September 16, 2022

Friday Forum is an All Hands meeting for the Levels team, where they discuss their progress and traction each week.


Josh Clemente (00:00):

Let’s do it. Welcome everybody. September 16th, 2022. Get right in. All right, starting off with recent achievements. This week Casey had a huge podcast launch with Bari Weiss on the Honestly Show. Can’t reinforce how much I loved this conversation. I really, really recommend everyone listen to it. Casey really brought some new energy I feel like, and just took her surgeon scalpel to a number of incentive structures that exist and that are not well-known to people, and it was just a really amazing conversation. There’s going to be an op-ed to compliment that release, which will come out this week. Super excited to read that as well. And then we also had Taylor on Zach Bitter’s podcast. If you don’t know Zach, he’s an extreme endurance athlete, does a lot of crazy stuff like running a hundred miles on a track without stopping and setting crazy records in there, and also doing so on a ketogenic or carnivore diet. Great conversation there.


And then we did Platinum Longevity round two with the Abundance Platinum Team, Peter Diamandis’ team out there in SF, which was fun. And Casey’s recording with Ben Greenfield next on Ben Greenfield’s show, so that’ll be great. That’ll be our second Levels performance on Ben’s show. We had the Apple Cider Vinegar Community experiment go live and rapidly got 200 signups in four days. I think for the original community experiment that we did with the Koch challenge, we had less than a hundred and it took some time to figure out or to get all those signups. So this is really great. People are clearly excited about experiments and understanding better how they perform relative to other people. We also have a fat burning experiment live with the internal team. I think about 12 people there. So yeah, excited to see these spinning up.


Caitlin and XinLiu are back. Both of them have been on leave for some time and so we’re very excited to welcome the two of them back to the team. On the engineering side end focus has been on UK beta, so continuing to debug and prep e-comm iterations so that we can really start to scale those. Collecting spike score feedback, so the spike detection has been out to our members for some time. I think the last thing that’s happening is backfilling member history for spike detection, so it’ll run all the way back to the beginning of a member’s data. We’ve got chart improvements happening, height and weight are going to be in app next, and event-based insights as well are a big focus for Eng Team. Support continues to really face an amazing set of obstacles that seem to change every week. It’s been certainly difficult for the team to stay on top of the effort changing circumstances post-launch and so just really want to support them and appreciate anyone who’s been digging through threads this week.


There’s just been so much rallying going on in the various channels to either directly relieve load that the support team has to face or think a few steps ahead on how we can avoid seeing these things recur. So big shout out to the support team and everyone else supporting them. Logging V5 is live to members. I think that was approved yesterday. We’ve got lots of work happening on Labs 1.1 and version two coming next, and then UK beta in app demographics, healthier food choices, all teed up from the product side to start either shaping or implementing. This week spent I think just a very small amount of time, and I believe Paul took the lead on this, but optimizing the homepage for a performance. So went from a Google Lighthouse score of 38 to 57 with some very basic changes like compressing videos, hero videos, and really optimizing for mobile viewing.


So the last thing we’re going to do here is move the partner pages off WordPress to a platform that will allow faster iteration. So definitely paying attention to the website and generally across surfaces, making things mobile first as opposed to considering mobile as a second strategy. So another experiment here is the mobile optimized ultimate guide PDF experiments. You can see a picture over on the right side here of what that looks like, but essentially taking some of our, what I would call our pinnacle content that really is deeply researched and people refer to it, it’s actually some of the main sources of traffic to our website, and repackaging those in a form that can be used as a reward for those that want to sign up for the email list and want to digest that content in a different channel.


So we’re going to be experimenting there. We’ve also got the exclusive content experiment going live right now with Casey and Dr. Mark Hyman, which is essentially an option to get additional information right to your inbox that you wouldn’t otherwise get if you don’t sign up. And then we’re doing micro content YouTube videos. You can see a number of these with our advisor, Dr. Ben Bikman, basically taking larger interviews and breaking down very small bite-sized clips that are relevant to a specific topic. And this drives additional discoverability for people who are, say browsing for gallbladder health on YouTube. They could find a little clip there that will lead them down the rabbit hole of Levels content. We’ve got an updated financial model coming including August results, so that’s going to come out very soon. So stay tuned there. And lastly, the Levels Test Kitchen. This is a big project, it’s in final editing.


I saw an early near finished clip of it and it’s just amazing. I’m so excited to see this. I think I mentioned on threads, emotional to see the quality of our content just continuing to get better and better and more in depth. There’s going to be five videos here, a docuseries. We’re going to be leveraging sub cuts of this across all of our channels in the app on social. It’s going to be amazing. Had a ton of great user generated content this week. We had some activations live on Huberman, Rich Roll and Max Lugavere’s podcast as well as Doctor’s Pharmacy, A couple of great whole new Level episodes with Dorothy Kilroy, Azure as well, diving into Hormone Health. We had Rachel Sanders from Routine and let’s see, I had a great conversation with Dr. Cary Gunn who built Genalyte, an amazing supporter of what we’re doing. We got some new renders coming live to show Levels product out in the wild.


VIP is from Zumba CEO to Steve Aoki in the product and some other great content. Also, we’re shifting metabolic pearls from Casey to the research team. So we’ll continue to use that concept and reinforce the research learnings that we’re getting. All right, I think that’s it on this one.


This slide is not coming through. I want to welcome Rian Edman. Rian, did you jump on the call?

Rian Edman (06:48):

Hey, good morning.

Josh Clemente (06:49):

Hey Rian, how are you? Sorry, I think your hero image is not loading. I think that might be on my side, but I want to welcome Rian, member, engineer, outdoor enthusiast. He has a couple anecdotes here about a history of diets and fast food that not necessarily proud to admit, but he’s learning and on a long journey with the intention of being able to climb mountains well into late life. And I really personally can relate to that specific anecdote. And Brian, I appreciate you taking some time to hang out with us. We’d love to hear more about your experience with Levels and metabolic health generally and what you’re excited about.

Rian Edman (07:26):

Yeah, absolutely. Thanks Josh. So I got on the beta access program, I think it was 2019 was when I used Levels, and I invested in 2021 because I really believe in what you guys are doing. Yeah, I guess from an early age I’ve been interested in metabolic health. My dad was diabetic growing up and he would use the old pinprick glucose monitors to give me instantaneous readings when I was not feeling good as a kid just to make sure that my blood glucose was normal. So I’ve always been aware of it and given the hereditary nature, I’ve always been conscious of what I eat. Even though, yeah, growing up in Ireland, diet isn’t so great as I’ve learned since using Levels. So I tried a lot of the foods during my month long subscription that I used to eat just to see how my body would react and sure enough, wasn’t pretty. I used to eat oatmeal for breakfast every day. Yeah, don’t do that anymore. That was a pretty bad food score.


Yeah, so I guess I had a similar experience to you as well. I think I read your story about mid-twenties, you were very active and you started to get sluggish and that’s how you ended up going down this road and discovering the importance of metabolic health and found myself in the same situation and using Levels even for a month, allowed me to find some of the foods like oatmeal and I used to love acai bowls, thought they were super healthy until I saw what they did to my glucose levels. And yeah, I don’t make that mistake anymore. So it was really helpful for me to learn from that experience.

Josh Clemente (09:28):

Amazing. I would love to hear along the way any long-lasting behavior change that has come from this product and or just generally other tools that you’re using to drive behavior change in the long term, and then also what you feel is missing and what we can improve on that you would really benefit from, whether in the Levels existing CGM product or if there’s additional metabolic health tools that you feel really need to be built, need to exist.

Rian Edman (09:57):

So the things I have changed, yeah, food choices. So some of the ones that I know now don’t react well with my body, don’t do oatmeal anymore, don’t do acai bowls. More cautious with things like bananas, fast digesting starches that cause my glucose to spike. So I still have them in moderation but in limited quantities. I also, I try to eat some protein or some fat before the meal because I saw that that was able to stunt the glucose spike. And then I also try to walk after meals now, which has really helped how I feel. There’s less of a slump 30 minutes to an hour after eating.

Josh Clemente (10:44):


Rian Edman (10:48):

Things that I didn’t like. I love the app, I love the experience. I think what you guys are doing are great. The only thing I didn’t like was the cost and I know you’re working to bring that down and I know that’s a matter of scale and timing, but yeah, I would use it much more frequently if it was just at a lower price point.

Josh Clemente (11:14):

Amazing. Yeah, that’s really important for us to always remind ourselves that the key here is to make this accessible to anyone who can benefit from it, which is clearly more than the current price point enables. What’s the one thing that you’re looking to learn more about in your personal health journey over the coming years that additional feedback loops could support? Is there anything that you don’t feel you have enough information about that you would love to be able to tap into?

Rian Edman (11:44):

Yeah, I think one of my key interests is longevity and being able to delay the onset of aging and improve my health span so that I can not be decrepit when I’m 70 years old. I live an active lifestyle and I want to keep doing that for years to come. So any new data I can get that can direct my behaviors to preserve that, that’s what I’m looking for.

Josh Clemente (12:13):

Feel similarly. Well Rian, thank you for joining us. Thanks for setting aside time on Friday morning. We can’t overstate how much this sort of conversation, hearing directly from use cases, supports what we’re doing and reinforces the features we’re building and the long-term vision. Lastly, if you’d like to stick around for the rest of the meeting, we’ve got a bit more here on the plate and would love to have you join and just observe and chime in. And finally, thank you for supporting us both as a member but also as an investor in the crowdfund round that we were just talking about replacing our Twitter river on the website with the testimonials that we received from that crowdfund round because it was really a life-changing moment for many of us as we saw all that support flow in. So thank you for doing that. It really means a lot.

Rian Edman (13:01):

Thanks very much and thanks for having me on here. And yeah, I’ll be glad to invest again if you guys open another crowdfunding round. I think it was capped at a pretty low amount, but I really believe in what you’re doing, so I’ll definitely get behind that if you have another round.

Josh Clemente (13:20):

Well, thank you. I certainly would like to make that happen. All right, there’s the picture. It was have an animation. There’s Rian. All right, jumping forward, culture and kudos aside, some great meetups action this week. Well, firstly, happy two years, Miz. Wow, that feels like five years. Amazing. That was a densely packed two years. Can’t of course overstate how critical Miz has been to Levels and shaping the culture in the company and the logistics and everything we’ve done here. So excited for many more. There’s some great meetups this week. JM and Stoddy hanging out in New York, Ryley and Paul up in Banff, and then Ryley somehow teleporting to San Francisco with some of the ops team, products team. We got to hang out Platinum Longevity and some of the ops team were hanging out in research group jumped in as well. So we had a nice dinner, some really good conversation, always loved these.


And then I want to highlight quickly, I touched on this on the first slide, support team has really been facing headwinds, not just from the volume and from the changes that launch brings, but also some systems and partner reliance that we have that we simply can’t overcome easily. So we don’t have control over every step. We have to rely on partners in certain cases. And when those don’t pan out or aren’t hitting SLAs, it really adds an additional burden on top of what is already a difficult time for support. So I just want to reinforce that this difficulty is happening, the support team needs our support, and we’re seeing that in a really amazing way. So on the lab side, we had some setbacks this week and there’s just been a flurry of activity across product ops engineering to find alternatives to get us to be able to support our members, to reduce the load on support and also to improve the experience for everyone, both on the team and in our member base.


So just want to say that it’s been inspiring to see that cross-functional, just flexible, rapid, just exceptional work that’s happening and everyone’s operating in good faith. Fingers are not being pointed. And I think that’s the most beautiful thing here. It’s just culture in action. Love it. So thanks everybody that continues to improve our systems and improve our cross-functional work. Okay, I’m going to step back and let Zac take the culture slide today.

Zac Henderson (15:40):

Hey everyone. So I’m going to talk about make yourself obsolete. I will say this is absolutely one of my favorite core cultural ideas at Levels. So this is in the Culture Handbook, just a quick snippet. Everyone’s goal is to make themselves obsolete, build a tower, hand it off, make it better. At Levels, we definitely reward and encourage this behavior, we don’t punish it. So when I think about a concept like make yourself obsolete, I like to first think about what the opposite is. That helps me understand it. So what’s the opposite of trying to make yourself obsolete? Probably trying to make yourself indispensable. And what I love about this is that sounds great on paper, because we all totally want to be needed. The problem is that in practice, it completely fails and it will just make you unhappy. Number one, if your focus is making yourself indispensable, it will inevitably make you a bit territorial.


Not only is this very anti short toes with the people you already work with, but this is the root of that feeling that some people have where a new team member comes on and their skills overlap with theirs. And instead of saying as you would, if you have a mindset of make yourself obsolete, “Oh my gosh, I have another person who understands my role, let’s work together. Maybe we can automate or hand off some of this stuff that I haven’t had the time to get to,” instead, your reaction is, “Oh my gosh, new person, my skills. Are we in competition? What’s going on here?” And it just wrecks your work mindset. It leads to gatekeeping, knowledge hoarding, a lack of openness, all stuff that just doesn’t make us happy. And I’ll tell you, this is very relevant to me as a lawyer because I see this all the time.


This mindset incentivizes, convincing others that our work is way more complicated than it is. So instead of having the goal of trying to make things more transparent and helping people actually see the path through the difficult thing, we end up sometimes unintentionally creating this space where people have to come to us just to understand stuff and it’s messy and it just puts up walls. So that mindset inevitably pits you against the company and your colleagues and it just ends up stinking. So take the opposite though. Making yourself obsolete or having the goal of making yourself obsolete. Starting just from any smart company’s perspective, this is awesome. It will free up more of your time so you can work on new and different challenges. It creates skills redundancy. This helps the company and all of your colleagues in a huge way. Now there’s more than one person who knows how to do a thing.


Or maybe you made it so efficient that it’s now fully automated and it’s just not something anybody has to worry about anymore. Think about the positive effects this has just for taking time off. If you know that there’s someone who actually knows those core skills, you just can take that time off feeling even better about it. And then when you level up your skills, you end up helping the company too. This is the double win of making yourself obsolete from the company perspective. Not only do we get you as someone who demonstrated the leadership and self-management to clear your plate up, but now you have time to learn new stuff and we get to take advantage of that as well, which is awesome. And then as far as what’s in it for you, making yourself obsolete, in my view, it’s the ultimate professional flywheel.


It requires practicing good leadership and management, even if you’re just a department of one. It inevitably leads to constant learning and professional development. It always results in more opportunities and more doors being open. Never less, this will never close a door for you. And it will just make you happier, more open and just make you and your colleagues have just a better working environment. And again, I think the core concept of Levels is we see all this and this is absolutely one of those things that we will always elevate and reward. So how do you make yourself obsolete? Look for ways to share the workload, harness Athena, delegate tasks, find a way to make processes, create efficiencies in your own workflow. All of these things are ways to make yourself obsolete even in a little way. So core Levels, I love this. Make yourself obsolete if you can.

Josh Clemente (19:49):

Beautifully articulated. Yeah, definitely can’t add much to that, so thank you Zac. And of course, if you’re looking to learn how to do this, real world examples, there are plenty out there, but just reach out to someone like Zac, someone like Miz, and ask for examples of in their workflow, how are they making themselves obsolete? It’s everywhere. It’s amazing. Thanks Zac. All right, Level shows you how food affects your health. No change to the main thing. Everyone should be working towards this. Our company objectives here, member retention, new member acquisition, member health improvement, stack up to that main thing. And you can see functional OKRs coming together that support these three top level objectives. Okay, we’ve got demand gen from Ben.

Ben Grynol (20:33):

Okay, here we go. So this is about ramping up video quality. And this framework is in tons of memos. We typically talk about it with product, but it applies to everything we do. So the skateboard, scooter, bike, motorcycle and car. Next slide please. So we have been producing a ton of video. A lot of it looks like the one on the left, that’s David Letterman and Martin Short doing some talk show thing in the 80s and that’s about the quality that we produce at. But what we’re used to watching as a team, as a society, is things that look like the frame on the right. And that’s one from Chef’s Table previous season, and those are shot on these cameras that are called red cameras. They’re a hundred thousand bucks a piece and they’re really high quality and we get used to that type of content.


We get used to that narrative and used to that storytelling. And there’s actually merit in both, but for different reasons. So next slide please. So we made a conscious decision when we started investing more time in video and we went for quantity and velocity in exchange for quality. And the caveat to all of this is that they come with trade-offs. When you don’t have as high quality content, not from an information standpoint, just from an aesthetic standpoint, you lose this feel of a brand, but you do it in hopes and hope is not a good strategy, you do it with the heuristic that doing so will allow you to A, get a ton of learnings and B, achieve certain metrics to be able to get to the next level quality. So you can keep going along in that skateboard through to car cycle. So next slide please.


So why do this? Well, quality equals time and money. We care mostly about time right now and it doesn’t mean we don’t care about money, but how we invest our time as far as opportunity cost goes makes a big difference. And so can you outsource to get greater quality? And the answer is yes, but many agencies will charge 20 to 150,000 bucks for short docus style content that looks like Chef’s Table and it can be five to 15 minutes and it can take 12 to 16 weeks to produce this type of content, which is insane, and it’s not hands off. It actually takes us 12 to 16 weeks worth of heavy, heavy involvement. And so you go, “Well, let’s punt on that right now.” Because we’ve got the capability internally with skillset. So Tony is an incredible cinematographer, incredible editor. Whether or not we have seen it because of the David Letterman type of content, it is there.


It’s just we have an exercised that muscle right now. So next slide please. What do the metrics look like? So this is on the left, it’s January 21 to August. We didn’t do anything on YouTube specifically. And these metrics, sorry, these have to do with YouTube as a platform. We didn’t do anything with video on purpose. We just said opportunity cost is way too high. We’ll address it sometime in the future. We’ve got to get some infrastructure in place. Tony came on board October of last year. And so then we said, “Cool, in January once we’re ramped up, let’s start actually doubling down.” So you can see we had 37 videos, we had about 3000 views, 166 hours of watch time, very few subscribers. And then we hit the gas in January and we said, “Cool, let’s start shipping a whole bunch of stuff.”


So we’ve shipped roughly 700 hundred videos, 34 million impressions, 325,000 hours of watch time, which is an insane metric in itself. And a lot of this is driven by having high quality information content, that being the videos that Casey’s been doing with Rob and Dr. Perlmutter, these all lead to these metrics. It’s not the, I hate to break it to us, we’re not that exciting, Friday forum videos, those are not leading to 325,000 hours of watch time. But you can see that we’ve got a lot of growth by going quantity and velocity. So next slide please.


So then when do we actually get the quality reps in? Well, it’s incrementally, it is, hat tipped to Seth Godin, drip by drip is his famous line. So next slide we started here. So we started controlling what we could control with fast execution. And I don’t know if everyone can see the difference, but on the left, that’s a video on inflammation that we recorded with Taylor and it’s just the built-in hardware camera, poor quality audio, not a ton of attention to the background. And on the right is one with a DSLR camera. So it’s a couple thousand bucks for the audio and video setup. And it’s not just our psychological viewer or the psychology of like, “Well, it looks better to us as a viewer,” it actually makes a huge difference for the algorithm. And we can do this remotely, we can do it with a very fast execution, not a ton of time invested, but these little incremental things get us to a better point.


So next slide please. Then we bring to the decision tree. So it was a Sunday night, I believe, correct me if I’m wrong, Casey, and this was I think the end of May. And Casey unprompted was like, “Hey, I’m just going to turn on my camera, my iPhone and let the audio record and just film myself cooking. We’ll see what happens.” And she ended up filming this video, it was 30 minutes and she had, I can’t remember if it was texted or something and she was like, “What do you think?” I was like, “This is good content,” she sent it to a bunch of us, “I bet it will land with people, even if the quality looks crappy, I bet it’s going to land with people if we ship it. Let’s see what happens.” And the feedback was people said, “This is super authentic, it’s genuine. I just want to see Casey Cook. I want to learn more about food philosophy.”


And so the decision tree becomes, next slide please, what do we do? Do we outsource to try to get that next iteration or do we bring this in-house so we can start to exercise the quality reps? And that let us down the path of saying, “No, let’s actually do this now.” Let’s go Chef’s Table, let’s get the red cameras, let’s film this. Let’s prove Tony actually is a cinematographer and can edit this stuff, but let’s control the treatment and the concept because outsourcing it’s not going to get us the result we want anyway. And we knew this, so we said, “Let’s build the muscle, get the reps in, experiment with treatment/concept and go back to outsourcing once we have that, because then we have levers to tweak.”


So these have not been released yet, but they’re frames from, we’re calling it Levels Kitchen. It was a shoot that Tony, Sonya, Stacy, and Casey were involved in Portland and it’s a bit of a mini doc and then there are four recipes associated with it and it’s all long form content. So data is to be determined, but Sonya’s going to report on all that in about a month’s time. So next slide please. Let’s see how we’re doing on time here. I will try to keep this quick, couple more to go. So why does this even matter? Who cares about quality? Well, one of our OKRs is audience growth across platforms and how does this actually happen? Well, the algorithm optimizes, we’re just going to talk about YouTube right now on two things, performance and personalization.


So performance is when a video gets served up, does it have a high click-through rate to the number of impressions it gets?` When people click through, what’s the watch time and what’s the retention? If you trick people into serving up a piece of content and they click on it and the retention sucks, well then that gets you screwed with the personalization aspect of it because YouTube thought, the algorithm thought, “Hey, this is for this person based on what they’re interested in, but clearly it’s not,” so you get downgraded super quick. So next slide please. And so the algo scoring can be hit-and-miss. This is where we use some data and we take things into account. And there isn’t one algorithm for YouTube, there’s a whole bunch behind the scenes and the scoring is based on frame and object recognition in the entire video.


It scrubs all frames and it goes is what is represented here going to match what I think is said in the information credibility? That being the transcription, in the description, in the title, in the thumbnail are all these representative things that blends this together. So it goes, “Is audio quality good? Is the information good? Are the cuts quick?” All these factors. And then last is are the topics and trends relevant to what’s happening societally, what people care about in search? So these images, these are two separate thumbnails and this is where you have to be data informed and not driven. It’s cloud vision. So you can upload and you can say what’s going to happen with the thumbnail? And you can see that’s facial recognition on Dr. P. And you can see it’s saying, “Cool, that’s got an element of surprise to it. We’re 88% confident this is going to be good based on what is said.”


And the bottom one is 99% confident and you think that’s a better thumbnail, but it’s rating it really high or very likely to be spoof content. So it might serve that up to a comedy audience because it thinks the thumbnail, that’s what it’s going to recommend. And somebody clicks and they’re like, “Oh man, that is not what I thought it was,” and you get downgraded. So you have to use all these levers and say what is going to happen? And keep iterating, iterating. So next slide please. One more after this. So what does this actually mean? Well, attribution isn’t a perfect measure of performance, but this data keeps us informed. So why do we care about all this quality and video content? Well, YouTube as a platform is our highest leverage traffic source. When we think about eyeballs, emails, orders, for traffic, coming to the website for organic content.


We get over 2X the engagement of traffic and the lowest bounce rate. So people are staying on pages on the website and it’s not perfect because people could come in through other avenues, but we get over two minutes of page view time when people are viewing that content. So it’s important to understand that we keep pushing on this as an angle. And then the last slide, here we go. So the takeaways are demand gen and demand capture, they’re inextricably linked. So this idea of audience growth across platforms, high content cadence, increased quality, and all of these things will help us with this framework of eyeballs to drive traffic to the site, to capture emails and nurture them, to get orders. And that is your update.

Josh Clemente (30:30):

Wow, that was killer. Things I never thought the computers would know they know already and have known apparently. So love it. Thank you Ben. I’m going to come back and watch this one again for sure in the [inaudible 00:30:43]. Okay, over to product, David, Maz.

Maziar Brumand (30:47):

All right. Well that was exciting. I’m really glad that Ben and James are driving the boat and the team. That content is just so deeply thought about and we’ll be the life blood of the app too. So super exciting update. All right, so welcome to the product updates for September 16th. You’re probably wondering is that cake? And has Maz lost his mind? Why is he showing cake on Friday Forum for Levels? Don’t worry that cake is almond flour, allulose and olive oil with deep coloring from raspberries. So don’t worry, it is metabolically healthy. Actually that’s not true. So I’m just kidding. But the point of this is continuing on the theme and metaphor of our product really being in layers just like a cake. And here what we try to do is show that we need to engage people. That’s the first layer of the cake, that’s getting people in, building that trust.


We need to keep people, retain people over a long-term to really create product that people want to use and continue to use for a long time as really a partner. And then really going into the delight of what else can Levels do for me? Like how Apple’s trajectory was. Apple first made the iPod, they made the iPhone, they made the iPad, and then people were like, “When is Apple going to make a fridge? When is Apple going to make a router? When is Apple going to solve all these other problems?” And so it was that trust that the company built and people wanted them to do more. And I think that is the trajectory for product that we want to create. And if you break it down and say, “Okay, how are we thinking about that? How are we driving that engagement? What are the features that are driving that engagement and what is the features that are driving that retention? And then eventually delight, what does that look like?”


And so looking at our product today and where we’re trying to go, the first layer that we built is first and foremost, content. Content is the lifeblood of the app. So starting with event-based insight that Mike had been heading and now we’re up to 28 event-based insights as of end of this week and keeping going. We have a roadmap to create 10, 20 pieces of content in that format a week, a month, and we’re going to keep on investing there and we’re really bringing the rest of the company to help us create this roadmap. Whether it’s working with research, whether it’s working with clinical, whether it’s working with KCN team, it’s really taking the best out of the company in creating this content. The second one is healthier food choices. So this is proactive content, creating something that people can follow.


Third one is the personality based insights that really brings a human element to it. So we’re delivering content to engage people. Second is logging. We want the interaction to be interactive, so that’s logging, to engage people in what they are doing. And then the third one is obviously the scoring work that’s creating feedback to people, this biofeedback to keep on the right track. So content logging, scoring, creating that feedback to engage people that 30, 90 days to really teach them and give them insights about how food is affecting their health. The second layer is labs. We’re creating this lab product to really check in on your health throughout time. How am I trending? Am I reaching my goals? So it is effectively a check on you. The second one is personalization.


We want to create a product that people are investing in, it’s getting to know them. It’s a tool that people say, “Why would I move to another product and have to start from scratch? This product really knows me.” The third one is community. We want to invest in community. Obviously we created this beautiful community outside the app where it feels like people in there helping each other. We want to bring that into the app and we’ll start light so it feels like it’s progressive but build out from there. And the third one is helping people make the right choices that improves their metabolic health on the food choices to purchase things that will help their metabolic health and really demystify that and make them actionable. So those are the things that we’re thinking about incoming and then eventually we move on to other features which we will share in a future update.


All right, next slide please. All right, so what is the equation? This is so simple, why would people buy products? It’s really in their head, they’re calculating this very simple equation, is the value I’m getting out of this thing is more than the effort and cost that I’m paying into it? Pretty simple. This becomes extremely important especially for retentive products because you can have a surge of motivation and say, “I want to buy that blender and I’m going to go buy it.” And you buy it and you’re like, “Ah.” The value goes down over time. And the effort of cleaning the thing increases over time. So we want to create a product if you want it to be retentive, maintains its value, has a low effort, and obviously doesn’t have an escalating cost. It is very hard for people to maintain high cost and high effort over time.


So we definitely want to bring people in, but the retentive layer of our product has to be something that’s low efforts, maintains this value and doesn’t have an escalating cost. So we’re thinking about that even at each of these layers of how does value, cost and effort play a role in a product that we’re trying to make? Next slide please. The value effort cost is also correlated with the behavior model. This is a chart from B.J. Fogg’s book Tiny Habits, which is we make decisions to do things when things are above the action line. Meaning our ability is higher to do the thing, we have the ability to do the thing and our motivation is high. And those are related to this concept of value and cost and effort. So we want to create a product that people can take action on and are motivated to do. And that’s associated with the formula we just talked about, which is value and cost. So it also ties to the behavior model that we were implementing into the product. Next slide please.


All right, so how far have we come in the last few months? So these are all the features and it’s a big thank you to everybody in product design, engineering that have really come together to launch these features. Most of this is since July, so about nine weeks or 10 weeks of engineering time effectively to launch this. The [inaudible 00:37:33] Now 2.0 that creates the interaction monitor and smart procurement that makes sure that we send people what they need and no more. The glucose game, which teaches people interactively about how food affects their health before they even get started with the CGM. UK 1.0, for expansion and allowing more people in the world to actually benefit from this product. Event-based insights it is reactive to you, feels personalized, it’s insightful, it’s fun to watch.


Scoring 2.0, really creating that feedback loop in a way that’s understandable, in a way that is rewarding and in a way that is positive. It’s positive psychology, encouraging. Logging, making it seamless, making it frictionless. Bringing that friction to the lowest possible so people want to do it and can sustain. And that low effort we just talked about can be maintained for a longer period. And of course demographics, which is our latest, which is enabling us to start thinking about personalization, to build that investment layer. So big thank you to everybody that has made this a reality. Next slide please. And this is a visual presentation. When we talked about this in July, we’re at that first box and now we’ve made progress through that and we are where we are and we are marching forward. We’re not even started yet. Just the beginning. Next slide please.


All right, so what are we’re working now? So shaping is Levels’ levels which Allen and team will give us an update next week. So it’s coming, I’m excited about it. I think you’re going to like it. Labs 1.1 and 2.0 that Cosmos really been driving and pushing the envelope thanks again to all of our advisors and we’ll continue to push there and in app content, which is everything from event-based insights, personality, event insights. And David is going to talk about today some of the feedback we’ve gotten from the [inaudible 00:39:19] videos and the personality and obviously continuing on personalization with the help from engineers to build demographics, to pull in data from healthcare and scoring, which Alan’s going to talk about some of the thinking of the next iteration of scoring. Thanks for that. I’ll turn it over to David.

David Finner (39:36):

Can we refresh Josh before you move on to the next page?

Josh Clemente (39:48):

Okay, there we go.

David Finner (39:53):

All right. There’s a few project updates before we get into the personality driven content. So first off, on the event based insights front, I wanted to highlight that we now have a retrospective available. Please go ahead and read that. You can hear all about what we did in the first round, what we learned, what went well, what our opportunities for improvement are. Mike put this together. It’s really good. Thank you, Mike. He’s at a conference today, so I just wanted to plug that. And then as Maz was just mentioning, one of the things that we’re really trying to do is increase that ability for our members and their motivation. But one of the features that we just launched was hyperlinks and stories and Steph did a great job on this.


And why this is important is that in these stories that we’re pairing, like try these three or four things instead of oatmeal, there was no way previously for our members to open up each of those different recommendations and now they can open up each one. Maybe they were only interested one out of four. So this is a great way to increase that ability, a small win, but I think it’s going to be a big effect. And then we’ve got a bunch more insights queued up. They all have zero eng effort, which is a big celebration. All of the insights coming up are going to have zero eng effort. So big thanks again to GLI and Mike for working on that. Next slide. On the logging front already mentioned, but logging 5.0 is complete. The whole project is complete wrapped up. I’m doing a retro for that 5.1 launched this week and that was jam packed full of updates based on feedback from the first one. And the biggest one was faster, one click access to the logging screen.


There’s no more fly up. Now you can go directly to the logging interface and 5.2 is underway. So 5.2 is a logging shortcut for photo-based logs that I previewed last week and there’s going to be more to come as well. Next slide. And closing the loop on scoring 2.0. So after we launched 2.0, we were gathering member feedback and some of the things we learned were people were confused about onboarding into it. So earlier we launched an onboarding interstitial. We recently just launched as well and inclusion of that into the glucose game. So now there’s one unified onboarding experience where you learn about the glucose curve, food logging, the stability ring, very nice work by Marillo on that one. And then on the design and product side, we’re taking to heart some of the feedback that we heard from our members who thought that they wanted something a bit more of an abstraction layer over their data, something on top of the stability score that help them see how are they doing at the end of the day and day over day.


When you look at the spike counts, it’s hard to visually grok two or four when you’re scanning in your calendar view or in my data. And so we’re thinking about how can we bring back some of the visual glancability of the metabolic score that you had day over day into another score abstraction that uses the same system, the stability ring and spike duration, things like that is woven into it, but is just a very quick visual glance when you’re in my data. More to come on that in a little bit. Next slide. I think you can skip that one.


And all right, next one. So today the highlighted experiment is personality driven content and to set the stage, as you know, we’re shifting into a program where we’re trying to actually change people’s behavior and not just show them their metabolic awareness and what we’re trying to do, you can read more about this in our in-app content framework strategy memo, but pair our effective interventions with an appropriate delivery style because there’s a variety of people and they resonate with different delivery styles, whether it’s a long form article or a video. We know this from the blog, we know this from Instagram that we’ve been exploring different formats on Instagram.


And I think even as Ben mentioned on YouTube, there’s some formats on YouTube that are really killing it and some that aren’t doing quite as well and that’s expected and the same things on Instagram. But what we’re finding is that the things that have people in them, especially on Instagram, are performing way, way better than the things that are just more rote data. And so the experiment with personality driven content was to take one video, put it in app, one video from two of our most highest performing personalities that we’ve engaged with for our Instagram platform and see how that would work. Next slide.


So in general for this, we went into it thinking that the human element matters in app just like it does on Instagram. I think the stats, and Sid, you can correct me later if I’m wrong, but I think the stats are three to four times higher engagement for videos with Mallory or Janet here who you see on the screen compared to something that it’s the same kind of content that was delivered but without a human face. So our hypothesis was that this personality driven content would increase the efficacy and retention which we’re shooting for as a company by better delivering our interventions. So we have all these interventions and this new behavior change system that we’ve queued up with the event-based insights, increasing ability, motivation, et cetera, but translating that into the medium that really resonates with each and every person. So what we did was we had Janet on the left, Mallory’s here in the middle, and we injected those into the app.


Next slide. We had a survey that went along with it and shout out to Sissy for helping us reach and reengage the people that we sent these to and get feedback from it. We were able to get 32 responses and the feedback was very strong. In summary, we had a couple proxy questions for whether it would be efficacious and retentive and 72% of people said that they learned something from the video, that they would take away from it. 76% of people said that they would come back to the app if they had this sort of content in it every day. And so this is very, very early. We can’t read too much into this, but as a very first baby step, I think directionally this was good information to get back and encourages us to continue on in the experiment. As a personal anecdote, and I’ll show you these graphs in a second, but this is the highest response to a product survey that I’ve sent yet across the product surveys that I’ve sent for different features from the company.


So next slide. We asked people if they would enjoy more personality driven content like this Mallory video on the app, and it was very far skewed towards strongly agree. This will be a theme from all these slides, but next slide. How interesting would you find content featuring other people, other members, personal experiences with Levels, also high skewed towards very interesting. Next slide. Now this question was a bit different. It was saying not just Levels members who are personalities but experts. So how interesting would you find content delivered by experts? Again, a huge skew towards the five out of five. Next slide. And then going back to those two proxy questions for efficacy and retention. Did you learn something from the video that would help you manage your glucose? A vast majority of people said yes, and three-fourths of people said that they would come back daily to see this in the app.


Now proof is you can’t always take people at their word in any survey. So we want to test this out and actually see as the next experiment and not just have one video, but have a series of five videos and see do people actually come back and is this really driving retention? Next slide. So some interesting takeaways for this. One of the things I wasn’t expecting, but the highest factor that people really took out of this was that they love the food experiment format of it. So the Mallory video and the Janet video, they were saying they took one of their normal meals that they loved and they showed the bad response and then they showed a practical tweak that they could do to reduce the glucose spike. And just the format of an experiment really seemed to resonate with people. So this is feedback we should take into the design considerations as we come up with new formats to bring back the challenges or experiments experience.


And I think this is good for all the things that Azure’s doing as well. It bodes well for our future research in the app. But the other big things were practical. It was short, they liked the video, they really liked the personality that was leading it. Next slide. There’s a smattering of the qualitative feedback here. There’s a lot. So if you want, check out the link that I’ll send after this in threads but efficient use of time. I think a big one was just demonstrating that you can still enjoy your favorite foods by pairing it with something that reduces spike. So not everyone’s going to drop everything on day one and become just the healthiest whole foods eater. They can take baby steps towards eating that. But simple, short, seeing someone normal like me, I think seeing themselves and these people is key and practical. So next slide.


So two other slides on results of the experiment itself. Some of the takeaways and learnings, one thing that we learned was authenticity and human connection are very important. So both of these videos were high quality, but in the Mallory video, I think she really was very relaxed and connected with people on an authentic level that just felt like you were chatting with a coworker or something like that. And so that seemed to resonate a little bit better in the survey than the one with Janet. Janet was also an exceptional video, very well done, but it was a bit more produced. And I think for the vibe for these ones is that we’re trying to make an equal playing field for people that they feel welcome and in it and not separate. So that’ll be feedback for iterating on this going forward with other personalities. Next slide.


And the last one was we tried three different ways of surveying people. And the takeaway is that it’s really important to be mindful of your calls to action. And oftentimes you get one shot to get results and you want to make sure that you maximize the results you get. So what we did was we had the in-app video and that was paired with the survey, but you had to tap through multiple screens to get to it. We got 1% of the people to fill out the survey from the in-app route. We also had Sissy work to, she pulled a bunch of people who were related to our target personalities and she emailed them the video and also had a link to the survey. But the primary thrust of the email was around watching the video as it should have been because they hadn’t seen it.


But the survey response request was a bit buried below. And we also didn’t get a very strong response from that one. Something around, I don’t know, two or 3%. And then we sent a third version of this where it was an email and it was both a link to the video but also a call to action in the survey in the first paragraph like you see here, and that one significantly increased the response rate to something like 8%. So if you’re doing an experiment, be really mindful on how you engage to get that feedback because you are going to get one shot and it’s going to be spammy if you ask them again. So some of these subtle takeaways I think can really help. Next slide.


So the next step is we want to test that daily engagement. This was a really strong validation that people resonate with the format, but we don’t know if this is going to actually work day over day. People say one thing, but do they actually follow through with their actions? So we’re going to commission five more videos starting with Mallory and we’re going to try to see how that resonates. Now, content creation, this can’t happen all in one day from some of these influencers because she has a full-time job somewhere else. So one of the things that we really need to do to have a strategic foundation for this is finding win-win ways with our influencers, with our experts and our internal staff and our costs and our budget to scale content while maintaining the level of quality that we need.


And I think this is where it’s going back to what Ben was talking about, really leveraging our team’s expertise in that and figuring out if this does prove out to be successful on an ongoing method, we’re going to want to scale this out, add personalities, and come up with a content calendar and pipeline of things. And so we’re going to have to find ways to do this effectively. And right now that’s a challenge that’s not trivial, but definitely solvable. So short term, the daily engagement and also short term we’ll be jamming on how we can increase this in a scaled way. I think that’s it.

Josh Clemente (51:32):

Amazing. Thank you David and Maz. Really awesome deep dives on product, very excited for this. Those results are amazing. All right, quick hiring update. We have Priya, Evan, and Sean all joining us now. Evan will be starting on Monday, Sean a little bit later this month, and then Priya in October. So super excited to have across a couple functions in the company, a new team coming soon. And then open roles. So still looking for engineers, still looking for people ops generalist and still looking for R&D engineering support. And then generally speaking, if you’re interested in what we’re building here at Levels, if you watched this meeting and felt inspired, reach out to us at or one of the team I think we’ve got our contact info basically all over the internet at this point. All right. I’m going to stop the share here.


We’ve got a few minutes. And we can jump into a couple personal shares, individual contributions. As a reminder, we do this by raising hands, so you can hit the reaction button and raise your hand, like so. I’ll kick us off. I’m hyped up about this meeting. This meeting was … I don’t know if we’ve had as much knowledge density transferred in a Friday Forum in a long time. This was amazing. I learned a ton and it was really exciting. I’m excited about meeting Cary Gunn yesterday. I had a great hangout with him. He’s built some incredible technology. If you want to read about it, I highly recommend just going to Genalyte. It’s basically, this sounds bad, but it’s Theranos except it actually works. It’s FDA approved. There’s hundreds of tests that they’re running in doctor’s offices with the bench shop product. It’s amazing.


And the technology underneath it is brilliant. And Cary is just a very deep, intentional, thoughtful person. And so learned a ton from him yesterday. Excited from more of that. Lastly, Casey, on various podcasts, that was such an inspiring conversation. It was pretty thrilling. And then personally, my wife bought a dog this week and I have not met it, but I’m going to go home and have a puppy waiting for me. So I’ll say that’s my personal share. Jenn?

Jenn Palandro (53:37):

On the Level side, I’m excited about Josh’s dog and everything going on in content, but I’m biased and personally I am excited because my husband’s last day at his full-time corporate job is today and he’s going to start with his skateboard biz starting on Monday. It’s full time, so Ben, I only resent you for the skateboard to car metaphor, but all is good.

Josh Clemente (54:05):

Love it. Ryley.

Ryley Walker (54:08):

Just shout out to one of my favorite Levels weeks. So it was super special getting to ASF to meet everybody and then I guess running back here to see Paul, but I think I got this social equivalent of a glucose spike this week, and so coming down and feeling tired and probably going to lay low this weekend.

Josh Clemente (54:30):

Great hanging out. Enjoy the rest. All right, anyone? Oh, there we go, Casey?

Casey Means (54:37):

I just have to plus one. This meeting was so inspiring and so many amazing things happening. And then personally, I just got back from two and a half weeks in Italy, which was awesome, and leaned into the remote work in Europe thing. I did a week of vacation in the Dolomites hiking with my dad, and we celebrated his 76th birthday while we were there, and it was so inspiring because he just crushed it on huge 3000 foot uphill days and such a testament to just staying healthy over the long haul. He wore his Levels the entire time. Some people may have seen this on Instagram, but he takes his Level stickers and puts them on all of his suitcases, his backpack, his arm, his water bottle. So he was just like a walking Levels commercial the whole time and kept his left sleeve rolled up the entire time because he wants to promote our European international expansion by showing Levels everywhere. So it was adorable, but it was a really, really fun vacation and great to get unplugged and hike a bunch. So super glad to be back. Yeah.

Josh Clemente (55:49):

Love that. That’s amazing. Ben?

Ben Grynol (55:56):

Casey on Bari’s podcast was super cool. So it was knowledge drops left, right, and center, but now listening to these, I’m listening as it relates to that doc that she wrote. So if you haven’t seen that doc, it’s amazing in the way that she preps and suggests people prep for podcasts from like she’s speaking in the soundbite that they’re going to cut for social and it’s just thinking through all these things and it makes it really easy to digest. So love that. And then also super stoked, XinLiu’s back and Caitlyn’s back. So welcome back. It’s great to have everyone back and personally heading to Austin next week with Pam, which is going to be super fun. So excited to go get some heat down there.

Josh Clemente (56:32):

Looking forward to it. Sonja?

Sonja Manning (56:36):

I think this is turning into shout outs in addition to highlights, but my biggest highlight this week, I think two things is just Josh, your Platinum Longevity presentation and Casey, the Bari Weiss podcast, both just really like zoom out, big picture, reminder of what we’re working on and made me feel even more connected to our long-term vision and our ultimate goals. So if you haven’t listened to Josh’s presentation at Platinum Longevity, highly recommend. I sent it also to my parents who said that they finally understand, I think what we’re working on, after millions of articles. So just a phenomenal job to both of you, and thanks for re-inspiring me, not that I needed it, around what we’re working on. Personally I’m bopping around the East coast for a bit and in Ann Arbor for the Michigan game, go blue.

Josh Clemente (57:25):

Amazing. Well, thanks. Shout out to the team, everyone that supported on Platinum Longevity. Again this week, Jesse, Mike, and many others behind the scenes, it went really well and we’re continuing to just resonate with people, so thanks, everybody. Feedback on that talk is welcome as well. Critical feedback. I know it wasn’t a hundred percent. Miz.

Michael Mizrahi (57:45):

We’re just out of time, but I’ll give a 30-second one. The People Fireside that we did yesterday was very enjoyable, so if you didn’t catch that discussion, go for it, but also enjoyed the conversation. Forum was great. The life lessons that Zac is dropping is awesome, so very much enjoyed that. Welcome back, Caitlyn and XinLiu. And then, yeah, great to see Ryley and Zac and all the new hires come in. Excited about that. That’s a lot.

Josh Clemente (58:12):

All we’re right at time here. Action packed meeting. Can’t believe that was only an hour. All right, everybody have a killer weekend and puppy pictures coming soon.