should finish. I think part of it is that it just takes time to build relationships between loan officers and real estate agents. And one of the signal events for doing that was the sales kickoff, which added $8 million to our Q1 expense. By getting the lenders and the agents in 1 room and just making sure everybody understands that when we work together, we can deliver more value to the customer and a better customer experience was probably what drove this uptick in attach rate.

There's also been some online marketing where when we sign someone up for a home tour off our website. The website says, do you need to figure out what you can afford. We can connect you with the lender. We've been surprised at how many people said, sure. We thought that there wouldn't be much uptake on that. Bay Equity really wanted it. And I said, fine, we'll give it to them. it's not going to do much. It has. It's only been a few days since it went nationwide but in the pilot at the end of Q4 and the first weeks of January, it was really good.

So we've been really encouraged about it. I think also they just really are delivering great service. We've gotten our agents to use Title Forward at a point when it wasn't delivering great service, and that just comes back in your face. But in this case, it's compounding in a positive direction because an agent sends one customer over there, the customer has a great experience and then we do more. What we're really proud of is that 21% for February, that's not in a pilot market, that's not in one place. It's everywhere that the equity operates. And so we think it's a sustainable durable result.

Chris, do you have anything to add?

Chris Nielsen

I don't. You hit the main points.

Operator

And there are no further questions at this time. So I'll now turn the call back over to Meg Nunnally for closing remarks.

Meg Nunnally

Thanks, everyone, for joining the call today. We can go ahead and wrap up now.

Operator

Thank you. This concludes today's conference. All parties may disconnect. Have a great day.