Good day, and thank you for standing by. Welcome to the Root Q3 2021 Earnings Conference Call. At this time, all participants are in listen-only mode. After the speaker's presentation, there'll be a question-and-answer session. To ask a question during this session, you'll need to press star one on your telephone. Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Christine Patrick, Vice President of Investor Relations. Please go ahead.
Good morning, and thank you for joining us today. Root is hosting this call to discuss its Q3 2021 earnings results. Participating on today's call are Alex Timm, Co-Founder and Chief Executive Officer, and Dan Rosenthal, Chief Operating Officer, Chief Revenue Officer, and Chief Financial Officer. During the question-and-answer portion of this call, our presenters will be joined by Matt Bonakdarpour, Chief Data Science and Analytics Officer, and Frank Palmer, Chief Insurance Officer. Last evening, Root issued a shareholder letter announcing its financial results. While this call will reflect items discussed within that document, for more complete information about our financial performance, we also encourage you to read our 10-Q as well as our 2020 Form 10-K.
Before we begin, I want to remind you that matters discussed on today's call will include forward-looking statements related to our operating performance, financial goals, and business outlook, which are based on management's current beliefs and assumptions. Please note that these forward-looking statements reflect our opinions as of the date of this call, and we undertake no obligation to revise this information as a result of new developments that may occur. Forward-looking statements are subject to various risks, uncertainties, and other factors that could cause our actual results to differ materially from those expected and described today. In addition, we are subject to a number of risks that may significantly impact our business and financial results.
For a more detailed description of our risk factors, once again, please review our Form 10-K, where you will see a discussion of factors that could cause the company's actual results to differ materially from these statements, as well as our shareholder letter released last evening. A replay of this conference call will be available on our website under the Investor Relations section. I would like to also remind you that during the call, we are discussing some non-GAAP measures in talking about Root's performance. You can find the reconciliation of those historical measures to the nearest comparable GAAP measures in our shareholder letter released last evening and our filings with the SEC, each of which will be posted on our website at ir.joinroot.com. I will now turn the call over to Alex Timm, Root's founder and CEO.
Thank you, Christine. Good morning, everyone, and thank you for joining us on our Q3 call. During the quarter, we made progress on many of our key initiatives and deepened our founding commitment to building world-class technology that addresses the long-standing inefficiencies and unfairness in the auto insurance industry. We're seeing the movement toward fairness grow, as evidenced by some of the largest incumbents following our lead in questioning the use of credit scores and pricing. We're excited to be leading this charge in the industry and bringing others along with us. This quarter, we placed the Chief Revenue and Chief Operating Officer responsibilities under Dan Rosenthal. In his new role, Dan oversees three critical areas of Root: marketing, insurance, and business growth development. This gives him the ability to drive profitable top-line growth and streamline operations.
With Dan's history of entrepreneurial success, I think he is the perfect person to lead this charge for our organization. I'd now like to walk you through highlights of the Q3. We made progress toward diversifying our distribution channels, an initiative we detailed on last quarter's call. Root's technology enables us to offer a quote in under a minute, providing meaningful opportunities to build differentiated access to customers and broaden our distribution. By embedding our product offering in moments that meet consumers when they need us most opens up a sizable and largely untapped opportunity in adjacent industries. This also creates a vastly superior customer experience. To bring this to life, we focused on building an embedded product that powers our exclusive partnership with Carvana. Since announcing our partnership in mid-August, cross-functional Root and Carvana teams have been hard at work developing a fully integrated point-of-sale offering.
We are currently testing the first iteration in 12 states. Early signs are positive, and we look forward to providing additional detail on results in the months ahead. We have made investments to bring the speed and ease of use powered by our technology to the independent agency channel, giving Root access to a larger demographic of consumers. We have developed a quote and buying process that takes a fraction of the time required compared to other carriers' platforms. This increases productivity for our agency partners. We're currently testing this product in five states with largely positive feedback to date. While investing in the diversification of our distribution channels, we have significantly dialed back our spend in performance marketing, reducing sales and marketing spend by 40%.
This move has greatly reduced our customer acquisition costs, which can be seen in a $45 million improvement to our operating loss compared to the Q2 of this year. This is a prudent use of capital, particularly given the current inflationary loss environment we find ourselves in. We believe the diversified channel approach will enable us to further lower customer acquisition costs, attract customers with higher lifetime values, and provide additional levers to deliver strong, profitable top-line growth. We have substantially improved our pricing and underwriting models. We continued the rollout of UBI 4.0, our latest and greatest telematics model, and MacModel 4.1, our national pricing model. UBI 4.0 is currently active in 20 of our 31 states, representing roughly 70% of our addressable market. This model enables a precise telematics quote, improves overall segmentation, and does more with less data.
Since the rollout of this model, we have been able to extend quotes to an additional 10% of users. We are also in the process of rolling out MacModel 4.1, which leverages Root's growing dataset to expand modeled coverages, improve segmentation, and better predict the lifetime value of a customer. As I stated previously, by growing the dataset, we can accelerate the momentum of our flywheel. As our pricing algorithms improve, we bring down the cost of insurance for good drivers, which allows us to attract and retain a more profitable mix of business. In addition to these segmentation improvements driven by model enhancements, we're also tightening our underwriting by refining contracts and endorsements and leaning on our state management group to unlock state-level efficiencies. We will see the results of these improvements in the quarters to come.
The actions we've taken this quarter demonstrate our awareness of deploying your capital thoughtfully and balancing growth and profitability. We still have work to do, but we know what we need to do to positively impact our bottom line. This is an ongoing process, and we will continue to find ways to drive toward profitability. I'm thankful for the continued support and trust of our customers, team members, and investors. With that, I'll turn the call over to Dan.
Thanks, Alex. Our results for the Q3 of 2021 reflected a decrease in our marketing spend and continued pressure from the loss environment. You'll find our full GAAP financial results contained in the shareholder letter we published yesterday evening, but we wanted to give a few of the key highlights. On the top line, we grew gross written premium 24% year-over-year to $205 million. Our gross earned premium increased 23% year-over-year to $189 million. Our gross earned premium from seasoned states increased to 76% of total earned premiums. The top line's outperformance was driven by the tail effect from performance marketing dollars previously committed.
Overall, for the quarter, sales and marketing expense declined 40% as we took action to reduce cash burn, especially in the face of the surge in performance marketing costs and the current inflationary environment. We continue to expect gross written premium to reflect year-over-year declines in the Q4 and first half of 2022 as we take active steps to reposition our marketing investments, pursue more cost-efficient distribution channels, and otherwise reduce our customer acquisition costs and operating loss. Shifting to profitability, gross accident period loss ratio was 91% for the Q3, a 12-point increase year-over-year against Q3 2020 when compared with the low loss environment last year. More importantly, we recorded a nine-point improvement from Q3 2019, demonstrating improved performance when stripping away the impact of the pandemic, which is partially offset by the current loss cost environment.
The year-over-year increase in the loss ratio was primarily driven by 8 points of severity and 9 points of frequency as inflationary pressures increased costs and miles driven remain above pre-pandemic levels in our book. Leveraging our technology, rating engine, and data architecture has allowed us to respond faster than industry norms with 13 rate increases taken during the Q3 and more planned in the coming months. Direct contribution was a $10 million loss for the quarter. The decline in direct contribution and related margin was driven primarily by direct loss ratio, as I covered above. Operating loss was $127 million, a $45 million improvement when compared with the Q2, driven by the aforementioned reduction in performance marketing spend. Following quarter close, we repaid both our Term Loan A and Term Loan B.
We have signed an exclusive term sheet with BlackRock Financial Management, Inc., on behalf of funds and accounts under its management to put in place a larger term loan facility with a longer maturity than our previous structure. We expect to work in good faith with BlackRock to close the facility before year-end, subject to negotiation and documentation of final terms and the terms and conditions contained in the definitive documentation. Turning to our outlook, we remain on the path we laid out last quarter. We have reined in customer acquisition costs in an inflationary environment and are actively laying the foundation for profitable growth. The work we are doing around cost management is a critical step forward to create the flexibility necessary to invest in and grow our business over the long term. These actions have also given us a more optimistic view of full year operating income.
We now expect to close the year on the favorable side of the midpoint of our original guidance of a loss of $555 million-$505 million. I would like to echo Alex's statement that we have taken a handful of actions that demonstrate our thoughtfulness around deploying capital. It is something we take very seriously, and the effort to improve our bottom line is ongoing. In my new role as Chief Revenue and Chief Operating Officer, I am tasked with aligning our customer experience, products, and market opportunities to drive profitable growth, giving further focus to these key drivers of our customer-centric business. I am excited to continue the work around securing our future.
We're excited about the opportunities before us and appreciate your continued support. With that, Alex, Frank, Matt, and I look forward to your questions.
Thank you. As a reminder, to ask a question, you'll need to press star one on your telephone. To remove yourself from the queue, press the pound key. Our first question comes from Michael Phillips with Morgan Stanley. Your line is open.
Hey, thanks. Good morning. Have you seen any change in the auto loss trend environment from last quarter? It looks like maybe your severity numbers came down a bit. I wonder if that's true, kind of what's driving that and just overall thoughts there.
Sure. This is Frank. Your question was on loss trends. I'd say that we're seeing a number of [crosstalk] Yep. We're seeing a number of factors there. First is we see higher miles driven, which translates into increased frequency. We've seen that in a number of different reports across the industry. We believe that that's an industry thing, not just a Root thing. As we've seen, there's inflation across the United States, which increases overall costs, but in particular in the cost of used cars and used car parts, which is causing a supernormal increase in severity, especially on the phys dam side. These two things together have increased losses dramatically over a very short amount of time. Historically, there have been trend turns in the past, but this is probably one of the most severe trend turns that I've seen in 25 years.
I guess, Frank, I think your wording was severity was eight points this quarter and last quarter was 16 points, so that seems to be down, but maybe I'm reading that wrong. Anything there that looks like it might be improving that?
Yeah. I think we saw a big step function from Q1 to Q2. We are still seeing increased trends, but we're not seeing the same amount of step increase from Q2 to Q3. There are still loss trends. We're still seeing inflation. We see that in the news. I would say we're not seeing as much of an acceleration away in Q3 as we did in Q2, but it's still the inflationary pressures are still high.
Okay. Thank you. You had guided to, you know, last couple calls, you talked about a delay of one of your reinsurance treaties on how that would affect your ceded ratio. It was still down this quarter. I don't think I understand that. I thought it was. Your prior wording was that's gonna tick back up to around 70% in the back half of this year, but it stayed at 55%. Was there something else going on or is it gonna stay this way until 4Q? Just trying to think about how and why it was down and then, you know, what we could expect that to be in 4Q and going forward.
Yeah. Thanks, Mike. This is Dan, and good morning. You're right. Our Q3 cession level remained roughly flat at 55%. That's really a function of the way that our cohort-based treaties work, where, as a refresher, they are multiyear treaties, where the first 12 months we take in new business, and then the reinsurance treaty follows that cohort of new business into additional years. Really just a function of the two treaties we've executed this year on 4/1 and 7/1 depend upon levels of new business. Given the fact that the pullback on the marketing spend, the lower levels of new business that we've talked about, that did lead to the cession level staying flat in the Q3. I'd expect that to roughly continue in the Q4.
You'll see a little bit higher cession rates as we put in place our 10-1 treaty, which is oversubscribed at this point. That's where we stand in terms of cession levels going forward.
Okay. Yeah. Thanks. I guess maybe one more here for now, and then we'll see. When you talk about further reduction in customer acquisition costs, I assume you mean relative to kind of what you saw in 2Q and the kind of the current environment there from what was happening in your direct channel. Do you mean further reductions from, I guess, prior to the spike that occurred in 2Q?
Yeah. No, we mean effectively further reductions in terms of sales and marketing spend on a quarter-over-quarter basis. You'll note that in the Q3, we said what we talked about on the Q2 call. We pulled back marketing by 40% on the sales and marketing side, focused on really utilizing the most efficient performance marketing channels. I think some of that level in the Q3 was that we were talking to you in early August, and we had some dollars that were already committed as the Q3 started. Now we're being able effectively to pull that back even further as we enter the Q4.
Okay, thanks, Dan. Although you're a triple chief hats you're wearing today and thank you very much.
Thanks, Mike.
Thank you. Our next question comes from Matt Carletti with JMP Securities. Your line is open.
Hey, thanks. Good morning. [crosstalk] First question around UBI 4.0, and you talked about how, you know, I think in the past and also in the letter, you know, it enables you to offer a quote with less data. What have you seen in terms of the impact that that's had on test drive periods, as well as the implications that might have for churn?
Yeah. Hi. Thanks for the question. This is Matt Bonakdarpour. Historically, when we first released our telematics scoring model, we were conservative in the way we defined our eligibility rules, which is how much data we need before we can produce a score to price with. As we've collected more data, we've been able to test different models that require less data to see if that degrades predictive power. UBI 4.0 allowed us to really dig deep into that question and do more with less. Users are becoming eligible much sooner in the test drive, and as a result, we can issue a telematics inspired quote much sooner. We're seeing that does translate into better conversion levels. People are getting more telematics inspired quotes, and we'll continue to work on pushing down that test drive period in the models that come.
Okay, great. Along those lines, just hoping you could kind of, you know, process flow. Just somebody who, you know, comes through the app and does a test drive. You know, we think I have a history of knowing how that works. When we think about Carvana or think about your independent agent channel that is growing, how might that work differently when he comes to an agent, you know, do they then have to do a test drive after downloading the app, or do they get a quote that gets adjusted over time? Kind of how does that work differently than kind of the direct-to-consumer process?
Yeah. You know, right now, this is Alex, and thank you for that question. Right now what we are in the independent agency channel, it's still very new, and we are continuing to experiment with different product flows and where we can introduce telematics. You know, we have a very flexible telematics platform which allows us to both collect telematics data upfront before a quote, after a quote, as we also have in our direct model, or even more of a traditional policy over a six-month period, and incorporate that telematics data where we see it to be the most beneficial for us and our customer. We're still exploring in the independent agency channel where we are actually introducing that telematics rate.
In the Carvana channel, right now, and we're very excited about that is an instant quote, basically with the vehicle purchase. It allows a consumer to really just get insurance connected to that vehicle. Right now we are not collecting telematics data through that flow. We do have plans to incorporate telematics in future product iterations.
Okay, great. Thank you very much for the color and best of luck.
Thank you.
Our next question comes from Josh Siegler with Cantor Fitzgerald. Your line is open.
Hi. Good morning. Thanks for taking my question. Have your telematics capabilities and segmentation allowed you to better weather the increased severity and frequency in the market, for example, by identifying higher speed drivers or drivers responsible for more miles? Or do you find that these macro trends are impacting a wide array of drivers on your platform?
Sure. This is Frank. Great question. I'd say first it gives us greater granularity and greater insight into the trends that are happening. Even before frequencies started to go up, we could see that miles driven was going up and could start to prepare rate actions in anticipation of the frequency going up. We have really great granular data to explain and help us understand what was going on. As we started to see frequency going up, other carriers may have been, you know, paused and been like, is this a one-time thing? Is this a catastrophe thing? We could actually tell that it was driving patterns that were causing the increases in frequency, and so could respond, I think much faster than the industry.
Not as much as at a granular customer level, although of course our score to the extent that customers, you know, may be speeding more because there's less on the roads, we would capture that in our score at regular rating time. The big benefit, I think has been the actual like granular insights into what's going on with trends that has allowed us to respond faster.
Got it. Appreciate the color there. You know, just continuing down this road, you mentioned you took 13 rate increases during 3Q. Were these rate increases incremental in nature or perhaps larger step ups to reflect the loss environment? How has the market reacted to them, albeit I understand it's early stage still.
Sure. I'd start also with we had some rate increases even early or late, sorry, in Q1. Based on that telematics data and the increase in driving and frequency, we actually started a round of rate increases in Q1, followed that up more with Q2, and then had the bunch in Q3. We still intend to take more in Q4 and Q1. Some of them are base rate increases, but as we mentioned, we're also rolling out our new MacModel 4, which we think will also give us greater segmentation power as well as helping us kind of capture more rate.
Got it. Thank you very much.
We have a question from Elyse Greenspan with Wells Fargo. Your line is open.
Hi. Thanks, good morning. My first question just goes back to the guidance you guys provided for gross premiums. Did you expect premiums to decline in the Q4, given that we still saw good growth in the Q3? I'm assuming you guys expect a significant decline in your policy count, given that, as you alluded to, you're getting good levels of rate increases across your books. Can you give us a sense of the expected decline in PIFs that you expect sequentially in Q4, as well as just the level of churn you're expecting across your books embedded within that guidance?
Look, Elyse, good morning, and thank you for the question. Given the actions we've taken on sales and marketing, that's why we wanted to continue to provide the color both for the Q4 and the first half of next year, consistent with what we shared on the Q2 call. I think the way you should model it and think about it is that premiums will loosely follow the policy count. We're not guiding to any specific levels of retention. I think Frank just articulated extremely well what we're doing in the context of the macro environment. We're not providing any specific guidance with respect to renewals. You can expect that as we continue to reduce that sales and marketing spend level, there will be an impact on PIF.
Okay. As we think about kind of working through like the frequency and severity issues and kind of your changing marketing spend, should we assume that, you know, if the issues persist right beyond the first half of next year, that you guys will consider to kind of continue to rein in the marketing spend, I guess, as kind of getting through the loss issues impact when you guys kind of return to increasing marketing and looking to pursue greater growth?
Thanks for that. This is Alex. You know, I think when we think about our marketing spend and really our growth drivers over the long term of the company, we're really excited, and we're seeing really strong results already in our embedded product. We believe that meeting consumers where they are in that moment is really going to drive long-term differentiated access to customers and growth. Really what we're doing now is we're making sure that we're preserving that capital, particularly given the inflationary environment that we are in, so that we can continue to develop that product and continue to develop more diversified sources of demand. In terms of timing, I'll turn it over to Frank, who can talk about when we may see our pricing sort of get to the levels that we wanna see as a correction for the inflationary environment.
Sure. As we think about kind of returning to growth, there's two parts there when you think about a rate increase. When you raise rates today, your new business kind of has the full impact of those rate changes, but it takes, you know, time for your renewals to earn in. As we raise rates and take these rate actions, we expect the new business to return to target profitability faster than what our calendar year or total renewal book might be. We might expect our ability to kind of turn marketing back on to be faster than what we might actually see in a calendar year loss ratio period.
We do have some pretty advanced marketing models the data science group has come up with that take a look at a state-by-state and customer-by-customer segment-level profitability. I think, you know, starting in Q1, it's likely that we'll be able to turn on either entire states or various segments on marketing and start to grow the new business again.
That's helpful.
Okay, our next question comes from Nick Jones with Citi. Your line is open.
Great. Thanks for taking the questions. I guess two, you know, one, I guess on the broader auto environment, you know, we've heard that people are actually starting to delay purchases due to how high used car pricing is, and, you know, there's still this chip challenge for new cars. So is that potentially a headwind heading into next year in terms of kind of triggering people to think about auto insurance changes? And then the second question. Oh, go ahead. And I have a follow-up.
We don't see that as a headwind going into next year. We think right now, actually demand, particularly for cars is still very high. Although some of the inflationary pressures that we are seeing are supply chain and chip related, we are also seeing a material amount of it being demand related as well. If anything, we're pretty excited that there is more interest there than historic levels.
Great. On the Carvana partnership, I know it's still kind of early days, but is there any kind of early indication of what kind of attach rate Root is seeing on the offering from you guys?
Thanks, Nick. This is Dan. Alex touched a little bit on this. We are really excited about the fact there's really no more opportune moment to connect with the consumer on auto insurance than at the moment they purchase their vehicle. There's no better partner, no more technology forward and customer-centric auto retailer than Carvana. Their enthusiasm and commitment to the success of the partnership have really been shown by the exclusive relationship, their investment in Root, which we closed on the first of October. Just to provide some more color, in fact, we had more than a dozen Carvana team members in Columbus this very week.
I commend people like Andy Lesko and Chloe Johnston and Josh Brown and Rachel Nelson and all their teams for all the hard work that has us already live testing the first iteration in 12 states. We were excited about the partnership back in August, and our expectations have really been exceeded. Now it's time to deliver a seamless experience and a fair price to consumers. I don't wanna get into specific levels on attach quite yet, Nick, but I would tell you we're off to a great start. We're gonna come back in 2022. We'll talk to you about milestones, timing, and growth at the right time.
Great. Thanks for taking the questions.
Thank you. Our next question comes from Tracy Benguigui with Barclays. Your line is open.
Thank you. Good morning. My first question, this is more of a philosophical question. You set your long-term loss ratio target of 60%-65% when loss cost trends were more benign due to the pandemic. I'm wondering if you may be revisiting that 60%-65% target, since I noticed even your renewed business is well north of that range in the last two quarters within your seasoned state.
Yep. This is Frank. Great question. I don't know that the pricing philosophy has changed a lot. We still wanna price to lifetime profitability on a cohort of customers. As we think about new business coming in, we still would like that cohort of new business to have a 65% target loss ratio based upon, you know, anticipated long-term expenses and so on. Now, that is somewhat balanced by, you know, what the in-force book that we have currently. For both new and renewals, we do want to bring that loss ratio down given the current environment.
I also have a question on your marketing efficiency. You discussed your work with independent agency channel, and I'm wondering if the idea is that maybe, at least in the short term, CAC is lower for agency distribution. If so, how meaningful do you think this revenue source could be in your mix?
Thanks, Tracy. I think we're not gonna get into specific CAC levels by channels. I think the key for us, and the discussion we had over the last several quarters is diversifying our distribution channels. For us, obviously, we were too concentrated in the performance marketing channels the first half of this year. That's the conversation that we had following the Q2 call. We are executing on that with Carvana on the embedded insurance product. We are executing on that with what we are exploring on the independent agent product. We're focused on testing and iterating, focused on diversification, focused on return on capital, across the investments, which obviously has to do with CAC, but also the lifetime value of the customers that come through those channels. We'll come back and talk specifically on each of those channels as that testing develops and we execute further.
This is Frank. I would also add there, as we think about the independent agency channel, it plays well in both being, you know, an expansion into new consumer segments. It also plays to our strength in UBI. UBI is generally not as developed in the independent agency channel. There are a lot of carriers there that still do not have UBI, and there are agency groups that are concerned with that they're getting cherry-picked by a lot of the direct carriers. We do see some demand in that independent agency channel, both for our technology and ability to write a customer in under a minute or two, compared to 20 minutes with some of their other carriers, as well as being able to deliver them a lower rate for good drivers via our telematics. We think that there's good pricing competitive abilities there too.
Yeah, that's interesting 'cause we've actually done some work and reached out to some of the agents that you're, I guess, testing the waters with. One thought that I heard is that they're looking for consistency for their customers. Does that limit how agile you could be in pricing new agents?
One of the things we're doing right now with independent agents, and I think we referenced this a little bit earlier, is to actually make sure that the telematics program works for an independent agent. It will look different than the direct model, where we may be pricing, repricing midterm. We may be using now telematics data at renewal. We also have developed a product that allows consumers to really opt into telematics. We think that using that consumer experience both works well for an agent as well as works well for that agent's customer. That's really where we're starting to iterate on our product and develop a product that really does work for that channel.
Very helpful context. Thank you for taking my questions, and congrats, Dan, on your new role.
Thanks, Tracy.
Our next question comes from David Motemaden with Evercore ISI. Your line is open.
Hi. Good morning. I had a question on the outlook. I understand the gross premiums written are expected to be down in the Q4 and then in the first half of 2022. I just wanted to make sure I'm thinking about this right. You know, you guys obviously had a strong Q3 this year, the $205 million of gross premiums written. Is that, I guess, when we think about Q3 2022, the sort of timing when you guys think you can start to grow gross premiums written again?
Thanks, David Motemaden, for the question. This is Dan Rosenthal. Yeah, I touched on this a little bit earlier with Elyse Greenspan. We wanted to provide color on where we are through the Q4 of this year and the first half of 2022. Given the line of sight that we have to those quarters and given what we've shared on both the Q2 call and today about our sales and marketing timing, the development of the Carvana product and the like, we have a lot we're gonna come back and talk to the market about following Q4 on our thoughts for full year 2022. I'd rather not get into the specifics for specific timing quarter by quarter next year quite yet.
As Carvana further develops in the testing, as we develop the diversification of channels that Alex just touched on, we'll come back and talk to you. Right now we are focused on execution. We are focused on what we call internally, one play, this play, and the play is really focused on a lot of what Frank's been talking to you about on this call, our differentiated technology and how it impacts our ability to get ahead of the loss cost environment. That's our focus right now. We'll come back and talk to you in February about 2022.
Okay, great. That's helpful. You know, this is also somewhat of an outlook question on the 2022 operating loss. But I was wondering if you could just touch on thoughts on the loss ratio over that timeframe as well, not necessarily quarter by quarter, but just sort of thoughts in terms of how you expect that to trend over the next year.
Thanks, David. This is Dan again. Maybe I'll start and then turn it over to Frank on the loss ratio. On operating loss overall, we were pleased to report this quarter that we're expecting to close the year on the favorable side of the midpoint of our original operating loss guidance. That is a function of the execution that we've been undertaking and the dip in performance marketing spend that we highlighted during the quarter. I think it's fair to say that we expect to deliver a meaningful improvement in full year operating loss in 2022 versus the 2021 levels as we continue with that strategy and continue with that focused and disciplined execution. As for the loss ratio, I'll turn it over to Frank to speak to trend he's seeing next year.
Sure. As mentioned earlier, as Alex mentioned, we have a number of opportunities to improve the loss ratio. We started early with rate increases. We've got our new MacModel coming in. We're looking at improving our underwriting and our contracts. We expect meaningful improvement in the loss ratio in 2022. The timing on that is probably less clear. Certainly, as we put in these actions, it takes time for them to earn in, but we're still trying to understand what the macroeconomic trends are. We have, as mentioned earlier, we've got a little indication that it's not further accelerating in Q3 over Q2, but there's still only been a few months to actually have seen those trends. We'll have to wait and see exactly how much loss ratio improvement we'll see as we understand how these trends develop.
Okay, great. Thanks for that. Maybe I can just sneak one more in, just on the Carvana partnership. You know, good to hear about some of the early success in some of the 12 states you're testing in. Do you have a sense for, you know, I guess, when that will be rolled out on a broader basis?
We're gonna continue to develop that product, and as we do, the rollout will continue to accelerate as the product becomes more and more mature, and we continue to get those consumer learnings. We're really excited about this. We think that this is absolutely how insurance should be purchased in the future, meeting consumers where they are in a really seamless experience. We think it's a clearly superior product experience, and we're really excited by it, and we're gonna continue to work hard and focus.
Great. Thank you.
Thank you. Our next question comes from Mark Hughes with Truist. Your line is open.
Yeah, thanks. Good morning. What should we think about in terms of average rate hikes? Looks like your premium per policy was up about 6% in the quarter. Is that a good proxy for what you're putting in place, say, Q4, Q1?
Yeah. I'd go back to the actions that we're taking, right? We do have rate increases. We've got new model segmentations, and we've got improved underwriting. As I think about that, part of the loss ratio improvement is gonna come from just kind of flat rate increases. We also expect that to not necessarily be totally flat, given that we expect up through Q1, maybe 12-15 different states picking up our MacModel, our new MacModel version. The actual increase in premiums will vary greatly by customer segment. I don't know that we've got kind of a target, like, this is how much the average premium is gonna change as much as we've got an idea as to we think we'll get meaningful improvements in the loss ratio.
Okay. When you think about an agent, [crosstalk] y eah, go ahead.
No, please, continue.
Okay. All right. Thank you. When you think about the independent agency channel, how quickly can you roll that out? Are there bigger chunks of maybe agency associations or groups that you can target to speed up your penetration of that market?
Yeah, thanks. This is Dan, and I appreciate the question. I think we're focused right now on testing and iterating the channel. Again, this is about diversification. This is about testing. This is about driving towards the path to profitability. We'll have a better answer to that as we continue testing with the select group of agents that we're focused on right now. Absolutely, we believe that there are opportunities, as Frank touched on, to take our product and leverage it with the broader agency market. We're gonna focus right now on testing, iterating, and executing, and then come back and talk about specifics around growth and timing.
One more, if I might. The repayment of the loans and the signing of the term sheet with BlackRock. Can you talk about the rationale for that and what benefits you think that'll bring?
Yeah. I think for us, there's sort of two parts to that. The first is, we're really excited about the partnership with BlackRock. We talk about in the letter that it's a facility that we think will work really well for us, and we'll come back to the market as we have a fully executed deal in place and provide more specifics. For us, it's about just thoughtful and prudent management of our balance sheet and thinking through our capital overall. That's what we took into account in terms of maturity level or maturity timing of the prior facilities, as well as the partnership with an entity like BlackRock.
Thank you.
Thank you. Our next question comes from Andrew Kligerman with Credit Suisse. Your line is open.
To follow-up on some of the earlier questions, just in the sales and marketing area, after the sharp drop off in sales and marketing costs from $112 million to $65 million in the Q3 quarter-over-quarter. You talk about dialing that down. Do you kind of, you know, kind of prior to the Q4 of last year, do you kinda get into that range of, say, $15 million-$35 million? Is that kind of the range that you're thinking about dialing it down to for the near to intermediate term?
Yeah, thanks for the question, Andrew. I think the way I would touch on that is to really focus on how seasonality impacts some of this. The Q4 we've talked about is both traditionally the most expensive quarter from the marketing channels that we have been in, as it's a retail focus quarter approaching the holidays. It's a quarter when, just frankly, consumers focus a little bit less on things like car insurance, and are focused more on their other purchases approaching those holidays. For us, seasonally, we would already approach performance marketing in the Q4 differently than in prior quarters. That is highlighted this year by the fact that we are pulling back in some of those more concentrated marketing channels and focusing on the most efficient channels.
You can expect the level to come down materially in the Q4, and we're focused on really investing our dollars in the right channels, and taking it from there.
Okay. No numbers around that?
Sorry, no numbers.
You know, on the last question with regard to independent agency, and again, you know, you're not gonna focus exactly on the numbers you said, and you're also looking at testing and iterating here. You know, as you think about that channel, the independent agency channel, is it gonna be sharply less expensive than the digital? You know, is it a situation where you can go into a large brokerage firm and you could get up and running pretty quickly? I just wanna get the sense of the magnitude of the potential upcoming spend. Is it materially less than what you need for your other channels?
This is Alex. In terms of the cost relative to the direct channel, I think what we see is it's just different. It's not necessarily less or more. You know, in the independent agency channel, we will be paying a commission. Instead of a flat fee to get a customer, as you see in the direct channels, we will be paying a commission based on the premium that comes through the channel. We like that. We think it's very cash flow efficient, because you do not pay all of the customer acquisition costs upfront, but rather you get to really pay that over the life of that customer.
We also know that there's very high retaining customer segments that still go to independent agency channels. We've seen that really persist. You know, a decade ago, the independent agency channel had 1/3 of all volume. Today, it's got roughly 1/3 of all volume. We believe that using our platform that allows us to deliver a quote in under a minute to those customers is really valuable for the productivity of agents. We think that it will have favorable unit economics as well, both in terms of retention and then again, that favorable cash flow profile to customer acquisition costs.
Helpful. Thank you.
Thank you. Our next question comes from Mike Zaremski with Wolfe Research. Your line is open.
Hey, this is Charlie on for Mike. Good morning. Wondering if you can provide some color on accident frequency and miles driven levels you're seeing. Has that relationship changed at all versus pre-pandemic levels? Is it consistent across the states, or does it vary? Did it vary across the months of the quarter? Any views or themes you could share there would be helpful.
Sure. This is Matt, thanks for the question. Talking about this year, we did see an uptick in accident frequency going into April. Since then, it's been a pretty consistent trend month-over-month up till now. We looked at time of day frequency. We have seen, as we've come out of the pandemic, an increased frequency both from vehicle mileage and accident frequency in the weekday mornings from, say, 6 A.M. to 10 A.M. These are in line with pre-pandemic levels. In terms of accident frequency within Root, because of underwriting and pricing model changes, we have seen decrease in frequency as our population has mix shifted into lower frequency customers. That's driven by not only underwriting, but telematics and other pricing model changes. Overall, we're seeing a lower frequency average on our book. Within this year, we are seeing the uptick going into April and beyond.
Thanks. That's helpful. Oh, sorry. For my second question, can you guys, you talked earlier in the year about raising rates in Georgia. Can you talk about what you're seeing there? It looks like it didn't slow your growth. Is Georgia a seasoned state now?
In Georgia, we did have a rate increase there. We are rolling out a new program. It's not fully in market yet, so we haven't fully turned Georgia back on. We do expect to in the Q4.
Okay. Thank you.
Thank you. Our next question comes from Chris Martin with KBW. Your line is open.
Hey, guys. Congrats on the growth in the quarter. I just have a couple follow-up questions on the Carvana piece. And the first one is, and it's really admirable the work you've done around, which, yeah, kind of bringing out the credit score from pricing. But if you are not going to be using the telematics piece in the Carvana side, is that going to be back to kind of a traditional pricing model or have you built something else a little bit different to, you know, build out that really kind of the immediate quotes?
We think that there's actually a big future and a differentiation in the pricing model for Carvana. It's not just a traditional underwriting models. One of the things that's a very interesting opportunity that we are exploring is when these vehicles are sold to consumers, they're actually in the possession of Carvana first. That gives us actually an opportunity to get closer to that vehicle and closer to the technology inside of that vehicle. As we continue to really expand this partnership and continue to do more product development, we fully expect to bring in more data science capabilities and more unique data sources, both from mobile phones and consumers and then also from the vehicles themselves.
Okay, great. Yeah, that makes sense. Just with like all the chatter, I may have missed this. My three-year-old kicked in the door about 10 minutes ago. On the independent agent side, will getting your quote from Carvana trigger anything to then be placed as an independent agent if you're in the right place, or will that be a fully direct business on that end?
That'll be fully direct business.
Okay. Thank you.
Thank you. We have a question from Ryan Tunis with Autonomous Research. Your line is open.
Hey, thanks. Following up on some of the independent agency questions, and my understanding has always been that table stakes for writing in that channel is, you know, like an A-minus financial strength rating from AM Best or S&P. You can maybe get away with Demotech if you're a coastal company. Could you remind us, you know, who's giving you your financial strength rating now? If you don't have an A-minus from AM Best or S&P, have you thought about trying to go out and get one?
We are not rated right now. We have not found a strong demand from any of our agency partners actually that we have spoken to. I think primarily because it is personal auto, and so typically that's not something that consumers themselves also are really demanding. We do not have a financial rating from AM Best or Demotech at this point. We have not seen that at all be a blocker to our partnerships with agents.
Got it. My one follow-up was, I guess just maybe looking for some color. I get what you guys are doing with the performance marketing pullback, which makes sense. Are you seeing that, you know, some of the advertising like are ad dollars becoming arguably a little bit more efficient given the pullback of a lot of carriers? Are you actually, I mean, theoretically seeing that if you did want to spend, you could maybe do it cheaper if you wanted to at this point? Is there a theme there?
Yeah. Thanks for the question. This is Matt Bonakdarpour. We are definitely seeing an improvement in efficiency on back of the pullback marketing spend. It's not just a pullback, but also more sophisticated segmentation and targeting to ensure that we are allocating our marketing budget in segments that are providing profitable lifetime value. At the new levels of marketing spend, we certainly are seeing return on that spend and more efficient spend as a result.
That's good to hear. Thank you.
Thank you. We have a question from Youssef Squali with Truist Securities. Your line is open.
Oh, great. Thank you very much. Just want to follow-up on a couple of questions my colleague Mark asked earlier, but just kind of maybe at a higher level, Alex, stepping back. A few years ago, you guys were one of the very, very few companies kind of chatting telematics and all the virtues of it. It seems like on the one hand, more and more people are or more companies are talking about it. Yesterday, we saw a deal between Lemonade and Metromile to kind of double-click on this. Your Carvana with its instant quotes seems to kind of move away a little bit from it, even at least at a first glance.
Can you just maybe kind of describe how you feel the industry is evolving around telematics and just the importance of it today versus how you thought about it maybe three years ago or two years ago? Second, maybe just give us an update on your campaign to Drop the Credit Score and pricing and any kind of traction there. Thanks.
Absolutely. Thank you for that, question. What we're seeing is that telematics, as consumers become more and more comfortable sharing data, that telematics is becoming more of a wave of the future. I think we have seen several points of validation across the industry. That being said, we also know that we are at the forefront, and we know that for a few reasons. You know, one, we are entirely focused and committed on disrupting and using new data science methods to disrupt the United States personal auto market, which is a $260 billion market. We have all of our resources, all of our investments, and really have for the last six years, entirely dedicated and focused to building the best technology in that space. We've seen it.
Carvana, we believe actually is just the next evolution of the technology. We believe that as we've developed this technology that is highly differentiated, that now allows us to embed into other vehicles and to actually ingest even more unique data sources. We do not believe that the evolution of data science and insurance stops at the smartphone. We think the smartphone is a pivotal piece of the puzzle, but we don't believe that's where it stops. We think we've demonstrated that, you know, we are now in more states really than any of our insurtech peers. We are also, you know, multiples the sizes of most of them. We believe that we are out at the forefront.
We think telematics is the wave of the future, and we're excited to be leading that wave and to really have differentiated technology based on rich behavioral mobile data there. In terms of Drop the Score and Drop the Credit Score, we also know that when we look at the industry, that consumers aren't getting a fair shake, particularly those consumers that we believe need it the most. We think that there's lots of discriminatory factors that do go into car insurance pricing today. We have decided to launch a campaign to really challenge the industry. We're out in the forefront here to Drop the Credit Score. We're really excited. Actually, a major incumbent has followed us into this movement and is following our lead. We are incredibly happy with that.
We've also now started to see progress at even the National Association of Insurance Commissioners level, where they are now considering laws and model laws for how to handle discrimination in car insurance. Those were things that did not exist three years ago. The industry is changing, it is moving, and we really know that we're at the forefront.
All right. Thanks. Good call.
That's all the time we have for questions. This concludes today's conference call. Thank you for participating. You may now disconnect. Everyone, have a great day.