JLuo's Profile

Thanks for writing, Peter! The model sounds really interesting. One area where I’m curious to look more into is the monetization model – it looks like they still give away the SaaS dealership management product away for free and doesn’t charge a transaction fee on any purchases sold through that platform. That to me means the one revenue stream it has is from the DaFengChe platform (new car leasing), where it takes some sort of transaction fee and maybe a cut on the issued ABS? How much revenue and volume does it generate from that part of the business? I’m also curious as to the details of how new car dealerships (4S) price the initial loan payments in the first year – cars depreciate very quickly once driven off the lot, and the resale value even after 1 year must be at a significant discount from the original MSRP.

One other thing that I would love to learn more about is the threat of disintermediation – it sounds like they take credit data and use e-commerce distribution channels to find customers in a more cost effective way – is there a way for OEMs to short circuit this and go directly to Ant Financial or Alibaba or someone else? Ultimately, they’re a middleman that operates between the OEM (supplier), the used car dealerships, and the consumer – how sticky is that relationship, or are there incentives for disintermediation directly between the supplier and dealerships for example? I’m not clear on how the dealership network works in China.

Overall really interesting concept!

Thanks for writing! It’s interesting that in the past, Apple viewed its app ecosystem as a way to sell more hardware (better apps that better integrate with the phone). In the future, it looks like the software itself should comprise a larger piece of Apple’s business. Ultimately though, the success of the two are connected (as we even saw in our simulation game earlier this semester). Apple’s been trying to do all sorts of other things with respect to its mobile phone sales to “smooth” out their revenue, the clearest example being the iPhone upgrade program where customers pay monthly and get a new iPhone every year. Subscription models are powerful for many reasons, but the predictability of cash flows and reduction in revenue losses and churn are the ones analysts get most excited about. It’ll be interesting to see how the once-hardware selling business re-invents itself as a subscription and service business across the board, not just on apps but on the hardware as well.

Thanks for the post! We spend a lot of time talking about automation of more manual tasks (e.g., manufacturing), but probably don’t spend enough time talking about automation of higher value / higher knowledge tasks like consulting. I think the model, at least in the near and medium term, is for consultancies to transform more into tech-enabled service businesses rather than pure service businesses – in these types of businesses, the solves a bunch of the more manual tasks (e.g., excel data crunching), but isn’t able to operate on a higher level when it comes to insights or judgment. You rightfully mention that a bunch of the input piece of the equation has become more democratized by things like Euromonitor, but as we’ve seen in some of the other data cases, there is plenty of judgment that needs to even go into finding the right inputs to put into the decision algorithm in the first place. I’m also less pessimistic on the decrease in more junior staff – when I was hired into McKinsey after college, I felt much of what I added wasn’t just the manual excel manipulation, but also the thought it took to do the right analyses and put the right information together on a page to show the client – I would have loved for something to make those things more efficient and free up time to do real problem solving!

Thanks for writing about this! I think it’s a really interesting model. There are a couple of challenges I see arising from this. First, it turns the workplace into some sort of surveillance state where every employee action is tracked – beyond the privacy implications and cybersecurity risks if that data were to be stolen, I think constant monitoring could dissuade potential employees from joining or create a negative fear-based work environment, if implemented poorly. A second question I have is how the company bridges the gap between correlating behaviors / workplace environment factors with positive outcomes, and taking that one level further to draw causal links. Like the concern about the pay-for-performance value capture model you allude to, I think it must be incredibly difficult to really pinpoint the drivers of organizational value creation from a behavioral aspect.

On the value capture side, I think they could both sell a SaaS data platform model as well as corresponding consulting services. With their core data platform, they could aggregate best practices across organizations that they then deploy on consulting engagements to improve workforce productivity.

Thanks for writing about this Zach! As a huge NBA fan, I’ve been fascinated by Morey for a while now and his “hyper-efficient” brand of basketball. I think there are pretty meaningful benefits to this strategy that’s borne out in the data (and as you mention in your post). I do see risks with going so far with this strategy though, some of which are common pitfalls related to over-reliance on data. There could very well be a point at which only taking 3 point shots and shots inside the paint end up hurting offensively, as defenses adjust knowing that the Rockets will basically avoid all mid-range shots. Because analytics are done on past data related to historical defensive and offensive strategies, the resulting predictions will be most appropriate for those same playing styles – when playing styles change, the new “out of sample” data may no longer fit that trendline. Empirically (and we’ll have to see this year), the strategy hasn’t yielded the ultimate reward, which is an NBA championship. In fact, if you look at the Golden State Warriors this year, they are a hyper-efficient offense that actually rank in the top 10 of teams that take mid-range jump shots (what Morey would call the least efficient shots). It’ll be fascinating to see how this plays out in the playoffs.

Thanks for posting about this, Hans! It’s an interesting point you bring up about the NFL being a copycat league, and is very reminiscent of the TSG Hoffenheim case we had. In some ways though, I think you could argue that the Eagles can always stay 1-2 years ahead of everyone else. When I worked with a couple of sports teams on a player care / development tech consulting project in 2013, the Eagles were generally viewed as leaders in player fitness development and training because of their use of technologies like on-body sensors, on-field motion cameras, etc. It seems that this embracing of analytics is not just in how they manage games, but also how they manage athlete development and down-time between games as well. I guess another interesting question here would be how this translates to “economic performance” if you think of the team as a business. To that end, I think you can make the argument that winning a championship is highly accretive to franchise value and income for the coming years, which made these types of investments highly worthwhile.

Interesting concept! I really like the spirit of the model, but I do have a couple of concerns as well. First, in theory, tutors work well for students because they really learn what the right process for solving certain problems is, and acquire knowledge of the important concepts needed to solve those problems. In this model, I’d be concerned that students simply ask questions and get answers, without any real learning – I can see cases where people would for example just put up the various questions on their math homework and get straight answers, without really assimilating any of that information. A second concern for me (alluded to in the above comments as well) is answer quality – a voting / trust system can help solve some of this, but I’d imaging the questions being asked on these platforms generally warrant faster answers (e.g., for homework due the next day), and the rating system may not be able to provide feedback fast enough. A third concern is the incentive system for answerers – because of the prevalence of other, more recognized certifications (e.g., high school honor roll), I’m not sure a certificate or other types of certification would be that appealing to the “best” students, who’d rather put their energy into getting awards at their actual schools. I still think this is a great concept, but there are open questions as to the details.

Thanks for writing! I’m not too sympathetic to movie studios (e.g., Sony and Emoji Movie example from above) who complain about reviews diminishing demand – bad products shouldn’t “deserve” to make money just because there was money spent producing it. Movie studios also benefit plenty from positively reviewed films, which to me makes any stance against reviews somewhat hypocritical. On the crowdsourcing element – I’ve always been curious why users post reviews onto these sites (RT and IMDB). What specific incentives are being offered, other than the opportunity for people on the internet to view your opinions?

Thanks Jesse, this is a fascinating concept. I agree that the social / ethical implications are significant – beyond racial profiling, there are issues stemming from potentially false accusations (e.g., a person who is accused and widely broadcasted via Citizen video could become known as a criminal in the eyes of the public, even if he / she is later exonerated) and interference with ongoing criminal investigations or activities (e.g., in a hostage situation, police oftentimes ask media to stop recording because it may inform the hostage takers of the police’s strategy). I think these issues are very complicated without one-size-fits-all solutions, and if Citizen wants to act responsibly and have a human editorial staff that verifies every incident, it will significantly reduce its ability to scale. The value capture model is also somewhat unproven – they could charge users, but many people would believe public safety is a public good and the service should be free. I’m also not sure why law enforcement companies would pay for data, because they can already subpoena key pieces of evidence without charge.

This is a really interesting concept! It feels as much operational innovation as it is tech innovation, because of how they’ve managed to reduce developed times down to 3-6 weeks. I wonder what’s driving the reduction in time, because that sounds like a TOM case in the making if it proves to be sustainable. Another concern I’d have is how you ensure that results from voting will mirror in some way real-life demand on the shelves. For the platform to surface the best ideas, it needs to have scale of the voting audience and also engagement that is representative of the potential market demand for a product – managing that seems like it’ll be tricky, especially if there aren’t clear incentives for voters to participate. I suppose voters should be incented because of the possibility that a product they want will actually be made, but it may turn out that only the most fervent believers in a product vote, which would skew the vote share relative to the wallet share.

Thanks for writing Liza! I think the question you pose at the end is really interesting. I think the origins of the accredited investor laws in the US stem from the fact that most people don’t have the expertise or any of the other things you mentioned (time, confidence, conviction, patience) that are requirements for managing and evaluating risk effectively. As a result, the US put in place restrictions on non-qualified investing (e.g., I don’t think most people can participate in IPO pre-sales or private security issuances). Interestingly though, I don’t think these restrictions ultimately applied to individuals when it came to Lending Club (i.e., no laws that I know of that prohibit me from participating), so maybe it’s some sort of hesitation specific in the minds of US consumers that makes them reticent to participate in direct lending. For LC at least, a pivot to focus on institutional lenders seems to make sense given that’s where the scale can be attained.

Thanks for writing! I had a few thoughts on some of the challenges you’ve described in Thumbtack’s business model. It seems like Thumbtack operates as a horizontal marketplace, as opposed to companies like Uber that operate in a specific vertical (e.g., transportation). As a horizontal marketplace, I think you either need to believe that there are transferrable network effects across verticals (e.g., suppliers can serve multiple verticals), or that there are significant benefits to the initial scale you can reach on the platform such that you create those network effects more easily. It seems to me that Thumbtack is betting on the latter, where they can get a larger corpus of users onto the platform initially by offering a wide variety of services they can be matched with. I think horizontal marketplaces are more difficult to get right, but they can be incredibly valuable if it works out.

A second thought is on disintermediation. It sounds like the model is quite unique in that it only charges a one-time fee for matching – this probably really helps them with the disintermediation problem, but without the ability to monetize an ongoing relationship, the long-term economics may not work very well – they’ll have to keep supporting the relationship through their tech platform, but gain no fee off of it. Maybe they can price high enough such that the fee from the initial transaction is enough to cover initial and ongoing expenses, but higher initial fees will also reduce the propensity to adopt. I think it’s likely they’ll have to tweak this part of their revenue model going forward.

A last thought is on customer acquisition. It’s actually pretty remarkable that they have spent very little on growing their base of users – it sounds like most of it is organic growth, and VCs love that type of trajectory where you can reach 9-figure scale with little investment. This strikes me as a positive – there must be something compelling about the model that drives users to adopt even with little prodding, and is a positive signal for the long-term unit economics of the business.

Thanks for writing Kat! I didn’t even notice that we had written about the same company until now haha. I think ClassPass’s recent changes to its pricing model to the credit-based model makes pricing more dynamic and aligned with the costs of serving each customer – this is a change for the better from a long-term sustainability perspective. The issue though is that a lot of VCs may have invested money in the business assuming a certain size of market that no longer is true, since the value accrued to the consumer has now decreased (e.g., in the fixed-class model I could go to Flywheel 3 times a month, now I can only go once or twice because it costs too many credits). They’re probably likely to see some churn out of the business as a result of this, and it’s an open question of how large the sustainable mass of users is under this new model.

When it comes to live classes, I’m not sure what their competitive advantage is over something like Peloton, who actually provides the equipment and has their own owned content. For ClassPass Live, it seems like they still have to depend on studio instructors, and have really low barriers to switching (in Peloton’s case, the cost of the equipment itself is a barrier to switching). I’m not sure they should be branching into this adjacent model, chasing what’s trendy, when they could instead shore up their business model and insulate themselves against challengers. I think the competitive threat from companies like MindBody is very real, especially since MB is ingrained in most studios already and has a consumer-facing app.

Thanks for writing about this, Eliza! I think the Opendoor model is really interesting, and a few of us actually visited the company on Westrek last year. Like you say, there appears to be pretty meaningful value for sellers – the speed and certainty of closing on a house sale, for many people, is well worth the added commission cost to Opendoor. Based on notes I wrote down during our meeting, it sounds like they make 40% gross margin per transaction, and they focus on houses in the $150-500K value range.

There are, however, several key risks to this business model. The first is the inventory risk that others have alluded to – in the event that Opendoor cannot find a liquid market to sell the house they’ve just purchased, they incur inventory carrying costs, which include maintenance and probably most significantly financing cost. This financing cost is the second risk that Opendoor incurs, namely that they lever up meaningfully to purchase these homes and are left paying the interest on that debt. Extended holding periods / reduced turns certainly increases this holding cost, but so does potential changes in the rate environment – with interest rates at all time lows and only expected to go up, Opendoor’s cost of holding inventory should also go up (and compress their gross margins). A third risk is macro housing valuation risk. By holding onto the inventory Opendoor holds all the risk associated with home values, both on the upside and the downside. Depending on the macro environment, Opendoor’s leverage profile coupled with rising rates and decreasing home prices could spell significant trouble for the company. To manage all of these risks, they attempt to use data algorithms that compute value-at-risk for the home portfolio and build in some margin of safety. In theory, this should help them manage risk, but one thing to watch out for is that all the CDS algorithms in 2008 were supposed to manage risk as well, and we all know how that turned out.

The cybersecurity space is certainly a fascinating space, but I think the winners are losers here are going to be hard to call. Already, the space for enterprise-level cybersecurity solutions is relatively competitive. The traditional security firms (e.g., Symantec, Intel McAfee) have begun to seriously refresh their offerings, large enterprise software conglomerates have instituted new initiatives to stem share loss (e.g., Oracle), there is a host of other billion dollar public next-gen security companies (e.g., Palo Alto Networks, Check Point, Fortinet) that have taken significant strides in the market, and a tremendous number of highly valued private startups (e.g., Okta, Ping) trying to carve out a space in the cybersecurity market. Ultimately, cybersecurity is such a fragmented and fast-moving space that the winners of yesterday could easily become the losers of tomorrow. Even FireEye has had its fair share of challenges. In 2014, coming off the highs of their acquisition of Mandiant (great decision, which you write about), the stock price hit over $85 per share. Today, the stock is at $15 per share as it has faced growth challenges and product competition from the likes of Palo Alto and Check Point. In this space, it’s always faster, better, cheaper wins, and that places huge R&D pressure on all of these companies.

Interesting piece! It seems like Unity is well set-up for continued success as a third party video game engine, especially for mobile. However, I did have a couple of thoughts on potential risks for the business down the line.

First, the video game market is growing more and more concentrated with the largest game publishers globally continuing to gain share. Part of this is due to these publishers’ efforts to increase the amount of time players can spend on the best “AAA” games (e.g., Call of Duty, FIFA), but the end result is that user time spent seems to trend more and more toward fewer titles. With these fewer titles concentrated in the hands of large publishers who have every incentive to keep development in house using proprietary engines (e.g., EA’s FrostBite engine), is there a potential cap to the size of Unity’s addressable market? As these AAA publishers move more of their games to mobile, is their a larger competitive risk to Unity?

Second, while mobile is the fastest growing piece of the gaming market globally, dis-aggregating that growth shows that the majority of growth has come from the Chinese gaming market, which is dominated by local players like Tencent. The chinese market is also notoriously protectionist – foreign publishers actually can’t distribute games in China without a partnership with a local company like Tencent or Netease. Can Unity effectively tap into the developer ecosystem in China, and can it navigate the regulatory hurdles that might exist?

Unlike the writer of the previous comment, I think it might be a little too early to write the conclusion to Snap’s story. As you rightfully point out in your piece, there are still significant assets that Snap has, most notably its large corpus of users in the younger demographic brackets. In fact, according to a recent Piper Jaffray survey, Snap has a 47% to 24% lead on Instagram when teens were asked for their preferred social media platform. Ultimately, these teens are the ones who will age into the high-spending 18-34 year old category, and the huge preference lead Snap has over Instagram and Facebook here is still quite valuable.

The challenges that the original poster and the first commenter pose are very much real – Snap has done a bad job of conveying its value proposition to the non-teen non-young demographic, and as a result its user growth has shown significant signs of slowing down. It remains to be seen whether the product re-design will make Snap friendlier to older users and whether it will allow Snap to enhance its monetization strategy, but with such a large corpus of active young users, the story may be far from finished.