Well, I'm very excited. Today we have Daniel Schreiber, Lemonade CEO and co-founder, and Nick Steed, Lemonade's new CFO. Thank you so much for taking the initiative to do this interview and coming on, Daniel. This is the third time we've done a little talk like this. Every time it's been so valuable to me and so many others, the detail and color you provide. So, thank you both for taking the time, initiative, and for being here.
Right back at you. We're both big fans of the Paper Bag, the depth and thoughtfulness of your analysis. Truly, it's something that we marvel at. So, thank you for your close attention and thoughtful commentary. Great to be back with you.
Thank you. Let's start with the most pressing topic on investor minds, which is operating expenses, especially excluding the growth spend, that portion of it. And years ago, Shai wrote about the autonomous organization. The idea that if you build a company right, you use AI to its fullest form, you can scale premiums with very little growth in operating expenses. For a long stretch, you delivered exactly that from Q3 2022 to Q2 2025. OPEX including growth spend was around 80 to 90 million a quarter, even as your top line was just cruising, was soaring. But lately that number has crept up. In Q2 it stepped up to 117 million, which is— some of which is higher stock-based compensation and also some of which is interest from the synthetic agents program as that is scaling up. But even if you back those out and you adjust for an employee retention credit from Q2 of last year, then those OPEX expenses excluding that growth spend still went from around 72 million to 86 million or about 20% year-over-year. So how should we think about operating expenses excluding growth spend going forward? What's the outlook for its growth rate? Is that something that can stay under 10% on an overall average? Some quarters are going to fluctuate higher or lower, but how should we be thinking about that?
Sure. Happy to spend some time chatting this through and I know this is a topic of great interest and I know both you and Matt Smith and others have begun to really peel this back. So I'll try and give an overview, some of which will be repetitive of the work that you've already done. And I'll just give a quick promo because the autonomous organization that Shai wrote about almost 10 years ago— I think he first wrote about this in 2017, if memory serves— is alive and kicking. And the premiere is for Investor Day where you'll really see behind the scenes and what that looks like. And I think there's a lot of exciting things to look forward to then. So I think the thesis— I'll start off by saying my conclusion, which is the thesis is robust and nothing much has changed.
In general— I'll kind of interject and then I'll come to your question directly, but my own view, and this was kind of obviously not the more broadly adopted perspective, but was that Q2 was amazing. It was a very strong quarter. We had a beat on every metric that we guided to and around street consensus. So we didn't miss anything and we guided up on a couple of things. And even— and this is kind of what's behind your question, which is about efficiency— we gave some insights into efficiency on our investor deck on our website. You can see IFP per employee or you can reverse into it. And we spoke about LE and Jay gave some numbers which are to my mind extraordinary and very, very telling. So having said all of that, I totally understand the concern. Total operating costs rose by 41% in a quarter that grew 32 and a half percent. So operating expenses outstripped growth. And even if you back out, as you say, the growth spend, which we've always done in these kind of conversations, the remaining opex grew like 48%. So there was really substantial growth, faster growth than the business grew. And that begs exactly the kind of question— if we've spoken in the past about mid-single digits and you're asking now could we stay under 10%, spoiler: yes. I'll come back to that, subject to the stuff that we'll discuss in a minute. But it's natural to ask questions on a quarter like this. So you want to know, does such opex growth undermine the operating leverage thesis? You want to know whether it's great that we're going to be adjusted positive just a quarter from now when we're recording this, but is that masking accelerating costs that could hamper whatever investors have modeled out over time? And then I think there's also a secondary or tertiary concern, which is: is there some lack of transparency here? Are we obfuscating by changing metrics midstream? Are we— you know, there's a chart that was there, now it's not there, and what's behind that? And as I say, I think that's probably secondary or tertiary, but I'd love to get a chance to address that straight on as well. So, let me— actually I want to share a waterfall chart that will just help me talk this through, and then I'll use some visual aids. So, okay. Hopefully you can see this just fine.
What you see on the left, the purple horizontal line is the total increase in opex. And these numbers, I've rounded them and I've used kind of public disclosure. So hopefully this matches everything that's broadly familiar. And then as you start going down the waterfall, you say, 'Well, what is that $53 million in added OPEX? What is it comprised of?' And then you go from left to right and you can see all the elements. And I'd love to talk you through them really quick. So the first is the one that you already kind of discounted, which is the growth spend itself. And obviously, as we grow our business, we're going to spend more. It's good news that we're able to spend more at a strong LTV/CAC that has remained very robust despite spending more. And every dollar that we can spend with a 3:1 ROI, we should be spending. So you want to see us spend as much as possible with these kind of ROI— tells you nothing about operational efficiency. We do tell you a fair amount about our marketing efficiency and we have a whole section in the letter about that. But this is obviously money well spent and it's not the money that you're asking about. But just so we see the full stack.
And next along is an interesting one. It's the interest payments and it's been growing. We have an off-balance-sheet arrangement. We call it the synthetic agents where we use somebody else's balance sheet to finance all of that growth and we have to pay them for that. It's a very, very profitable arrangement. We take money at an IRR traditionally of 16% and it's dropping down to the single digits as you know in the newly negotiated deal and we're generating CAC IRRs in the high double digits or low triple digits even. So clearly this is a very, very smart investment and I think categorically you should think about this as part of our growth spending. It is just one of the associated expenses to do with growth— we have to pay for the money with which we grow. So even though it doesn't appear under that same line, we put it under G&A. Other companies wouldn't put this in G&A at all. Interest payments often appeared in an entirely different place. But because we regard the entire synthetic agents arrangement as a core part of our operating business, we include it in here and I think that's appropriate. But in terms of the categories— if you're looking for a signal of are we failing to scale, this does not give you that. This is clearly something that is part of the growth spend and as I say, a pretty profitable and compelling arrangement. So you want to see us there as well, do more of that. Parenthetically, it's also not long-term. Eventually this goes away. In the meantime, the interest payments as a percentage at least are declining because we're managing to get to better and better arrangements. So we spoke about that. That was covered in the letter as well. So I think everyone should feel pretty good about these $3 million as well. Not a signal for the concern that you raised.
The next one along is even more easy to dismiss because it's just a pure one-off thing. We didn't fire people during COVID. And the US government a year ago gave us an employee retention credit for COVID. It's a one-off thing. It benefited our opex last year. There was no such benefit this year. So that's clearly just a phantom growth number. There was no real growth. There was just no ERC credit this year and there never will be again. So I think that one also is fine to dismiss. Now look, we've just spoken through the biggest chunks of the jumps, right? You've got the growth spend with its growth funding is 40% of the growth. The ERC is another almost quarter of the growth in the OPEX.
Then you get to stuff that's a bit more interesting and that we should dwell on a little bit more. Other insurance expense at 10% of the growth is an important one. So what is this? This tends to be the kinds of things that we have no control over such as processing fees for your credit card. We use Stripe or others to clear credit cards. They get paid. Credit card companies get paid. So you've got a couple of percentage points that are always going to go to processing fees. Scale doesn't much help with this. Maybe you can shave off a fraction here or there, but at first approximation, this is something that will keep growing. The more IFP we have, the more premiums we collect, the more processing fees we will pay in dollar terms. Expect this to grow. The same exact thing is true of premium taxes. Premium taxes is a kind of value-added tax that all insurance companies have to pay. All regulators collect these. They're in the low single digits and they scale linearly with premiums. There's nothing to talk about. It's just part of the premium goes in taxes if you like and that's just something that is going to continue to scale. No signal here about operating efficiency. This is just the taxes that we pay.
We do have a small portion of a business that is sold through agents. Those commissions appear here as well and that might fluctuate up and down a bit as we use more or fewer agents in different arrangements. But at first approximation, I think it's perfectly fair to say that all of other insurance as a category says nothing about operational efficiency. It is just inherently volume-linked. And something that is again, if you're looking for signal about the autonomous organization thesis, you're not going to find it here. But, and this is important, it has certain implications that we best be aware of. One is that it will grow and if you want premiums to grow, expect all these associated expenses to grow more or less linearly. Actually, we grew less than linearly. IFP grew 32%, this category grew like 25%. But again, rounding up, assume that this will grow in line with premiums over time. And don't find that startling. It is absolutely to be expected. It is normal and it again says nothing about the hypothesis or the thesis or the promise of an autonomous organization. If we had zero employees and zero operational expenses in the sense that we as operators use the term, this would continue to grow.
An important artifact of that is that if we hold all the rest of our true operating expenses kind of flat or single-digit percentage increases and other insurance expenses grow linearly with the business, call it 30-something percent a year compounding, what follows from that is not only that you will see this category continue to balloon, kind of quote unquote, but it will track our ballooning business. It's going to be very similar in its trajectory. But it will take up more and more of the operating expense category because the others will be relatively stable. This one will be relatively unstable. It will scale— not at all, it will grow linearly with the business, which means that this will become a greater and greater percentage of our operating expenses. If you followed that logic, as we start thinking about forward-looking, what can we expect, to your question, will it be single digit, will it be 10%, will it be whatever— ultimately this thing is going to push into the double digits because even if we are a million percent efficient, even if everything is automated— especially if everything is automated, this will then become the overwhelming portion of the operating expenses ex-growth. And it will be growing at something that will track to the growth of the business. And as we— I'll plant that seed now because as we come back to the question later of why are we switching graphs and what is the best way to think about that, this is something to be aware of that over time you'd get a false signal. It would be noise but it would appear a signal because you'd start seeing operating expenses creep up. The things that you're asking me about could stay damn straight or decline and you would still see operating expenses climb because of this line. I'll come back to that a bit later on.
Next up is tech. So I'm not talking about the SBC at the moment. I'm treating that as its own— stock-based compensation, treating that as its own category. We'll come back to that in a minute. Ex-SBC, without looking at that, you're talking about something like a $5.1 million growth in tech investments. About half of that is headcount, about half of that is tokens and other kind of associated expenses. The first thing, just to put some dimensionality about it: this is a small fraction of operating expense. This is really not a big deal. You know, it's 1 point something percent of our opex or whatever it is. It's not very— the headcount piece of it, but let's talk about the headcount for a second. We're talking about something like a 5% increase in headcount year. I have to tell you that we have an incredibly stringent process around hiring. We scrutinize every single hire several times over. We hire people when we think there is an absolutely compelling ROI. We are working on some very cool stuff and we need some really smart people to drive it and they use AI to amplify their abilities in powerful ways. But we're not yet in a humanless future. So highly talented people help us do incredible things. But my first comment is rest assured that each of those people are going to— if you like the 3:1 ratio of our growth spend, I think we're getting much better than 3 to 1 on every headcount that we hire.
The other thing to say though in terms of the thesis that you're asking me about is that you got to expect modest fluctuations quarter to quarter or year to year. Zoom out. Four years ago in Q3 of 2022, we had 1,379 employees, about 40 more than we have today, and our business is two and a half times bigger now than it was then. The autonomous organization hypothesis or thesis is robust. It is secure. We have fewer employees now than we had four years ago. The fact that within that time we've had a certain wave of ups and downs is to be expected and I want to explain to you why.
Our workforce is continuously shifting. It's not that we don't hire people. We have open positions all the time and we also have people leave us all the time. We have thankfully, mercifully been able to avoid big waves of redundancies. We've not had to do that. It has been natural turnover and that's very good for the health of our organization and for motivation, etc. But we have been not only replacing people but replacing entire functions and shifting people from some departments to others and we always are kind of prejudiced in favor of product and engineering building the core. If we can have fewer people doing customer support and claims because of the automation levels and we can recycle those freed-up headcount in order to accelerate our development and ambitions, we're going to do that.
But the timing of neither one of those things is perfect. People leave when people leave. And people accept job offers when people accept job offers. And therefore, you should just understand that there's a certain fluctuation that is entirely within the realm of just normal ordinary course of business. Every hire is scrutinized. We're very strict about that, but we're not very strict about the timing. If when we need somebody and it makes sense, we will hire them and other people will leave. And you're seeing that over the last four years, this has netted out as a smaller team over time with a business that was then $600 million in IFP and is now at $1.5 billion and yet we have a smaller team. I think we can all feel very comfortable that this is— even a blip I think overstates it. This is 5%. And let's just kind of leave it at that. On a forward-looking basis, I think Nick already referenced this. We're not anticipating any big hires. We did some of the hiring that we anticipate. So I think roughly where we are now is roughly where we'll stay for the next while as far as we can see. But I'm not making promises. We'll do what's right for the business from time to time. And as you can say, even this quarter with big hiring, relatively in the scheme of things, zoom out a bit and nothing really happened.
Which brings us to G&A. G&A had a dramatic headline because it reported what appeared to be an 85% increase in G&A. So that looks alarming. But all these other items that we're talking about, SBC on the one hand, ERC on the other, they all fold into G&A and we've already accounted for those elements. And once you take out the SBC, the interest expense, the ERC, each one of which we've spoken about, you're left with a 1% change in G&A. Nothing happened in G&A. The headline says 85% growth. The reality rounds down to zero growth in G&A. Which leaves us with one piece. I don't know, should I stop and do you have any questions on that before I plow into SBC or should I keep going?
Oh, that's really— everything's really fantastic. I think it's very helpful to think about those costs that no matter what are always going to stay lockstep with your top line, because that could become a developing story in a few years from now where your other insurance expenses, the biggest category or growing or significant category that keeps growing, well, simply maybe because of credit card fees that are always going to be there. No, that's super, super helpful. One detail was within that sort of fixed cost would be tokens spend, AI spend. I know a few people are bringing that up when it was crowd-sourcing questions. How is that measured? Is that a significant cost? Is that a concerning cost? Because some we don't really know from the outside and some people— that was their quick explanation: Oh, AI costs are just through the roof and they're spending a fortune. Can you speak to that specific cost for a moment?
I'll give you my perspective and then Nick, feel free to pile in if you've got anything to add. Not everybody agrees with me, but I'm entirely unconcerned about tokens. In the first quarter, to some extent second quarter, the industry at large went through silly things— token maxing. Some companies prided themselves on how many tokens they burnt through. Companies, us included, became much more sophisticated about model selection and having an orchestra of models and an orchestration layer that will choose the appropriate level of intelligence for a given task. You don't want to be doing everything with Fable. But the broader reason why I'm absolutely unconcerned is the cost per token for a given level of intelligence, a given intelligence density, is dropping by about 10x every year. That's just been true for a while. With the trillions of dollars being spent, I envision that being true for a while. I think ultimately the cost of tokens asymptotes to the cost of electricity. And therefore, yes, as these models still quickly evolve, the absolute frontier models with tremendous intelligence density are very expensive. But already now you go back, you can use a plethora of open-source, open-weights models. You can use slightly older versions of the labs' models and you get to very high levels of efficiency. So I do think that over time some of our headcount will shift into tokens. You can start thinking over time about a budget on intelligence rather than on AI and humans and you can start thinking about different framings. But my short answer to your question is: yes, it will grow as a cost. No, it's not going to blow up the model. Nick, anything to add on that?
No, I totally agree with that. And yes, the spend towards tokens is increasing in recent periods, but it's a pretty small percentage of that total expense base. And we do track that spend very tightly across all use cases. Internally, we have customer-facing use cases and internal use cases. And across all, there's a significant ROI orientation across each and every use case. And what we see currently, we really like.
That's great. Yeah. I'll mention too— Daniel touched on it when you're talking about hiring, and it's a bit of a blip this quarter of oh the hiring's going up, but actually yeah, like you said, zoom out even a little further. You were at the same level or higher previously. And if you look at in-force premium per employee, that's still rising. And I was thinking about even at the investor day of the last one, 2024, Adena shared this goal or aspiration of 4 million of IFP per employee when you get to 10 billion roughly, maybe when you get to 10 billion total. So that's 2,500 employees. So it's almost doubling your size from between now and whenever that is— I could estimate that might be 2033 or something like that. So that you're actually even gaining about 40-some employees per quarter is on track with that ambitious goal. So it makes sense and if you can beat the goal even better. I mean I know you guys if you could you would. So that gives us some perspective that there's going to be times where you're going to hire and maybe there's 100 people or 150 added and then other times it's totally flat for periods. So anyways, that was very helpful. Just myself thinking about that bigger goal and your commentary there too about how stringent you are in the hiring process. That's something I really appreciate about Lemonade. You are very cost-focused and centric and you're not just going to maximize tokens because that sounds good that you're using lots of tokens, which some management and some companies might think that's a good idea. It sounds good because you're using lots of AI or something. But you're the complete opposite. So yeah, appreciate that. If you want to dive into stock-based compensation.
Absolutely. So a couple of broad comments and then I'll talk about the specific grant that Shai and I received as well. First, in terms of your broad question about scaling and linear growth versus the other kind of efficiencies or the scalability of the business, SBC, stock-based compensation, which means the GAAP accounting treatment of allocations of options and RSUs to our employees— that has trended rather favorably compared to any metric that you want, whether it's IFP or revenue or gross written premium. So you're definitely seeing the growth in SBC, a fraction of the growth of the business as a whole. So it's certainly, even with the changes and the surprises, it's done no violence to the basic hypothesis that you're asking about. And forward-looking, and I'll come back to this maybe in a minute, but we don't see any big— we have seen a bit of a step function change. People wised up to it or focused on it in Q2, but actually we disclosed it already in Q1 and it was curious that it went unnoticed then and suddenly people kind of spoke about this, but we guided to this a quarter ago so nothing really— it didn't surprise us, put it that way. It was in our disclosures and in our guidance and of course every allocation that we give to senior executives like myself is disclosed in real time within 48 hours so all the information was out there. We spoke earlier, we have no plans to do any big hiring sprees. We also have no plans, and I'll elaborate on this in a second, to give any big option packages to senior executives. There'll be some ordinary course of business, but nothing that would register on the horizon, so to speak. So I think that once we lap this year, you know, next quarter will still be compared to a year ago and that will be true for the quarter after that. But once you lap a year from those allocations, you'll see the growth flat out more or less or go back to the kind of single digit that you were asking about. But do anticipate for the next couple of quarters to see something similar to what you saw in Q2. Nothing new will have happened. There will be no new information, no new allocations. It's just the lapping needs to take place before you see the growth stop.
I think in general— I don't want to dwell on this and every investor will take their own perspective, but it's an interesting question how much investors do or should care about stock-based compensation. Clearly more than zero, but how to think about this. So a lot of times growth investors are EBITDA focused. EBITDA exists in large part in order to neutralize SBC and say I want to look at the underlying core business, not kind of option allocations, and that's fair enough. And more traditional investors perhaps, or value investors and others— you know, Warren Buffett famously, but he's not alone in this— I think the majority of investors say no, no, no, SBC is the same as cash. If you have to give Daniel this much value, I don't care what currency you're paying it in, that's a cost to the business. And I totally respect that. For what it's worth, my own vantage point is more dilution-focused rather than GAAP accounting. So we have seen dilution hover at around 2%, 2 point something percent on a multi-year basis. Not every year is the same, not every quarter is the same, multi-year basis since the IPO. And I'll tell you right now that on a going-forward basis, I think that remains true. You'll see some ups and downs, but I think likely as not, the multiplier begins with a two for the foreseeable future. And that compares very favorably to the industry at large and to competitors. And I think ultimately GAAP accounting treats the value based on various measures— Black-Scholes for options and Monte Carlos for other kind of equity grants, and we'll come back to that in a second. But I think if you look at the dilution as the true cost to an investor, you'll see that our SBC allocations, GAAP accounting notwithstanding, have actually been very efficient.
A word more before I kind of jump into the specific grants. We as a company and I personally really like using stock-based compensation significantly at this stage of a company's life. I think it gives you several things that are all investor-friendly. So I know investors say, 'Oh, you're causing dilution.' Cash has its own impact on investors, of course, on earnings. So if you're looking on an earnings per share basis, I prefer stock-based compensation to cash compensation even if you treat them in GAAP accounting as entirely the same. One is I think it aligns management and employees more broadly with investors. You don't get some kind of a bonus that could be gained, you know, some metric that could distort value. You start getting unintended consequences when people are striving for a particular metric and they can sacrifice other metrics. You're saying employees and executives are going to be aligned with investors. If it's good for the investors, ultimately it's good for them, and if not, not. So I like that aspect.
I think there's a cool kind of arbitrage, which is we expense these options the way we need to under GAAP accounting. So you'll use Black-Scholes or whatever the metric is for whatever instrument or equity is being granted. But unlike cash where everybody views a dollar as a dollar, here it's very much in the eyes of the beholder because GAAP accounting is strictly regimented. But what happens in an employee or executive's brain is not. And if the specific employee is very bullish on Lemonade and they believe that this is one day going to be an incredibly valuable stock, then in their mind the psychic value of the option far exceeds the GAAP accounting treatment. So you get far more attentive value for a dollar of GAAP expense than you would if you just gave somebody cash. And I think that is the case. It selects for people who really believe in Lemonade.
It means that when you give them a dollar of SBC, you may be buying $10 of retentive value because they think the company's going to 10x over time. And therefore, I really like it from that perspective. I think it actually is less dilutive in terms of EPS and you can get more bang for your buck as an investor and therefore we actually skew quite heavily to SBC. The overwhelming majority of my compensation is in stock-based compensation, and that's true for others in the company as well. And I would like to think that it is a strong signal to investors that the senior leadership is willing to take so much of their compensation in the same currency that you carry as an investor and therefore I hope it also sends a reassuring signal.
Now, specifically to the significant grant that Shai and I got, there were a couple of other grants to senior executives, but the ones that Shai and I got rightly get the focus more than the others. Important to note, I have not once in the 11 years that I've been in this job, never once asked for any compensation from the comp committee. I've seen some people saying, 'Why are they paying themselves?' That's not the way this works. We're a public company, a financially regulated corporation. There is a compensation committee of which neither Shai nor I nor any other executive is a member. They are the ones who determine compensation.
We have a methodology across the company, which is we try to run every now and again, preferably annually, a survey to see if we are paying our employees on par with what they would get outside. We don't wait for them to complain. We try to give them a compensation that would align them and give them no good reason to leave the company. That same methodology was applied to Shai and myself, and the grant that was given was a consequence of that same methodology being applied. So there's nothing particularly unusual. This is a pretty professional process of looking at what could Shai and Daniel get elsewhere. Frankly, we actually got significantly less than the initial survey results were. We were very comfortable with that, and we were given a very long vesting schedule. Vesting schedules tend to be four years. This was an eight-year vesting schedule. This is something quite extraordinary.
And the board set the expectation with us that there may well be no major grants to us for years to come. So this was kind of a special one-off grant that's meant to last for several years. There's no promises that there won't be any others, but that is the expectation that was set. So just important to understand the process and how we got to these numbers. We do have senior leaders who companies like OpenAI and others would love to headhunt, and being sure that we and our executive team has been incredibly constant and our tenure of our employees is very good, very unusually good and long. That is hugely valuable to the business. You know, you've been around for a while and you see the same faces again and again, and that is just incredible because the collective learning and the muscle memory and the team dynamics are preserved. So these are important dynamics and I think our compensation committee is doing the right thing.
So on a forward-looking basis, I think this is a one-off or something kind of like a one-off in the sense that it will continue to be accounted for. You will still see it next quarter and the quarter after and two years from now and four years from now, but you won't see a step up. And therefore, once the anniversary passes, this will stop being a growth factor in the operating expenses. And yes, this is a category that will comport with your question about kind of single digits versus this kind of growth. So if I try to kind of step back and summarize it all, I totally understand the concern. I have to say going into the quarter, we knew that there was an issue. Nick and his commentary put a few sentences to try and give it perspective. We didn't think it would loom as large as it did. Since IBITA and everything else was better than guided, we thought that that would give the signal and the reassurance that people wanted. So I'm sorry that we misjudged on that.
But truly once you unpack it, I think there is nothing here to give you or any of our thoughtful investors cause for pause. The hypothesis is rock solid. Everything is really as it should be. And if you want, we can talk a little bit about the appearance of lack of transparency as well. Happy to address that if you'd like.
Yeah, I know that and that all is super helpful. Yeah, I think with the stock-based compensation, I think it is a good strategy because it is aligning interests and I'd rather see that than if it was just simply paid as a cash payment. I think that's much better. Um, yeah, with there we could dwell on that for just a little just a moment longer. The and I don't even and I think these are fair and my own assessment of the grants when I looked at them versus competitors' other CEOs was that these are, they're not unfair grants or something. You know, we weren't as investors giving you guys the moon or something for no reason. But there are criticisms. People like to bring up are that these came now before profitability was here. You know, is that too soon? This could have been delayed. And the other one was people say, 'Oh, why couldn't they have been more performance-based?' You know, I know there's an aspect of them that was performance-based. And I mean my own sort of devil's advocate to that criticism would be that you're still getting shares no matter what and that aligns interest. You know, if the shares, that no matter what that aligns your interest in the long-term performance. So I don't know what if either of you could share those more to that and those criticisms, the timing of these or could they have been more performance-based? Even how that sort of decided or chosen.
I'll share my, you know, I talk about everybody else. The board determines my compensation and it does in a room that I'm not present. So I don't want to speak for them too much, but I do oversee the compensation of the rest of the organization and we don't have performance-based bonuses. And we give people good options way before profitability. We've done it for 11 years.
So I can talk to the broader principles. Profitability is super important, but it's a point in time. We get to EBITDA profitability next quarter. We're recording this in Q3, Q4. We're expecting to be EBITDA positive, and that is super important. We've spoken about it for a long time, but honestly, we've been trending towards it and then we will trend beyond it. And it's a point in time and nothing actually changes for me in Q4 of this year. And we don't need our team to do anything different in Q4 of this year versus Q4 of last year or the year before or the year before. We need them highly motivated, working hard, increasing efficiency, satisfying customers, launching products. And profitability is a byproduct of them all being highly motivated and doing the right thing. So I don't understand. I know it exists and there may be a good explanation for it, but I don't understand the philosophy that says wait for profitability. We need highly motivated people to get to profitability.
How do you think we'd get to profitability if we didn't have a team who are highly motivated and hardworking for 11 years? So, I'm not sure I followed that line of logic. At any rate, it's not what we've implemented from day one 11 years ago. We've given people a stake in the business and that's how we get to where we get to. And we will continue to do that when we tick over into profitability as well.
I will tell you, and we don't have to delay on this long, but Lemonade has an unusual compensation principle when it comes to performance-based. We have zero employees who get bonuses based on performance. All of our variable compensation is stock. It's very unusual. Some of the investor advisory firms ding us for that. They seem to think it's a bad thing. We think it's an amazing thing. If we hit a certain target, I've been in too many companies and seen this too many times. Any metric that you choose, people can game it. They can do things and suddenly that will hurt the business in order to achieve a particular goal because their financial welfare requires that one goal to be met.
And you create internal politics. Lemonade is an organization of remarkably little tension and politics. We all got the same coin that we're measured by and working towards. And you know, if we hit an IFP target this quarter, is it because our marketing team did a great job, or because the product team did a great job, or because the customer support team did a great job, or because the cross-sales team did a great job? The list goes on and on. And it takes a village to hit a goal. And therefore we want everybody pulling in the same direction. Which is why conceptually we don't use performance bonuses. We use the stock as the rallying motivator for all of us all of the time.
Mhm. That makes sense. That's super helpful. Thank you. No, thank you for that. Yeah, another criticism related to is sort of moving target or sugar coating maybe something in a letter. So I don't know if I share this one as much, but it definitely comes up with other investors, is that there's maybe a sort of sugar coating a metric in the letters when it looks good, or quietly dropping it when it doesn't look so good. So the most recent example was you showed for a number of quarters the IFP growth versus opex excluding gross spend, and then that was removed, and then that turned into employee count versus IFP which looked a little better, and then that was removed to neither was shown. So do you want to address that kind of criticism that we're going to show or hide things? Another example I've heard in the past is that renters and home are combined. 'Oh, they don't want to show the home because the numbers are too poor' or something like that. And I can understand too, like I think every company, every person, it's natural you're going to highlight some things and not highlight other things. And so I don't think that's even dishonest. And there's often usually explanations or sort of details or thought behind the thought, but would you just give you a chance to talk to those criticisms.
There is definitely always, you know, we're very proud of our business and we want to present it in the best possible light. And so I think the temptation that you're describing is there and there may be occasions where we're guilty of some of that, not with malice or anything like that, but just things that we're excited about and we think are great and we want to share them. There's I'm sure some of that psychology at play, but we cut ourselves some slack for a couple of reasons. One is we disclose everything. You want to know what our IFP per employee is? Take out a calculator. It's all in there. None of this is being hidden or withheld. As it happens, that particular chart exists. It just exists on our investor website, not in the letter. We actually did produce the chart this quarter. But even if we didn't, I don't think that's an obfuscation or hiding or dishonesty. We choose what we highlight in the letter. Every quarter, we choose what we think will be interesting. And if we were to remove key metrics that you need or change our disclosures, that would be more significant.
But generally speaking, we disclose everything and whatever it is that you want to get at. Look at this quarter specifically. We didn't show that chart that you're asking me about, but you and everybody else had no difficulty reproducing it. All the data was there. It takes a second to get there. So, you know, these are visual aids to try and make a point that we're trying to get across in the letter. I don't think they should be taken as disclosures per se. Everything is disclosed. We're hiding nothing. But there's a broader point. We're actually trying to convey a hypothesis and to align with investors about what we think Lemonade is all about. And disclosures are disclosures, but what we choose to highlight in the letter and the charts that we put in the letter are us, you know, these are teachable moments where we try to connect with our investors and share with them how we see Lemonade. And you can accuse us of being somewhat distorted in our vision of Lemonade. And everyone can draw their own conclusion, but that's what we're trying to do. And it isn't always about cherry-picking the best-looking metric.
Here's an interesting example. Our trailing 12-month loss ratio was 59% in Q2. In past quarters, that would have been the front page in those bullets where we put all the data and it would be followed by a chart showing our loss ratio over time. It is our best loss ratio ever. In our letter, you have to get to page 34 out of 35 to see that mentioned. In fact, we've kind of stopped talking about loss ratio at all. And that is because we think that the general-purpose investor, the tech investor and general-purpose investor, doesn't know loss ratio. Over time it's going to go up and that could become alarming. It didn't go up. It went down. It was the best ever. And yet we pushed it right to the back in the disclosures that we promised to keep going because we think that gross profit is actually a much more important and helpful measure and it's how we manage the business. We don't manage it to a loss ratio. We try to maximize gross profit and sometimes that means raising loss ratio. Sometimes means lowering loss ratio. And look at our cover page. It has gross profit.
Now you say, 'Oh, but two years ago you would have shown loss ratio.' True. We switched out loss ratio for gross profit. Not because we're trying to hide some dirty secret, despite loss ratio being the best ever. You'll have to search to the back to discover that fact. We didn't talk about it in our script. We didn't put it in the letter. So it was thoughtful. It was reasoned. It was cherry-picking of a sort, but it wasn't driven by an attempt to obfuscate or to hide some embarrassing metric. On the contrary, it was despite it being the best ever.
