Josh Clemente (00:00):
All right, jump into it. Welcome to the forum, October 7th, 2022, and we’ll get straight into the recent achievements. So a lot of stuff on the blog this week. So we’re considering a restructuring to enhance discoverability, essentially breaking out into buckets of content to allow people to follow a journey of discovering everything that we produced on the blog thus far. We’re also evaluating a why article about our editorial platform. So we take a pretty different approach at Levels to producing content. This is not your run-of-the-mill content marketing operation, and that isn’t always clear to people, I think. They see a blog, they assume, okay, this is just one of those standard things for a D2C Company, and actually this is one of the core pillars of the company’s mission is to produce the best educational information about metabolic health. And so, there’s tons of value locked in there and we want to make sure that people have an approachable article that explains our rationale there and what we’re doing and what the mission is.
(01:06):
We’re also completing an SEO cleanup, so just to make sure that we get rid of old stale code and stop crawling on deprecated articles and things like that. So big improvement there. And then we’re also going to shift to a headless CMS architecture, which is something I am not super familiar with, but essentially breaking out the backend of the content we produce and allowing us to serve it into different platforms in a sort of super agile and modular way. The IRB side, so there’s been a lot of chatter about logistics on the backend for IRB fulfillment and opt-out paths this past week with our partners and made some big progress. We’re super back on track and this is all good news. So some major improvements to improve cost, fulfillment, speed, as well as reducing the burden on the member who is signing up for our IRB. So that’s all great.
(01:55):
For the past five weeks running, we’ve been at a replacement rate of less than 10%. This is a big improvement over a few weeks back when things were quite a bit higher. I think I attribute that to the new platform that people were getting adjusted to some of the differences between the platforms. And then we’re still running into volume challenges right now on the support side. So with Farmsemblage last week, we are definitely continuing to have a lot of inbound support requests and the team is still doing a stellar job keeping up with that massive volume. On the people ops side, we’ve gotten an announcement, our candidate accepted our offer, so our lead candidate is going to join us. This is huge. Big shout out to Miz for running a great process. We had well over a 100 candidates who came inbound for that role and pretty amazing, so excited for that one.
(02:45):
And then on engineering, upgraded the Prisma backend to version four, check out 2.1, so some improvements on the design and explainability of our product shipped, so that’s out, slightly better user experience, we’re going to learn a bit on that. And then weight and height and the new day score, which we’re now calling again, V0.1, which is more of a standard deviation variability score, have both released internally. So the team, update your apps, check these functions out, make sure they’re pulling properly, give feedback. It’s going to be exciting to see how this plays out. And then healthier food choices and recipes, which will be in app along with labs. V1.1 presentation and scoring. V2.2 are both currently in work on the inside. Casey showed up on Doctor’s Farmacy this week again, so she’s had several appearances. This is the first one with a partner offer live.
(03:36):
So we’ve talked about this. There’s some essentially exclusive content that is available through Mark Huberman with Casey and that went live, so we’re learning from that. And then Ben Greenfield and Casey’s episode goes live this week. The Moonshots and Mindsets ad, which is through Peter Diamandis, this is a brand new podcast. Those podcasts went live this week and we are out there in the wild, so we’re tracking progress there. And then Huberman and Tim Ferris spots are live. And lastly, the Honestly Podcast discussions continue. So a lot of amazing stuff. Those are huge names, kind of the biggest names in our field, all in one update, which is great. Digital audiences grew 34% in September. This is actually despite a pretty significant reduction in volume of posts this month. So that is a really interesting signal for us in that the quality of our content is clearly improving and making up for quantity. Also, the virality of the content.
(04:29):
So I think we’re really nailing the way we go about distributing this stuff and there’s going to be a much increased attention to this as Sonia and Casey are going to be helping to continue to improve our reach across digital channels. So this was an exciting update this month and looking forward to seeing continued growth there. We had a huge week for IRB signups. This was a 17% increase in overall numbers since the IRB went live. So pretty substantial. And we had a revision completed for the subject facing material. So one of the requirements for our IRB is that we keep the materials very up to speed in terms of what potential participants are experiencing during checkout. So Static took care of that, research team is continuing to stay on top of just making sure that the moving parts are all taken care of on the backend. Make sure the IRB is well aware of how we’re evolving.
(05:22):
Then lastly, we had some very interesting results, which Azure did a deep dive on. There’s a video that you can see the thumbnail for on the apple cider vinegar community experiment. So it showed what I thought were really compelling results in that the benefits of apple cider vinegar are substantially higher for people who have lower time and range. So people who are experiencing more difficult blood sugar control can benefit more from an apple cider vinegar experiment. There’s a ton of interpersonal variability, but that seems to be a pretty strong finding. So that’s something that we can absolutely introduce into our product and making sure that when we’re releasing these insights that can be targeted at people who could benefit even more. Let’s see, a couple other things here. So some great articles showed up on our blog this month, the month in Metabolic Health, obviously being a favorite, but we also had a Healthy Nut Butters article, which I love.
(06:11):
The first of our sort of women’s health content initiative around higher cardiovascular risk is out. And then that conversation with Andrew Hare font is live. We had Mark Schasker and Emily Strong on talking about Alzheimer’s and cravings, great conversations, and then a ton of love for our Levels’ company culture documents. We’re continuing to just be, I think recognized for both leadership in the culture and company building space as well as metabolic health. And Twitter is really resonating for us. Some other great UGC here and huge response to the seven factors that significantly impact your blood sugar. 23,000 clicks, we made it onto the Google Discover page for this and that really resonated this week. All right, jumping ahead. I want to welcome Melissa McAllister, a partner here at Levels and nutritional therapy practitioner. Melissa, I know you had some really interesting magic moments through your experience with Levels. I would love to just hear generally what you’re excited about in the world of metabolic health, what you’re paying closest attention to and your thoughts on your experience with Levels so far.
Melissa McAllister (07:21):
Yes, thank you. It’s funny, I actually moved just a couple of weeks ago from McKinney, Texas to Melissa, Texas. So now I literally own my own city.
Josh Clemente (07:28):
Amazing.
Melissa McAllister (07:31):
I love this because both my son and my husband are both mechanics and so, the world of social media I think is wonderful in the fact that they can go to YouTube and they can plug in how to change oil filter or how to fix a valve. But on the opposite end, I think with us in the nutrition aspect, we have social media where someone will put out a diet that they think works for everybody. And what I have found is from my own personal experience and then also being able to use Levels with my clients is they are able to see the difference between maybe a diet plan that they thought was working for this one individual.
(08:11):
They give it a try, but to have that CGM on, they get to see that it maybe isn’t optimal for them to where they can kind of play with it. Like you mentioned, the apple cider vinegar or for them to go for a walk after they have a meal or mealtime makes a difference even if the foods are the same. So to me, it takes that ambiguity of social media where everybody thinks that this one diet works well because this fit influencer looks amazing. And to be able to customize it to themselves and to be able to experiment with themselves to see that one thing might work better for them that doesn’t work so well for somebody else. So it’s really been able to help me help them individualize the nutrition plan.
Josh Clemente (08:54):
I love it. And I think the objectivity of the data is such a unique element here. Like you said, there are all these people can pull on essentially any thread that could be contradictory to another one that somebody else posts. And having an objective data stream from one’s own body is potentially the future of all nutrition influencing, I think is being able to show data that backs up the claims and the statements. And that’s something that in concrete mechanic terms, you can always try out and see if something is actually going to fix the problem, but this is unique to the human body. I’d love to hear from your own experience maybe one of the strongest realizations that you had that maybe was counterintuitive. And then really if you could share something that you feel like we need to see in the world of metabolic health in order to improve people’s lives, where are we completely missing? And not just Levels, but just generally speaking, where are we completely missing and where is the gap in people’s intuition or awareness right now?
Melissa McAllister (09:58):
To be honest, I found you guys initially from your blog. I think you guys are absolutely hitting the mark when it comes to that. It’s probably the best research that I’ve ever found. So I know that’s not what you’re asking, but I think it’s really the best out there. As far as missing the mark, I do think that we do tend to fall short when it comes to people understanding that because a little bit of pushback that I’ve gotten is there are type 1 diabetics and there are some professionals out there who think that people like me are misusing CGMs and that they should only be for type 1 diabetics, maybe type 2 diabetics, but not for the common people. And I wholeheartedly disagree with that.
(10:46):
I think if you can get ACGM on most people, at least 88% of the population to really get a grasp on how food is affecting their metabolic health, I think that would be very powerful because if they’re hearing from some people that this is unnecessary or this was not designed for somebody like you, they might shy away from getting it. Where I think that if we really make it clear that this is just a really big and powerful step for them, understanding their own bodies and getting back to being metabolically healthy would be great.
Josh Clemente (11:19):
Yeah, I think that’s part of the educational mission that we’re on. Accessibility to this technology is a key challenge that we have to overcome and that means improving the price point and improving the understanding and awareness of why this might be helpful to anyone. We all have a metabolism. We all experience blood sugar control and lack of it to some extent. And so, being able to learn about one’s own body in the way that a blood pressure cuff or a weight scale or a heart rate monitor are helpful for people today, this goes one layer deeper. So Melissa, if there’s one feature that you would love for Levels to build that does not exist today in your experience with the product, is there anything that comes to mind?
Melissa McAllister (12:05):
From personal experience, I have learned here lately just in the last probably three or four months, and I don’t know if it’s something that can be incorporated, but I have learned that increasing my need on a daily basis. So maybe being able to remind people to take a 15-minute walk in the morning or after dinner or something, but to move more throughout the day because I’m a fitness person, so I do my workouts and I work out really hard, but I sit a lot of the rest of the day. And that had a huge impact on the rollercoaster ride of blood sugar was to be able to just remind myself if the Levels app could remind you, “Hey, kind, I think like the Apple Watch, you’ve been sitting a while, why don’t you get up and take a 10 or 15 minute walk or move a little bit?” I think that would be great.
Josh Clemente (12:52):
Real time interventions. I love that. That’s definitely something the team is thinking a lot about and I think we’ll have some exciting stuff on the roadmap that you’ll love. Melissa, thank you so much for taking some time to join us on Friday. Our team absolutely loves and could not overvalue these interactions and hearing from me directly, so I really appreciate it. And if you’d like to stick around, we do have a pretty packed meeting. We’d love to have you join us through the rest of it, but if not, thanks from the whole team.
Melissa McAllister (13:21):
Okay, thank you guys.
Josh Clemente (13:22):
All right, thank you. Okay, jumping ahead, quick culture slide. This one is just all about meetups this week. So Sam, Scott and Taylor, you can see they’re in Seattle doing a fireside session and then the entire gang out for the Farmsemblage 2022 in Montana, which looked absolutely amazing. Got some horseback riding, there were some visits up to Glacier and some awesome dinners, which I’ve been hearing bits and pieces of and it just sounded like an amazing success. So thanks everyone involved in making that happen logistically, traveling out there and spending time together and getting some deep sync time. Love to see it. Okay, people process culture. So I’m going to chat a little bit about confidence is earned this week and we got a memo on this and it kind of emerged from personal experience and also just watching how the communications culture at Levels has evolved.
(14:12):
And so, what I want to talk about here… Oh yeah, we’ve got a memo on this. If you haven’t read it, definitely recommend doing so. But ultimately confidence is the foundation of a functional organization. So if you think about a professional sports team, people have to have confidence in one another. They can be exceptionally talented individuals, but if they can’t operate together as a high performing team with confidence in the other individual, you don’t actually have success as a possibility. So being that we are a team, not a family, and we are a high performing one, we have to have high communication and reliability amongst each other in order to have confidence in one another. And the interesting thing about confidence is it can’t be assumed. It has to be built and it has to be maintained. So really, in the world of remote and asynchronous work where nothing is within sight unless it is shared deliberately, confidence sort of atrophies very quickly.
(15:06):
So the absence of information is bad information for most people. If I’m trying to execute on something and I am not clear on where the progress is in terms of people who I rely on in order to succeed, I will have low confidence in my own ability to execute. And what that means is now I’m carrying an additional cognitive burden that I have to essentially resolve by seeking out that information. The alternative in our world is that we surface that information proactively and transparently. We take great pride in building confidence in other people and this can feel sort of like a burden, it can feel like a lack of trust, but that’s a fundamentally different thing. And that’s what I really want to zoom in on is that lack of confidence does not mean distrust, it’s not suspicion, it’s not ill will. It’s not a negative evaluation of character, it is the absence of a positive, not the presence of a negative.
(16:01):
So if someone is not feeling high confidence, we each should take the responsibility to restore that and that means seeking out the source. Why is there a lack of confidence here? Is there somewhere where I can be communicating more effectively, more proactively to get this person and other people who may not be as proactive in telling me about this problem to get their confidence back up. So it’s an individual responsibility for each of us to take great pride in building the confidence in everyone around us and earning it over time. So what we’re looking for here is primarily taking that responsibility on, having short toes. So if somebody comes to you and says, “Hey look, I am feeling low confidence in this project, or I don’t know what’s going on, can you give me an update?” It’s not ill will. Again, we have to assume best intent in each other. The objective here is the same.
(16:48):
The way to approach it is to increase the cadence of communication, make sure it’s more discoverable for the people that need the info. So this can mean uncomfortable degrees of transparent communication, making information available proactively so that people don’t have to come and pull it from you. They can go and find it at their leisure and striving for ways to be very transparent about that. So if I’m just communicating everything that comes up to people, that can put an additional burden on the recipient unless we’re clear. So things like saying FYI only or no action needed or action requested is a super helpful mechanism to again, transparently communicate, build that confidence without requiring other people to do added work to interpret it.
(17:30):
And then lastly, this can take different sort of approaches. So sometimes we can achieve a high confidence project or interaction without needing to go synchronous. So a lot of basic comms can fill this gap, but every once in a while just given how fast things are moving and how quickly we can get off track track with just a very small misalignment, even daily synchronous communication may be necessary to maintain confidence in a fast moving environment. So this is all up to our individual judgment. What makes sense given a certain circumstance is up to you. So that’s confidence is earned. There’s a lot more in the memo. Highly recommend digging into it and thanks everybody for working on this one. All right, main thing, Level shows you how food affects your health. No changes here, we’re all marching towards this. If you feel you’re misaligned on this priority, definitely raise that. Top objectives, member attention, new member acquisition and member health improvement. And I’m going to hand it off to Sonia.
Sonia (18:35):
Awesome, thank you. Happy first week of Q4, everyone. So to restate our objectives of acquisition and retention in a different way, it is build a product that people want, sell it and do so while keeping the lights on. And you can see how all of our groups really align to these areas. And the first part of this process is building a product that people want, which is why product market fit is the focus right now and why product is our top priority for this quarter. So if you go to the next slide from you, please, Josh, just like our tagline that we hear every week that Levels shows you how food affects your health to create clarity across the team, each group has created its own tagline here. So the goal is that you cannot understand and remember these just as you do the Levels tagline. So for product, build a delightful and safety product that help our members improve their health. And over the next few weeks, you’ll hear a spotlight on each of these groups and this week we will start with product. So I’ll pass it over to Maz.
Maziar Brumand (19:37):
Thanks, Sonya. All right, this is basically Boz trying to explain to Woody that, don’t worry, and Woody looks really skeptical about this OKR process, but we’re going to go through it today, hopefully we’ll be Boz and not Woody. So let’s do it. Okay, so just going back to the tagline, why do we pick this tagline? It’s the three concepts that we have in our company objectives. So obviously we want the product that’s the delightful that people want and want to buy. We want people to use it for a long time. We want it to be sticky. We don’t want to be a gadget where somebody uses it for a couple of weeks or a month or maybe two months and then just moves on, we want it to be part of their life, part of their metabolic health and really be that trusted partner that will be there for a long time.
(20:23):
And obviously this is all for naught if we can improve people’s health. So all of this is in the service of creating value for people and improving their health. Next slide please. We picked six OKRs for product org, which is product design and engineering. And really what we try to do is split the things that are here and now on the rails and add things that will create a fast engine that will help us find product market fit faster. So that’s kind of how at the high level we thought about this. If you look at the first four, which is content, logging, glucose variability and labs is really the foundational rails to make sure that we can educate people, create content that will help people improve their health, enable them to log really fast, frictionless. So we have the opportunities to give them personalized feedback and information so they can improve.
(21:24):
Growth is variability, which is the core of our biofeedback today, to be able to see that variability is going down and will lead to better health outcomes, whether it’s quality of life or health metrics, which we’ll get into. And then finally, the Blood Panel, which we want people to use more of it. We want it to be part of the core experience because this creates another lens on people’s health and it’s really tied to the north star metrics, which is, is my health actually improving? Is my underlying health actually improving in addition to the quality of life? And then the last two, we really want to be able to iterate, experiment fast. We are in the pre-market, product market fit stage, so we need to be able to make as many bets obviously all thought out and strategic and work in a system.
(22:18):
But we need to be able to iterate fast and not be sitting in our offices dreaming up solutions for months at a time and then find out that nobody wants it. So this enables us to actually do that quickly from a design and product perspective. And we want to be able to focus our stuff, our engineer, which is really the engine of the company on being able to accelerate what the customers care about and not spend our time either bike shedding or working on stuff that’s good for us internally but our customers don’t care about. So these are kind of the overall how we got to these metrics. So if you go to the next slide please. Okay, where are we on these metrics? And we’ll go through each of these one by one.
(23:03):
But right now, we’re on track for three and we’re slightly off track for three. And part of the off track is because we haven’t been thinking about these OKRs, these work streams in terms of OKR, and this is why I think OKRs are going to be so powerful. This is why I think Boz is right, he’s excited about OKRs because he really puts the focus on what matters in a quantitative and quality way versus going to work every day and trying to do the best that you can and not knowing how does that add up to actually the success criteria for these things. So we can jump into each one of these. Next slide please. Okay, so the quality and quantity of content, we feel pretty good about this one. I think there is a lot of work that’s being done. Big thank you to the team on this one.
(23:50):
I think everybody’s come together big thank you to Mike D, big thank you to Casey, Haney, everybody really, this is really a company-wide effort and we have been producing action more than 15 per month. I mean I think can keep that up and even do better. So I think we feel pretty good about this one. The next one’s interesting. So when we looked at general, all of our contents in the app before we kicked off the event-based insights, it was around the click rate of people opening and reading it. It was around 15 to 25% plus kind of our average of insights. And now, we’ve built a dashboard in Snowflake that we can actually see these insights and it’s somewhere in the order of 35 to 55%.
(24:34):
So it is quite encouraging that A, it shows that the content is good and B, actually gives it to people when it’s the right time and see it uses the behavior change paradigms which make it engaging, make it actionable and make it rewarding. And also, we’ve seen a lot of people, 80 to 90% of people are clicking on that learn more, which is actually amazing, which means that they enjoy the content and they want to learn more and then they go into either a short form video, like a reel, a long form article, the great work that Hagan is doing and more types of things that we’re providing. For example, some of the swaps now link to an Amazon Whole Foods dashboard, which basically they can add the swap now directly into their cart. We’re making the content in a way that we think people want it and can actually action on it and we’ll find a reward.
(25:23):
So anyways, the data is good now and we’ll continue to make improvements and push this forward. So overall, feel pretty good about this. Again, health content is the life blood of the app and it will help us retain people and improve their health. Next slide please. Okay, increased logging. So we’ve done a lot on logging, we’ve simplified it, we’ve made it fast. If you haven’t used it yet, the fast logging where you hold down the log button and the camera pops up is awesome. I love it. I don’t know if you guys played with it. Love to hear your feedback on that one. And if you want to add voice, actually at least within iOS, there is a built-in iOS feature where you can hit the microphone and you could actually spell out what you ate in the normal tag. So basically with two taps, you can log now and if you want to use voice feature you can on the at least iOS keyboard feature where you can hit the voice and you can just save what you are and it’s pretty good.
(26:20):
So logging, we’ve been doing a lot of work to simplify, make it fast and also make it responsive based on the event based insights. And we’re not quite at 85 yet, but we feel pretty good. We’re up 80% for people in their first month, so feel pretty good about this one. And again, what does this matter? It allows us to give people the content in a way that’s personalized and the way they will action. Next slide please. Okay, improved glucose variability. This is one of the case of we haven’t… I mean, we want to improve people’s variability or glucose control. This is the goal, but we haven’t really instrumented it in a structured way where we can actually track to see. So the team has been the data science team and big thanks have been running some analysis now to see how are members glucose variability doing.
(27:12):
And we’re using several metrics. We’re looking at a number of spikes, duration spikes, high spikes, and the early data shows that there is improvement. It’s not quite 25%, but there is some improvement and this is something we haven’t measured before. We are going to start measuring it. And also, a lot of the features that are coming out is it’s going to create a structure for people to achieve this, whether it’s through Levels, whether it’s through the metabolic day score, it’s a lot of work going on here to make sure that we actually do impact reduce variability. And also, there’s a lot of work going into, thanks to the research team is what does the reduced variability on day-to-day and CGM mean on our long-term North star metrics. So there’s work going on here. We haven’t fully gone there yet, but is one of the things that we’re thinking about and will make progress and hopefully in the next few weeks this will be on track. Next one please.
(28:08):
Blood Panel, one of the favorites. So a lot of work has been going on into this and there are two aspects of it. The aspect of working with our advisors, big thanks to Rob. We had a great session with him this week, it was really fantastic, really opened our mind into how we can create a product that will help people and people want hopefully so big thanks to him. Big thanks to Cosima. She’s been really trying to work to think through this in a way that will be delightful and big thanks to Victor for really creating designs that look phenomenal. The thing that’s off track here is really around our fulfillment of labs. So as you all know, we’re going through a reconsideration of our suppliers on this, who’s actually going to do the labs, who’s going to do the phlebotomy? So there’s a lot of work going on in there, which the team, Matt, Cosima, Mont, and JM have really been leading with the help from Chris.
(29:08):
So Dick thank you to everybody, but we’re not quite out of the woods there. So I think this will move to on track when we have firm plan that the new vendors will be on back on track. So I think I’m not worried about it, it’s just work that has to be done and the team is on it and executing really well and being really thoughtful about how do we build in a way that it’s more robust but at the same time does not add complexity. Obviously our one of one, number one objectives is not add complexity at this stage in the company. Complexity will slow us down and will bloat our headcount. So one of the big priorities is not increase complexity, but at the same time create a robust system that we can rely on. On the numbers here, we really want to take 10% penetration or more in our labs right now around for 5% of people are getting labs and then we want people to do multiple labs because it’s the trajectory of their health and a check on their health.
(30:07):
There’s a lot of work going on here both in terms of quality of our labs, in terms of integration and in terms of cost. So we hope to actually make 5% of people or more redo labs, which is about 50% of our penetration. Next slide please. Okay, so how does the product market fit? There are many definitions, but the definition I love the most is when so many people are banging on your door, you are struggling to keep up, really, that’s the simplest way to describe it. And obviously our growth has been really fantastic and we’re heading towards this direction. I think we’re on track on the concept of the engine of how do we do that? Really having high velocity of creating versions of app that’s released to customers. This is the whole shift from really spending the whole time on creating the whole experience in one go versus really doing the skateboard where we do the thing that is a full thought but then build on it through releases. And I think that’s worked really well and well on track on having 80% of our releases be really around five weeks.
(31:21):
Obviously the big rethinks like when we switched to the ring, that takes a lot more work because there’s a lot of the system that you have to rethink, it’s a fundamental shift and we were okay with that and that’s really part of the 20% that needs a lot more thought will take longer and that’s okay. But in terms of how do we actually rev these and improve it, we really want to be nimble and get feedback so we can get the product market faster. Next one please. And I think the last one, all right, engineering. So we want our engineers to focus 60% or more of our time on having direct customer impact here. So these are features that people see, feel, and love. And so, we really want to make sure that 60% or more are here and then 40% or less is really on the internal self, whether it’s QA, infrastructure or total tools and things that the customers don’t directly benefit from.
(32:14):
The reason for this is again, tied to product market fit. If we’re working on a lot of internal stuff, for example, let’s say we spend all of our time creating beautiful tooling internally but don’t have features that people want, we’re just not going to have the results that we want. It’s more akin to we’re currently in the exploration phase. In the exploration phase, you want to experiment a lot with customer features and the exploitation phase where we have a lot of customers. Imagine if we had 10 million customers, then it’s really important to make sure our internal tools infrastructure is really robust. And so, the focus can shift in the future, but for now it’s really important that we keep more than 60% of our engineering resources on building experience that people want. And why is it offtrack? It’s mainly due to one project. It’s the e-commerce project that we all went into it with expectation to make the system a lot more robust and scalable given that it was a lot of things, a lot of tech that we had taken on to move fast but redo it.
(33:18):
We went through a very detailed exercise to break down this into phases and define what is and isn’t necessary. And so, I think we’re getting on track here. I’m really confident that e-commerce is looking up and we can deliver it in a phased approach. So hopefully this will go on track, but the main reason this is off track is because of the e-commerce projects, taking a large chunk of our backend engineering to redo. It’s necessary change that we need to make as we go into multiple geos, multiple product SKUs, different lab, different vendors. So it’s necessary but it’s hopefully on track to be done and phased out so that we don’t spend a ton of engineering resources on that past the month of November. I think that’s it. Hopefully that was helpful.
Josh Clemente (34:07):
Great dive, super helpful. And I think we transitioned straight into David.
Maziar Brumand (34:13):
Actually I’m going to do the first slide and then pass it on to the respective folks.
Josh Clemente (34:18):
Cool.
Maziar Brumand (34:19):
All right, next slide please. It’s funny how it’s back to back this week. Okay, so what are we working on? Levels Levels, a really fantastic progress. We spent a lot of time at assemblage working together and with Casey, big thank you. I can’t thank Casey enough for really helping us think through some of the interventions. She created a map of everything from customer need to physiology to intervention and then subcellular path. It was just a work of art. So thank you Casey, that will really help us shape this experience in a way that is science-based. People want to use it, it’s learning by doing. It was just one of the highlights of assemblage for me. So thank you for that.
(35:08):
And obviously big thank you to David and Alan for really pushing the envelope on this one. And they’re just a machine. Anything from creating simulators to see how points can be awarded in a way that feels right, to create an experiences that is very thoughtful and later. So really impressed here and hopefully you guys all get to see the next rev when we’re ready to share the version there as UX are happening this week and maybe early next week. So we’ll update it based on that feedback, but really good progress there. That was 1.1 and 2.0 great work going on. We talked about that. We’re continue to push there in-app content. Again, really good momentum. Thanks for the team, Jen, Heidi and Caitlin to jump in and really offering to help here to not only increase our velocity but increase our quality. So this is really fantastic.
(35:59):
And then Stacy’s working with Mallory to create the videos. The first video should be ready soon, so really optimistic on this plan. And then we’re talking to figure out how can actions create a media platform that we can use potentially inside the app with Ben and his team. So really bullish on this and look forward to see how that shakes out. Metabolic score, David, I believe will talk more about this, but this is the idea of bringing back a version of a metabolic score that’ll be a supplement to our rank stability and spikes, which really help people figure out day-to-day, how they’re doing. David’s going to talk to that and it’s ready for internal testing.
(36:36):
Personalization, a lot of good work last week at assemblage as well. We talked through this, how can we actually phase this out and start simple and add more features and features. It’s great celebrations. This is the idea of being able to see visual feedback when you do things that are positive. And then finally, commerce. This is the idea of integrating with other platforms to help people change their pantry and their environment, but also increase their ability to find things that are metabolically positive. So with that, turn it over to I believe David.
David (37:13):
Thanks. And just before we get into this, Stacy informed me that the second Mallory video is ready and it’s pretty good. So she’s going to drop an update in Threads after the forum so you can take a look at that. I wanted to give you a quick update on one of the projects, which is as Maz alluded to, the metabolic score variant, which we’ve been calling the scoring abstraction in the past. And the reason we are thinking about this is just to remind us based on the feedback we heard from our members when we rolled out the stability ring. They love the stability ring and how understandable it is and how spike detection really makes it obvious where their opportunities are and how encouraging that is. But one of the things they missed about the metabolic score was the trending day. Every day they could very easily visually glance how a day was in my data and compare back and forth how good or bad that day was compared to other days.
(38:01):
And so, we’ve been working across data and the backend engineering team and designed to figure out, could we bring back something that helps us see the day every day trending, a quick way to grok was our day good or bad visually. But also not promote it back to the same level that it used to be. And also, it has to be super explainable. So the model we came out with was using a score base and standard deviation, which is basically variability that can be very easily mapped and explained using the stability ring, the spike detection, etc. So our ask to you is that this is now actually ready for testing. If you go and update your apps, you can click to my data and then you can swap down in the metrics selector, just choose Day Score. Actually Josh, can you refresh this? There’s a link here to a survey.
(38:56):
Yeah, sweet. So you can take a look at it on your data. I want to let you know that we tried to spend as minimal amount of time as possible on this so far. So even though it kind of looks designed, it’s actually not designed, what we’re asking you to do is take a look at the number and tell us does this number feel accurate based on how your stability was for the day? Does it feel too easy? Is it too hard? Is it just about right? And let us know in that survey link below, I’ll send that out in Threads as well. But levels.link/score-test and give us some feedback. Does the score seem reasonable for you? Based on that, we’re going to go ahead and incorporate it more intentionally into the app.
(39:36):
What we want to do is right in these screenshots you’re going to see it looks like the old metabolic score treatment, but we want to incorporate it in such a way that it’s married to the spike time, the spike count, things like that. And we’ll figure out a way that it holistically wraps up. So take a look. There’s a video demo in there as well that reminds you of what to look for and we’re really looking forward to this. This was really kicked off by member feedback. So huge shout out to our members for testing out the stability ring. That’s it.
Josh Clemente (40:13):
Awesome, thank you.
Alan McLean (40:15):
Hey folks, a quick one today. The next slide please. So as mentioned, the dashboard is getting some updates. We are preparing for Levels Levels, which is going to require some restructuring of the homepage. And one specific point that we’ve heard in response to user feedback was they’re sometimes missing the list of insights at the bottom. So we want to make sure we’ve got that little chip that’s probably not quite enough that this slide is cropped off. But for most phones, by moving the text that we have below the ring into the ring, we should gain back some precious vertical space. People will be able to see larger graph and see more insights dropping in below. So that’s a nice big change. But also, you’re going to have glucose in the top. That was another request from users. We’re going to give it a try. We’re going to see how that feels.
(41:11):
There’ll be a little toggle and you can tap on their ring and it’ll transition to glucose and that’ll persist. And if new things come in, little rewards or state changes like your spiking or crashing or recovering, then it’ll transition quickly to that and go back to whatever your preference was set. So we hope that also satisfies some of our members. Yeah, happy to see that go out. Next slide please. Oh wait, some more examples. So in the middle you’ve got an example of having the text in the ring with it spiking. There’ll be a version on the right there, earning some stable time. I’ve got a nice little transition and then a little toggle on the left. So you can want to keep the stable ring for many people, but if they want to switch back and forth they can with little tap.
(41:57):
Yeah, and that’s it. Next slide please. And I think some of this has been covered before, but I’m just so excited about it and it’s looking so great. So I want to bring forward again, labs 2.0. The visual design is being iterated on by Victor. It’s looking really good. We’re going to see a line graph so you can show historical change. We have some prompts there on how to improve your results and even just getting you to do it again, we know that’s going to be important to see that change. And so, we want to create a long-running relationship with people and plotting out the work that they’re doing every day against their overall metabolic health through results, through the test results.
(42:32):
Next slide. So some more examples of that. I think one nice thing here also is that Victor’s hard work on the design system is starting to propagate. So we’re going to start seeing some of these refreshes in terms of typography and color and design treatments. So the app is starting to feel a bit fresher, a little bit cleaner, and overall just more together and taken care of, which of course, we’re really excited about. And so that’s it for design. Next week, maybe Levels Levels.
Josh Clemente (43:01):
Something that stands out here to me is just the beautiful spectrum on these ranges, which we all know, these are not demarcated buckets that you fall into. This is a spectrum and so the visuals here really support I think what the educational material is describing. Love it. Thanks team. Right, we’ve got-
Cosima Travis (43:21):
I have one quick PSA that the UK closed, beta retro is out. So anyone who wants to check it out, it’s in the product and design forum.
Josh Clemente (43:30):
Yes, and actually something that I didn’t get it into the main page, but the UK beta, the first 140 invites I think went out to the wait list this morning. So that’s a big milestone that should have been mentioned. All right, over to Dan, who’s going to dig into the algorithm.
Dan Summers (43:54):
Yeah. Thanks, Josh. Hey, team, happy Friday. I’m here today to talk about a concept known as the algorithm and why it’s important to what we’re building here at Levels. So in short, the algorithm refers to the five steps listed on the slide used in the context of developing new products or technologies. So it’s best to view each of these five steps not as a literal business process, but more as a mindset. A mindset that begins by acknowledging that true innovation only occurs when previous assumptions are challenged. So step one, make the requirements less dumb. Said another way, challenge the requirement, challenge the assumption for our company in the context of Levels, our mission is to solve the metabolic health crisis. This fundamentally challenges the assumption that metabolic health overall is not in a crisis state. Our proposed solution to this problem is providing individuals data insights to better understand how food affects their health and the status quo assumption made by many elements of the healthcare industries that certain diseases are unavoidable, that reactive treatment is the priority for innovation.
(45:11):
We obviously disagree with that assumption as a fundamental aspect of our mission. So when you hear ‘make the requirements less dumb,’ just think of this analogy. Step two, delete the part or process step. The best part is no part, the best process is no process. I’ve heard this a 1,000 times at SpaceX. This is not meant as an excuse for laziness. In fact, if we relate this back directly to that analogy I just made, you do not need medication to treat a disease that could otherwise be prevented by changing your diet or lifestyle. The best medication is no medication. Step three, optimize. This is where most people start. You have a baseline, you work to improve it. That’s all well and good. Assuming you’ve completed steps one and two, I’m sure all of us can think of examples where we spent time or have known very smart people around us who have worked tirelessly to optimize a process only to find out that the thing they optimized was not needed and should have just been deleted in the first place.
(46:20):
So don’t fall into this trap, descope before you optimize. Step four, accelerate. Moving faster can feel reckless or chaotic at times. Accelerating something requires an increasingly strong force to be applied. Otherwise, that thing will slow to a stop-over time. Being that increasingly strong force can definitely feel uncomfortable. And that’s okay to reinforce a few of our other cultural values. Take the big swings. Don’t ask how you can help, just help. Use your best judgment, we trust you. Step five, automate. This is an increasingly popular term as Zach pointed out a few weeks ago in this forum. To make yourself obsolete often requires some amount of automation and automations are not easy to implement. The key here is to make sure that you remember automation is the last step in the process, not the first, worse than optimizing something that should not exist is automating something that should not exist. This only compounds complexity over time and makes it more difficult to be agile.
(47:31):
So lastly, I included this visual representation of the algorithm on the right showing a development version of the Raptor engine on the left and a production version of the same engine on the right. You can clearly see the production version of the engine is visually more simple, where numerous parts and processes have been deleted. The deletions here were primarily achieved by challenging fundamental assumptions and requirements about what was physically possible for the design. The production version of this engine ended up being orders of magnitude less expensive to build and operate and actually had better performance metrics on thrust and reliability. So just to recap, we’re leveraging the algorithm all over Levels every day, even if maybe we don’t realize it, all of us should try to put ourselves into the state of mind where challenging the status quo becomes comfortable because that’s inherently what our company mission is all about. If you haven’t read the memo on this topic, there is a memo on this, read it. If you have read it, great. Read it again. That’s all from me. Thank you all.
Josh Clemente (48:42):
Awesome. Thanks, Dan. I love abstracting this concept and just applying it everywhere, even in one’s own personal life, which I’ve started to increasingly think about as well. So I love it, I love the visuals. Appreciate that. Quick hiring updates here. So Priya, and as I mentioned at the top, and Miz, if you want to jump in and add more, please do. But Nicole Miller has agreed to join us as our people operations generalist starting October 17th, same date that Priya is joining. Super stoked for both of those. Miz, I’ll let you add a little color if you’d like.
Michael Mizrahi (49:19):
Yeah, I threw in an aloo into Thread, so plenty of context there just to give some background on the welcome video thread.
Josh Clemente (49:28):
Perfect. All right, excited for that. You’ll see our roles have been updated here as we’ve removed the people ops role, still looking for RD engineering and of course, software backend and mobile, and always looking for general culture and role fits. Please reach out if you’re interested in joining us here at Levels. All right, made it to the individual contributions, we’ve got a couple of minutes here. I’m going to stop the share. Remember, we run these by using the raise the hand reaction like this. I’ll go ahead and kick us off. I am professionally excited for just product and design and eng, those three core elements of our business just humming along. And that dive today on the OKRs and kind of seeing it all line up into an execution plan is really exciting. So, stoked there. Personally, I’m going to go to Austin City Limits this afternoon, our first real Austin event that we’ve done in a while. And I’m looking forward to it. Sam.
Sam Corcos (50:36):
For me, I’d kind of say it’s sort of both personal and professional. Yesterday, I spent a good chunk of the day playing around with DALL-E 2 from Open AI and Stable Diffusion, which is an open library for it. And it’s really cool. I’m curious to see with a little bit more effort if we could 10x the amount of content we’re able to produce in a really high quality. So just an interesting exploration. I’ll probably have more to share on that later.
Josh Clemente (51:11):
Love it, Andrew?
Andrew Conner (51:14):
Yeah, I want to give a quick professional excitement for Jin Lu being back. She’s been picking up a ton of just really high value things just improve everywhere across the code base. And so, I’m just incredibly happy that she’s back and contributing with Gusto. It’s really exciting.
Josh Clemente (51:35):
A 100%. Welcome back, Jin Lu. All right, anyone interested in a share? I will say that I’m also quite excited about the comms tool that is rolling out. We’ve migrated the R&D team over and I can see the light at the end of the tunnel in some ways. Obviously there’s room to continue to improve. It will be an experiment, but the performance and the consolidation of notifications by priority is really exciting. Taylor.
Taylor Sittler (52:15):
Hey, I just wanted to say a quick thanks. I wasn’t able to make the forum last week, so I want to say thanks for the thanks. And also, I know a lot of people probably talked about it last week, but Farmsemblage was amazing and just wanted to give Chris another shout out. Being able to walk out on Glacier National Park was fantastic and the whole way that he organized it, I was just blown away. So thanks again, Chris, made for an awesome week last week.
Josh Clemente (52:46):
Yeah, pictures looks incredible, Miz.
Michael Mizrahi (52:51):
You’ve had us on our toes here with these last minute hands. Just keeping the kudos going I guess, to Chris, Cosima, JM, Moz, and a handful of people that I’m missing. A lot of this kind of true pill scattering has been very interesting, but we’re firing on all cylinders to get it right. And so, I think we have a huge relief that the pendulum came back a little bit on some of our initial expectations, but we’ll be stronger in the long run with a lot of the projects that are running. So thanks to everyone that’s hustling on that. I know these rushes aren’t fun, but it’s been cool to see the team come together. And then on the personal side, I was traveling last few weeks for Jewish holidays and conference. Got to see some people in Austin, which I very much enjoyed and I’m equally as happy to be home.
Josh Clemente (53:36):
We didn’t get a picture of the Austin Meetup this week, but we did have one of those. We had a bit of a startle last week and just seeing the team quickly shoot out of the blocks and address it, it was pretty awesome. It’s good to stay on our toes. We’re still scrappy and it’s good to feel that confidence. All right, Maz.
Maziar Brumand (53:59):
I just wanted to add to that. When the news came out, I think it was Thursday and we were obviously there and Chris was calm as a cucumber. I mean, I’ve never seen anything like it. He’s just so professional and while he’s managing 20 people at his farm and all the craziness, we just had the most logical, straightforward conversation about it and he followed up on everything. I’m just super amazed. One big thanks. Obviously Farmsemblage is a big effort and I think everybody pitched in, but Chris took the burden of it and then handling this in parallel to that was just really amazing to see. And he was so generous with everything. He lent his car to us while we were there so we can go into the park. I mean, who does that? It’s just the quality of character is just off the charts. So a big thank you to Chris and obviously seeing everybody is just fantastic.
(54:54):
I think taking the opportunity to celebrate these rare moments together was just amazing. I think just seeing people, John was there for example, from Columbia. I think there’s only a handful of times that he’s been here, just happened to be in the US and joined. I mean, these things are super valuable and just being able to manage it with obviously the strong a culture that we have is going to be critical. But I just thought the team… It was just an amazing experience and big thanks to Chris that made it all happen. And then obviously people pitched in Sissy made it all happen on Saturday. I don’t know how she did, but she also made the hikes possible. So anyways, really fantastic. I just wanted to add.
Josh Clemente (55:31):
Huge. Thanks, Chris. Rebecca.
Rebecca (55:36):
I’m just going to echo that Farmsemblage was amazing. I never would have guessed that just after a few months of working here, I would have an opportunity to meet so many of the team members. I remember when I first started looking at the company, bingo, thinking like, oh, I don’t live near New York. Oh, I’m probably not going to meet these people that soon. And then here I am. Now, I’ve met most of the team. So that really amazing and seeing a lot of gratitude for that. And thanks to Chris and Sissy for planning all the hikes, it was really an incredible experience.
Josh Clemente (56:07):
Amazing, Miz.
Michael Mizrahi (56:09):
Yeah. I’ll add a quick note on the assemblage from assemblage. Bit, a lot of really helpful learnings coming out of this. I think it was the first kind of large meetup that we had separate from the Austin one, early in the spring. So I think we’re going to take back to our notes and figure out how to move forward, what this looks like in a virtual environment, what it looks like if we do in-person meetups in the future, and how they coexist. And so, a lot of learnings all around. Thanks to everyone that attended the virtual assembly events. Pumped for everyone that got the in-person time as well. And we’ll figure out what our strategy is for the longterm. But this was very helpful to learn from and I’m happy that it was as productive and enriching as it was for everyone.
Josh Clemente (56:48):
Experiments. All right, it looks like we’ve got a couple of minutes back, which I was skeptical of considering how action packed today’s meeting was. Go have a great weekend. Thanks everybody, and remember the algorithm.