• Alumni

Activity Feed

On November 23, 2015, Salem commented on Tinder: You’re doing it wrong :

I’m honestly not surprised to see this – several of their competitors have been doing this already (The League, for example) and giving users a secret “flakiness” score that impacts your future matches. Also, I’m not sure if they are doing it anyway, but there is a massive opportunity for mining the text data on the Tinder platform. There is a ton of information on user habits (including what activities they like to do, where they like to go for drinks, etc.) that Tinder should capitalize on before they start to lose their users to other platforms. However, the silver lining about multi-homing is that it’s actually an asset to each platform. Tinder, Hinge, Bumble, etc. only have value when users are single (or cheaters) – the paradox of choice implies that the more perceived options someone has, the less likely they are to commit – thus, it’s better for Tinder if their users are simultaneously on every other platform because it increases their likelihood of staying single and continuing to be a Tinder user.

On November 23, 2015, Salem commented on Gmail: ensuring a spam-free inbox with Machine Learning :

Cool post and I especially like the bit about the challenges, particularly the impact of false positives and false negatives. I think false negatives (and the low incidence in Gmail), as you indicated initially, are what elevate Gmail’s product above others. Enough false negatives, i.e. spam emails that end up in your inbox, could eventually drive a user away. What worries me much more are false positives – how often do users actually go into their spam folder and check to see if there are any relevant emails? Based on my own anecdotal evidence, that action is very rare. However, one case of an important email inadvertently making it into the spam folder is likely to be far more damaging to the Gmail product than many false negatives. As a result, Gmail should be very cautious with their algorithm and give a very high priority to data around “spam” emails that a user marks as “not spam”.

On November 23, 2015, Salem commented on “Also recommended for Sarah…” – Thanks, Netflix. :

Interesting post – I too had primarily considered Netflix’s use of data surrounding the recommendation engine, so it’s cool to see how else they use it. I worry a bit about two things. First, how defensible is their position? Many new and traditional players are getting into the SVOD space and it’s impossible to compete on licensed content alone. Their two sources of differentiation are their use of data and their original content. Other players can compete on the data front: HBO Go can gather just as much data from their users and curate the content. That leaves only original content. I think using data is a great way to figure out which shows Netflix should go with based on what their existing users already enjoy, but I worry that this will be at the expense of truly innovative artistic productions.

On October 28, 2015, Salem commented on How the Crowd Saved LEGO :

In response to both posts above, I think Elizabeth is right to note that there is likely demand for custom-crafted LEGO masterpieces that don’t necessarily make it into the development pipeline. I also agree that this would hurt sales of physical LEGO products. However, what is stopping a third-party platform from launching such a service? I’m not sure how they would necessarily monetize it, but it would still hurt LEGO’s ability to sell bricks. For this reason, it may make sense for LEGO to do it as a defensive play. Maybe they could charge users some nominal fee to unlock the instructions.

Alternatively, they could add user profiles that log all the LEGO sets you’ve purchased (and thus all the LEGO pieces you own, assuming they aren’t lost under dressers and impaled in bare feet). LEGO could then show custom designs that you are capable of building with your existing pieces and strategically show awesome designs that would require you to purchase a few new parts or a whole new set (FarmVille style).

On October 28, 2015, Salem commented on Waze: Leveraging the Crowd to Find Your Way :

Awesome post! I’m curious if Waze extracts more value from the direct or indirect crowdsourcing. The indirect crowdsourcing isn’t all that different from what Google Maps does for traffic, so I would assume the direct crowdsourcing provides the differentiation. That said, this direct crowdsourcing typically requires an active driver to be playing around on their phone. Waze “attempts” to prevent this by launching a popup and requiring the user to confirm they are a passenger and not a driver, yet they benefit each time a driver breaks the law and adds traffic and accident info. I imagine it’s not difficult to determine how often a Waze user is involved in an accident. I wonder what, if anything, Google could and should do with this information and if there is a way to make Waze use safer.

On October 28, 2015, Salem commented on Wikipedia – Who is the authority for the Truth? :

Great post! I particularly like the discussion at the end about value capture. I imagine that if Wikipedia tried to monetize the site, the reputation for information integrity would be at risk. Wikipedia was not always considered a reliable source – in the 2000’s, I remember it being banned as a resource for research papers in my school. I think the deliberate decision to avoid advertising/sponsorship enabled it to build this reputation over time — I can’t imagine trusting an encyclopedia page littered with advertisements. To that point, considering nearly anyone can be a contributor, what checks are in place to stop corporations from slyly promoting their brands on various relevant wikipedia pages?

On October 5, 2015, Salem commented on YouTube 1, everyone else 0 :

Part of the beauty of having a billion users is that there is bound to be some diversity — some users want a free experience and would never consider paying for a subscription service (which may very well be why they consume music on YouTube rather than one of the paid platforms), while there is a decent sized chunk that would love an ad-free product (and those other “rumored” features of background listening & offline mode). Regardless of which camp the content consumers self-select into, the indirect network effects between viewers and creators should remain just as strong as long as the “free” users don’t feel like their experience is deteriorating. For this reason, I think it’s really important to have a thoughtful promotion engine for the paid service so you don’t alienate or annoy your free users who will continue to make up the vast majority of YouTube’s user base.

On October 4, 2015, Salem commented on Strava: Social Fitness Veteran :

Great post – I’m an avid Strava user and I’m constantly surprised how much it permeates both my fitness and social lives. I’ve seen dozens of people download the app in social settings when the conversation turns to cycling or running. As a casual runner and cyclist that lives in very fit neighborhoods and is therefore never really near KOM/QOM status, I extract most of the value from Kudos, comments, and camaraderie with my friends in the app as well as PRs. I think it’s really interesting how they have been able to create value for both casual and highly competitive athletes, however, they really only seem poised to capture value from the power (i.e. highly competitive) users and I don’t think the network effects are strong enough between the two groups to effectively scale the paid tier simply by growing the overall user base.

Great post & cool company. I’m curious how representative this data is of each overall patient population, particularly for chronic diseases given the online platform inherently caters to a tech savvy crowd. Also, because of the strong network effects, certain conditions and symptoms may be overly represented and not indicative of the broader disease profile. It would be useful to both the patients and the pharmaceutical companies to clearly show who and what might be missing from the data.

On September 14, 2015, Salem commented on Venmo: Rapidly Replacing Cash in the Pockets of Millenials :

This was a great read. I wholeheartedly agree about the great UX within the app, although I seem to hear one common complaint that results from the process. When one person picks up a large tab and is venmoed by everyone else, they stand to earn all of the resulting credit card points. Also, as many digital venmo transactions are directly accompanied by a credit card transaction, I’m curious if there is a partnership opportunity between venmo and a credit card company, in which, for example, Amex will give bonus rewards points to a user if they pick up a big tab and use venmo versus conceding to their Visa-carrying friend. Even better, if they could distribute the rewards points among all people within the transaction (even non-Amex users) it could be a good channel for customer acquisition.

On September 14, 2015, Salem commented on Digital Matchmakers Win! :

Great post & I wholeheartedly agree that Tinder serves as a “gateway drug”, but with competitors like Hinge, Bumble, Coffee Meets Bagel, and The League curating your matches while retaining swipeability, I wonder if Tinder will instead funnel users to the competition.

On September 14, 2015, Salem commented on Waze: Redefining Carpooling :

I love the concept, and with Waze’s competencies in growth, I think they can definitely make this successful. I do have a concern about safety and privacy – users are generally ok with giving away endless data about their actions and movements to Google, but generally with the assumption that it will only be exploited by Google’s advertising and internal apps. This is the first time I’m aware of that this data is being shared with another user (the carpoolers you match with) and I think this has the potential to raise security and privacy concerns among Google & Waze’s users. They will have to be very careful about what information is shared across the two sides of the marketplace.