Portfolio Review - September 2020

As many of you know, I started at @SagaPartners last month. I want to reiterate the views I share are entirely my own and are not intended as investment advice.

September started out incredibly strong with Zoom reporting a mindblowing quarter and starting a SaaS buying frenzy but that proved to be too much too quickly and the portfolio pulled back before making new highs at the end of the month. Since I’ve started a number of new positions this month and I want to provide more commentary on my positions in general, I’ve decided that Substack would be the better format.

One of the new positions that I started this month is a healthcare company and I will be revealing it when I release my write-up soon with an announcement related to my role at Saga.

The other new position that may come as a surprise is DarioHealth (DRIO), a micro-cap in Livongo’s space that I’ve ignored because its financials haven’t been good but I have reason to believe a turnaround is likely. I am very bullish on the future of digital therapeutics and think barriers to entry and switching costs are much higher than in telemedicine for example. But because this company is largely unproven, it is a very risky idea and I intend to keep the position very small to see if the story plays out. I don’t recommend that anyone follow me or purchase any stock without doing proper due diligence first. Since this review is already getting too long, I will make another post introducing it.

At 14 positions (counting LVGO and TDOC as one), my portfolio is larger than I would prefer it though still quite concentrated with my top 4 positions making up more than 70%. Of my 5 new positions, I am most bullish on TTD and will keep it for the long-term. I am still developing conviction for PTON, GDRX, and the yet-to-be-revealed name and will re-evaluate accordingly after future earnings. DRIO is more of a moonshot and the next two earnings are make-or-break.

Performance: +203.9% YTD

Initiated: DRIO, TTD, PTON, GDRX, ???


Trimmed: CRWD, ZM

Sold Out: TWLO

Teladoc / Livongo

I think it's easy to underestimate just how much COVID has accelerated the adoption of telemedicine. TDOC was growing at an excellent 74% CAGR from 2013 through 2019. The headlines say that telemedicine usage is "plunging" but when you look at the data, 21% of ambulatory visits is still an extraordinary increase. A lot of talk on how e-commerce penetration has doubled (10 years of growth) in 3 months but providers are seeing 50 to 175 times the number of patients via telehealth than they did before!

Is this heightened level of usage sustainable? Well not only providers and patients acclimating to its benefits, but so are regulatory bodies. The NCQA taskforce on telehealth released its final report last week. They recommended lifting geographic restrictions and limitations on originating sites which makes telemedicine a much more viable product because it allows for "load balancing". What that means is that since providers are now able to see patients in all 50 states, we can better pair up demand with supply and enable access to better care in rural or underprivileged areas. The task force also found that "the evidence base for telehealth is strong, particularly when it comes to the remote management of chronic conditions".

In addition, Jason mentioned that "CMS says that they will reimburse through the end of 2021, but they are not going backward. This will become an issue of what's the level of reimbursement." So, this is huge in terms of adoption because hospitals couldn't afford to implement it if they had to pay the telemedicine provider and were making much less than in-person visits. It simply wasn't a realistic option, even if providers wanted to offer patients that flexibility.

Of course, TDOC can't capture the vast majority of the shift right now because most of the demand is still with health systems (as evidenced by AMWL growing slightly faster last quarter), but over time, I believe it is well-positioned to outperform AMWL because it is positioning itself within health systems with InTouch and now has a better, holistic offering with Livongo, as evidenced with the health plan executive saying that they will "have to work with them" and the taskforce discovering the benefits of remote monitoring.

There is a lot more that I want to discuss and will release more research in the future. But I’ve added significantly to TDOC and LVGO throughout the month. LVGO is still priced below the exchange ratio but the taxes on the cash makes up for it. So I’ve been adding LVGO in my tax-free and TDOC in my taxable. See my tweets below for a recap:

Updated guidance, helpful investor presentation:

Some very bullish management commentary:

Why remote monitoring is critical to building a great telemedicine platform:

Why synergy estimates are conservative:


Great quarter, revenue growth of 84% which is about the same as their last quarter and guiding for 72% growth next quarter. 124% net retention, 78% non-GAAP GM, and 16% FCF Margin, 969 net new customers (+91% growth), and 3.9% net profit margin. Margins were a little weaker than last quarter but other than that, nothing really to complain about. And if they beat by the same amount, they have in previous quarters then we should see another quarter of 80%+ growth.

CRWD is in a really advantageous position right now, WFH has increased the demand for cloud-native endpoint protection platforms significantly (Gartner expects cloud-delivered EPP solutions to grow from 20% of new deals to 95% by the year 2025), legacy vendors lack the ability to compete with CRWD's complete solution set as I explain below, both in terms of being bogged down by legacy architecture (re: the lightweight agent) and lacking the data (re: Threat Graph). As a result, competitors are getting bought out: Symantec (by Broadcom), Carbon Black (by VMWare), Endgame (by Elastic). In particular, there has been a mass exodus out of Symantec and George mentioned that it was the best competitive landscape he had seen in his 27-year career.

For more on CRWD, Chris released an excellent free report (among others) and Robert wrote a very comprehensive SA article on it as well.

This is another quote from the recent CC that captures George's thoughts on CRWD's competitive advantage.

"We spent the time and effort to build the platform out from the ground up, right? It's the same Salesforce, Workday, ServiceNow, CrowdStrike. We don't have an on-premise version, because that's not our model. So a lot of our competitors built on-premise versions, they try to move it to the cloud and call it a cloud offering, their data is still on each endpoint. The value is being able to aggregate this data at scale, which we figured out with our threat graph. And effectively, that creates a data mode plus the module expansion allows customers to add more modules not agents, right? And even our next-gen competitors, they still have three and four different agents because of their acquisitions. So people want something that's simple, want something that works and want something that's future proof and ultimately stops the breach."

I trimmed a little this month because I think its moat is a bit weaker than my other holdings. Frictionless deployment means that it is easier to replace than a ZS, OKTA, or NCNO, although no one else on the market can match it right now. Sentinel One is the closest comp in the endpoint market and they said that they see the market eventually evolving into a two-man show between themselves and CrowdStrike. Right now though, Sentinel One's capabilities are less comprehensive.


Datadog continues to perform well, although revenue growth dropped to 68% and they’re only guiding for 51% growth on the high-end next quarter, if they beat by 10% as they have in the past then that’s 66% growth. Margins and customer growth remained strong. However, commentary on the call suggested a “notable improvement” in usage in July. Whatever the case, the observability market remains in its early innings, Datadog remains the category leader by offering a frictionless, holistic platform, and cloud adoption has been accelerated by the pandemic.

Datadog also just announced a partnership with Microsoft Azure. This is not only a significant validation of its platform but makes implementation and integration seamless and opens up a new sales channel via the Azure Marketplace. This reinforces Datadog’s ease-of-use and should produce increased ROI for mutual customers.


After reporting one of the best quarters in enterprise software history, it was hard to believe that Zoom could outdo themselves again. But they certainly did, shocking the market and proving that we are still just beginning to understand the longer-term implications of COVID.

Revenue grew 355% year-over-year, customers with more than 10 employees grew 458% year-over-year, they achieved 71% gross margins, 42% non-GAAP operating margins, and 56% FCF margins! They also expect growth to continue, guiding for 314% year-over-year revenue growth next quarter and guiding for 282% growth in FY21. The quarter was so good in fact, that it kickstarted a FOMO buying spree in other SaaS names that had yet to report earnings like DOCU and CRWD. Although this move was quickly retraced, it looks like it was just another temporary pullback on-route to new highs.

I ended up trimming Zoom, not because I could find any fault within their results, but because I had a hard time seeing how this growth is sustainable in the long-term. As I covered in my Agora write-up, my Zoom and Agora theses are a bit at odds with each other.


I added to Roku after the Peacock deal, which has streaming rights to premier league soccer. It’s a big win over Fire TV which still doesn’t have it and positions Roku well to get HBO Max in the future. Needham estimates that the deal "could add up to 5%-10% to Roku's pre-Peacock ad revenues in 2021, at 50-70% gross margins."

This is a great summary of the Roku thesis for those wanting to learn more:

I also came across this great tweet on why hardware and content providers should remain separate. He mentions Peloton as a company that combines both and thinks that may be problematic, I agree and expand on that in my Peloton section:

My take: If Netflix’s job is to commoditize the hardware providers then it makes sense to be available on as many as possible to become an essential service and thus avoid paying rent. Roku’s job to commoditize content suppliers so it makes sense to have as many as possible on the platform and leverage their economies of scale to gain market share in hardware. Over time, Roku’s cost leadership forces smaller competitors out and Netflix becomes forced to pay Roku rent. Netflix has to be on Roku right now because they would take a big revenue hit and allow a competitor like FireTV to gain market share which basically restarts the problem. So, I think Netflix’s job over the long-term is harder.


I have always been a bit cautious on Fastly, although its recent COVID bump has been significant, I question how sustainable it is. Last quarter, enterprise customers only grew by 7. That's a deceleration while enterprise makes up more and more of their top line. Sure, it's a pandemic but with FSLY's valuation, there is no room for error. Here's the trend: Q3'19 – 12, Q4'19 – 14, Q1'20 – 9, Q2'20 – 7. Existing customers are spending more though which is great and expected (DBNER of 133% -> 137% is amazing). Nice gross margin improvements too (61.7% from 55.6%) and EBITDA positive for the first time. They also enhanced Compute@Edge with new observability features, including — logging, tracing, and granular, real-time metrics. The platform is scaling well to meet the increased usage. Forward guidance pegs them at 51% growth for the next quarter. Slight deceleration even with a similar beat as this quarter.

There is a lot of hype around Compute@edge without much evidence yet in exactly how much it will contribute to revenues when it launches. (It is planned for limited availability in 2H2020 and general release in 2021). Peer analysts who have a deep technical background like Muji (hhhypergrowth) and Peter (Software Stack Investing) certainly believe this will be significant. In a recent write-up, Peter noted that Fastly is in a good position to dominate this nascent but high-growth market both because of a tech advantage (FSLY's solution can start-up in 35 microseconds, which they claim is 100x faster than competitive solutions) and because of their background as a CDN, cached data is already being stored locally by Fastly’s edge network, so it wouldn’t make sense to transport the data somewhere else to process.

Clearly, FSLY expects this boost to last more than one quarter as well, as they've raised guidance significantly and customers like SHOP and AMZN are experiencing record traffic. It's not a surprise that the market is expecting an even bigger boost going into 2021. The permanently increased demand for bandwidth from COVID is a catalyst both for their CDN business and for the need to shift data processing from the traditional centralized cloud to the edge.

What got me to add more recently is their acquisition of Signal Sciences. This expands Fastly’s security capabilities significantly with Secure@Edge, which will help drive Compute@Edge adoption. It also seems their cultures are very similar (developer-first approach and company values). What is even better is that this is accretive both to growth and gross margins. Signal Fire is a very strong company with 5/5 score on Gartner as well as an NPS of +80, gross margins of 85%, and they have 265 customers and 60 enterprise (100k+) accounts, 70% of which will also be new customers for Fastly. These include Datadog, Under Armour, Twilio SendGrid, and DoorDash. They also have five of the top e-commerce companies, five of the largest software companies, and many others in the financial services, retail, healthcare, media and entertainment, manufacturing and educational sectors. So, this will be great for cross-selling opportunities.


Agora held up well during the initial SaaS sell-off at the start of the month but faded slowly for the remainder of the month. Price action has been disappointing but I’m not selling as nothing changed with its fundamentals. They recently announced that they will be hosting its annual RTE2020 conference from October 13-14 which might provide some exciting updates on product development. Some people have voiced concern over increasing US-China tension. I’m not really concerned as Agora has a second headquarters in the US, and last quarter mentioned that US revenues were only in the high single digits at this point. The biggest market for live streaming is in China now and US is a longer-term opportunity.

I think the primary risk is Twilio moving more in this direction. Twilio recently introduced Edge Locations which is a PoP network similar to Agora. They have 8 public locations globally so far vs Agora's 250+ but it is still troubling given Twilio could easily cross-sell to their existing customer and Agora doesn't seem to be making much progress in the US. China revenues have more than made up for it so far and Agora owns broadcasting which Twilio does not do at this point and would have to spend a lot of time replicating so I do think that once broadcasting really takes off here that there will be more of a market for Agora. It’s also harder to evaluate as much of its customer base is outside the US, but if you have first-hand experience with Agora, please let me know your thoughts.

For more on my Agora thesis, check out my write-up

The Trade Desk

The Trade Desk was a stock that I had owned in the past but sold after revenue growth declined and the valuation got expensive. After joining Saga, I read Joe’s take on it and realized that it did indeed have a very strong moat and more potential than I had imagined.

A Demand Side Platform (DSP) allows advertising clients to buy digital media on an ad exchange where Supply Side Platforms (SSP) help publishers sell their ad space. When DSPs like Trade Desk run ad campaigns for clients, they “look” at the available inventory on the exchange, analyzes the impressions, and place bids on the auction.

The key is there is a cost to the DSP every time they look at an auction and every time an SSP sends an impression to a DSP, it also costs them money. Therefore, SSPs send impressions to the DSPs that are most likely to win each auction, hence smaller DSPs with less inventory to choose from will find it hard to deliver the same value proposition to advertisers. Furthermore, since DSPs and SSPs only get paid if the auction is monetized, smaller DSPs with lower ad spend will not be able to take as many looks and thus win a fewer number of auctions while incurring the same costs. This will push further consolidation with DSPs, helping ensure The Trade Desk remains the category leader.

For the full thesis:



I’ve been thinking about adding to Alteryx, if one expects it to rebound when the economy eventually recovers, then it's quite cheap here (P/S of 17x, GMs >90%, FCF margins of 30-35% at scale). But I have been keeping tabs on job openings using @aznweng’s platform and there has not been much of a recovery.

Source: @aznweng on Twitter

Combined with management indicating that it expects no material improvement for the rest of the year and it has me questioning what could be holding them back when most others have indicated they've hit the bottom. Of course, the stock could rebound before the job openings do and I'd miss my opportunity to buy but I'm comfortable holding my current position.

Not much in terms of news this month, but Dean did a very informative interview.

He acquiesces that analytics is a value-add rather than a short-term priority and customers are prioritizing cybersecurity and collaboration tools. But notes that in the long run, analytics is as important as ever.

Some highlights:

  • Pricing is not an issue, 250 use cases across every sector, customers are seeing 757% improvements in operational efficiency. Even after the 30% price increase of Server, they aren't seeing significant impact on implementations

  • Adoption licenses are used to gauge demand from a customer, because companies don't know how many citizen data scientists they have, 5-8% of all workers are drowning in VLOOKUPs, most don't even have analyst in their job title. Even PwC which is a 55,000 user account, there is still room for more expansion. Adoption licenses convert faster than typical land and expand.

  • AYX has two main experiences: a design time experience, where you build analytic pipelines, and a run time experience, where you execute these automations that occur in the business. You can deploy in the cloud today. AYX believes there in a hybrid cloud future, they have 37% of the G2K and there's a lot who still have 0 data in the cloud. But they said we are going to see something in cloud next year, they've been talking about a browser-based designer and have one in alpha. They want to wait to see what's going to happen first in terms of market dynamics.

  • AYX uses SNOW in their own operations, AYX server sits next to SNOW and they build their design time experience in this line of business and populate it to server. SNOW's success is bullish for AYX because they want compute loads to monetize, AYX makes it easy for compute load efficiency in SNOW. They have several hundred joint customers without having a partnership, without doing go-to-market execution together. They're currently trying to figure that one out with SNOW.

  • Shorter length of contracts, will continue to be a headwind under ASC 606 going into 2H'20, because late last year there was greater than average contract duration. ARR better indicator of actual business.

Muji and Software Stack Investing have discussed this recently, but I see the primary risk being the trend towards centralized cloud storage like Snowflake potentially eroding AYX's TAM and value proposition. It seems they have been slow to adapt to a changing industry landscape so when Dean said "there will be more product innovation for Alteryx in the next four quarters than there has been in the last 10 years", it seems more reactionary. I’m keeping my position as is though because I don't believe that the competitive environment has evolved significantly over the last couple months and AYX has suffered a permanent setback but it's too early to tell for sure.


This is another new position which I’ve written about here:

I understood that Peloton did benefit from economies of scale (fixed costs of content production and hard to scale hardware production) and built an aspirational brand but there was still less visibility than I would have liked. What's to stop an Apple from partnering with an Echelon for example, and brands are very hard to measure.

But what convinced me was the network effects. Users have their own profiles and compete against each other’s scores. More users = higher chance your friends are on PTON = better value proposition. Workout apps like Strava already have some of these metrics but it's hard to standardize if you aren't vertically integrated. After I ordered my Peloton, I immediately found myself searching for friends and wanting to add them to track our metrics together and race against each other’s scores, even convincing those without it to just try the app.

It is similar to a very expensive gaming console almost, you want to get the one your friends are on to play together, there’s a difference in content w/some having more exclusives creating a winner-take-most market with more popular brands being able to invest in better games, and people tend to have a fierce loyalty to one brand or another.

In addition, the whole gamification of fitness opens up a whole new market of people like me who were not frequent gym-goers before. Being able to track your metrics and see consistent progress is addicting, you know in gaming where people enjoy doing boring, repetitive tasks for hours on end to level up their characters, the same instant gratification applies here. Apple entering the market just further justifies the long-term potential of this trend.

My main concern is not the other hardware providers, but other content services like Apple Fitness+. This author of this article thinks Peloton will eventually be forced to split its hardware and content businesses.

"Other content competitors will open themselves to the Prime Bike, Soulcycle hardware, etc. and over time assemble an addressable market that is more fragmented but substantially larger. In the case of Android, this strategy has produced a fragmented user experience. But do we really believe that the integration of bike content and bike hardware will need to be as deep as that between phone OS and phone hardware?”

Since Echelon or Schwinn can’t compete on content because of Peloton’s economies of scale, they may want to open up their ecosystem to integrate with third-party content providers. If Apple invests enough money and their content is more or less the same as Peloton, then consumers buy a cheaper bike that works with it or whatever other content provider they subscribe to. If Peloton makes their content available on third-party bikes then they lose all their pricing power for their own bike but if they don't then people will go with the cheaper option. Again, all assuming that Apple eventually matches Peloton quality on the content which shouldn't be that hard as the content is all of the same type (instructor quality matters but Apple could outbid Peloton and steal talent). Apple also already is easier to scale by upselling into their install base, thus nullifying Peloton's network effects if they reach size. The one caveat here is that Peloton has the patented leaderboard and standard hardware = more accurate metrics. Leaderboard makes the biggest case for integrating hardware with software but not sure if that is a big enough differentiator.

Certainly, this is not going to happen in the near-term though, and will be a significant challenge considering the fact that Apple hasn’t been very successful in its other content-production endeavors.


Another new position, growing at a 57% CAGR since 2016 with 95% gross margins and adjusted EBIDTA margins of 39%. check out my thread on the basics:

GoodRx allows consumers to compare drug prices and get coupons, which improves medication adherence. They have amazing reviews on the App Store and an NPS of +90! They also partnered with healthcare professionals (NPS of +86) by integrating the codes directly in EHRs as well as having a telemedicine platform through their acquisition of Hey Doctor and offering price comparisons on third-party telehealth offerings too. They have 4.9M MAUs and they claim their primary moat is their scale and consumer focus, there's definitely some network effects at play here as they partner with more PBMs and pharmacies to get more data than competitors, hence offering the best value proposition to providers and consumers. They also have more users, so they have some leverage over smaller competitors and more PBMs and pharmacies will be inclined to partner to access their network.

Price transparency is a big component of creating a fairer healthcare system and will be only more important as co-pays increase and healthcare becomes more consumer directed. Controlling the consumer is very powerful and I don't see much reason why consumers shouldn't use it. Similar to Booking Holdings vs Expedia. Of course, the market is huge ($524B for prescriptions) and they enable the commoditization of telemedicine and drugs so they will benefit from their large userbase to gain a good chunk of the market as they get to own the front door (users go to GoodRx first to compare prices across pharmacies and telemedicine providers) then they can also have their own products on there.

Two main concerns:

1. Seems they are screwing over the pharmacies, who even lose money on some items, causing them to offer price match so they can avoid paying GoodRx its cut. GoodRx claims it improves foot traffic, thus there's a higher likelihood of purchasing higher-margin grocery items. The real reason is that they have contracts with PBMs but that hasn't stopped some from not accepting it regardless.

So doesn’t seem like the win-win-win they claim.

2. Don't see why others can't copy its model. Seems they mainly have network effects and branding on their side, so flywheel is more partners -> more discounts and increased likelihood of finding cheaper prices -> more users -> more partners. So there is some barriers to entry, but don't see why PBMs would not want to partner/share data with competitors too so they can maximize transactions (whole reason for them to share data is to access uninsured patients and to help insurers save money when GoodRx price is lower than price of co-pay). Regulatory changes are a big concern too.

But overall, they definitely have built a great brand. They could be Booking of this space. And pharmacies working around GoodRx probably won't get too out of hand before PBMs do something, best thing about that is that most pharmacies are a part of a national chain so standard policy. Owning the front-door of healthcare is very powerful, by commoditizing their markets, they can continue to extract rent and use that to fund proprietary services. We see this with their expansion in telehealth with their own HeyDoctor offering, others will be forced to partner with them to access their consumer pool.


Another healthcare company, it’s a significant position in the Durable Growth Portfolio. Will be releasing a deep-dive on it soon.


A micro-cap digital health company. Very risky investment but with easy multi-bagger potential. Will be releasing a quick write-up on it soon.


Sold Twilio because I wanted to concentrate in my higher conviction ideas. Like The Trade Desk, I’ve owned Twilio in the past. I re-invested because I think they could do pretty well if the telehealth trend plays out. They became HIPAA compliant in February, and they've seen more than a 100 percent increase in active healthcare customers using its video product. Epic, MDLive, ZocDoc, and Doctor on Demand all use TWLO. I think as traditional healthcare systems digitize and as more and more new startups build their own solutions then TWLO can do really well here even though this vertical is non-material to their growth right now. The unbundling or Zoom will be another trend in their favour as indicated in my Agora thesis. But these are both non-material segments of their revenue currently and I wanted to concentrate in pure-plays TDOC and API.



Category leader with excellent financials. Too expensive out of the gate at >55x LTM revenue.

My Thread:


Again, too expensive at >30x LTM revenue. Amwell had slightly better growth than Teladoc last quarter but a much worse gross and net margin profile and a worse competitive position after Teladoc's Livongo acquisition. Keeping an eye on payment parity (equal reimbursement for in-person and telemedicine visits) will be key to Amwell's success in my opinion, but I still don't think that health systems will be the harbingers of change. There is more incentive for employers to adopt Teladoc/Livongo and hope to reduce their insurance premiums.

My Thread:

To learn more about Saga Partners and our investment philosophy, check out our investor letters and fact sheet HERE and mention how you discovered us!

If you’re interested in learning about opportunities to invest with us, feel free to contact me at richard.chu@sagapartners.com

Connect with me on LinkedIn

Subscribe to my newsletter to get updates on new articles right in your inbox.