• Alumni

Activity Feed

On November 30, 2017, BT commented on Collaborative Robots as a hedge against Isolationism :

Really interesting article, about a topic and technology I’ve never heard of. The first question I have is how effective are these robots compared to humans? Are they more / less / roughly equivalent? Another question I have is how quickly these robots can be deployed and ramped up? Is it faster than it would take for a human to ramp up? Are there increased safety concerns with having humans and cobot arms working side by side? Based on your comment above that various nations have entered into agreements to work with these cobots, my biggest question overall is how deeply these are being used today? Do they simply have a prototype with a contract to deliver the product in the future, or are these currently in use all over the world?

Overall, there seem to be significant cost reductions in terms lower shipping costs by manufacturing domestically with cobots through low fixed costs, even accounting for the $40,000 PP&E cost which can be amortized over 5 years.

On November 30, 2017, BT commented on Amazon – In Prime position for the last mile :

Really interesting article Max! You’ve done a good job capturing many facets of Amazon’s last mile opportunities and also dove into a potential LaaS (Logistics as a Service) business unit for Amazon.

1. Many have spoken about Amazon predicting and shipping future orders as you mention above, but I see a problem with this practice as it relates to channel stuffing. Right now, people have a positive, fuzzy image of many tech firms, but this practice opens up the possibility of Amazon probabilistically sending people items that they know a percentage of which will be returned but a percentage of which will be kept. There may be a problem with the latter because there are few signals at that point of how the customer interacted with the delivered product: did they enjoy it and keep it? Were they too lazy to return it? Did they intend to return it but forget about it? The latter 2 events could be construed as unethical practices if Amazon is even slightly perceived as forcing unwanted products on users. Furthermore, this might lead to fines and legal action, but also hurt Amazon’s trust with its customers.

2. To your point about Amazon’s potential LaaS operation, I’m not sold that this would be a success. It is a similar origin to AWS , which arose out of Amazon’s excess server and storage capacity (which can easily be scaled as demand for AWS increases). I wonder whether the same excess capacity exists in Amazon’s warehouses, and I question whether it can be as easily scaled as server storage. This will likely be a much lower margin business than AWS if it is profitable at all.

This LaaS business also brings Amazon closer to anti-trust concerns. You might find this video (though some of it is alarmist and ungrounded):

This video makes a strong argument that many tech firms have used a sweet, fuzzy public image to gain anti-competitive power in the overall consumer landscape and that other industries (like “big bad telecom”) are held to a higher anti-competitive standard while big tech gets off easy for now. After all, is the difference really that big in AT&T’s acquisition of Time Warner than in Google owning Youtube? There is a possibility that big tech’s image changes over the next decade where tech firms are held under a stricter anti-trust standard.

On November 29, 2017, BT commented on The next Google – should iRobot sell you a $0 Roomba? :

Nice article Amar. This is very thought provoking and brings up many fascinating ideas. However, in most cases, I feel very strongly that iRobot’s strategy is a poor one.

As Anusha mentioned, privacy is a huge issue for any data company. If Roomba were to start mapping and selling the data of a user’s home, it would then sit on very private data, considering that end users consider their home to be one of the most private aspects of their lives. One might wonder how security breaches would affect Roomba data. If the data were to be hacked, would it be possible for a burglar to see your home layout and illegally enter a Roomba customer’s home?

When we think about who will be the customers for this potential Roomba data, I think of major tech companies such as Amazon (who will in the future sense that you’re out of toilet paper and order you more) and advertisers. The question for each is how do you classify this data in meaningful ways that can be actionable but not overwhelming, as well as secure. In internet advertising, most banner ad retargeting is done by cookie’ing a user and anonymizing their user data. There are also many compliance issues with collecting personally identifiable information (PII) such as name in advertising and analytics. In a sense, an ad server is sending a ‘check out your abandoned cart’ display ad to USER123456 instead of Amar. The inherent problem I see with anonymizing and depersonalizing this data (and as a result protecting the privacy of the user’s data) in this case is that the dream for a smart home is to have a more intimate solution like Alexa, calling you by name and making smart recommendations, not to have a Roomba rolling around like an internet pop-up ad. In order for Roomba to even think about creating a significant personal relationship with a customer, PII compliance rules will likely need to be re-evaluated (which may take years) and iRobot will need to develop significant core competencies in information security for IOT applications an inherently insecure sphere due to the youth of the space. Personally, the core competencies for security and privacy are so high in the case of mapping a home that I would not entrust them to iRobot.

Finally, there is an interesting idea of pricing in that iRobot would give away Roombas in order to collect data on users. This is frankly awful positioning because it positions Roomba almost as a surveillance tool that is always watching the user only to collect data on the customer and make money. It is likely that by using this strategy Roomba would acquire low-value customers who do not have a lot of data to protect or who are careless about their privacy. In addition, looking at the iRobot home page for pricing, the Roomba is priced between $250 and $750. Roomba can easily monetize that user data over a few years, but the question for customers that will happen (and is happening across many industries) is how much do I value my privacy, security, and data? $250 – $750 is probably too low.

Nice article!

You bring up an interesting first question of “What steps can Walmart take to corroborate supplier’s independent reports of product sustainability?” For a company whose supply chain makes up 80% of the greenhouse emissions associated with the U.S retail and consumer goods sector, it is important for Walmart to start taking steps to change its global footprint for the sake of keeping a sustainable world and also maintaining its public image. However, as it does so, it needs to be careful to balance driving value for shareholders (who will be myopic and focusing on 2-3 year horizons instead of the 2030 vision Walmart has set) and keeping costs manageable with reducing its global footprint.

To this end, Walmart needs to be very efficient as it audits its 100,000 suppliers. Ideally, there would be audits for each location multiple times per year, but I wonder whether sending 500 auditors to 200 global locations for once-a-year audit would even be a manageable cost, or whether a yearly audit would even be a viable incentive to dissuade bad behavior from the suppliers. Similar to Ikea with its IWAY policy, Walmart would be best off by implementing a harsh penalty of contract termination for suppliers who do not buy in or underperform on their scorecards. In the long run as more IOT applications arise, Walmart may be able to deploy internet-connected sensors in supplier facilities that would monitor certain aspects of the suppliers compliance with Walmart’s scorecard.

Elli, you bring up many interesting strategies that JLR is using to mitigate increasing costs due to Brexit, including moving component manufacturing and suppliers to the UK, hiring skilled workers from the EU while it still can, and partnering with Chinese manufacturing companies. I wonder is it possible to relocate JLR headquarters and incorporation to a different country and avoid the upcoming UK Brexit penalties you mentioned? I am not certain what the costs would be to reincorporate in a different country, but for 600,000 cars to lose £2,372 in margin would be a roughly £1.4 billion dollar loss every year. If the costs to reincorporate were less than say 10 years of losses, then it might be in the company’s best interest to relocate.

In response to your thought “How should JLR balance its response to the ambiguous external trade environment with the requirement to increase investment in new technologies,” the big concern I see with JLR pursuing self-driving cars and automation of manufacturing plants as you mentioned is the timeline of Brexit. It is unlikely that self-driving cars will be ready in time to differentiate JLR, and I worry that automation of manufacturing plants (including delays, budget overruns, etc) will be completed in time. It appears that JLR will be hit hard regardless and that many of the strategies you outlined above are the best immediate steps they can take to reduce their exposure on supply chain costs.

On November 29, 2017, BT commented on Disney n Chill :

Great article Mehdi!

You bring up an interesting idea of day-and-date releases for movies, which was attempted as you noted for Tower Heist and a few other projects including one by Steven Soderbergh. You assert that film distributors may be reluctant to release a movie for fear of losing revenue when Disney can stream it directly through its own channel. However, there is likely a greater concern of cannibalizing in-theater movie ticket sales (which typically send back 90-95% of box office ticket revenue through to the studio over the first 2 weeks) by releasing a movie that can be simultaneously viewed at a cheaper price than buying a movie ticket. Consider this: a family of 4 goes to a 3D movie and pays 4 x $15 per ticket = $60. If the movie were to be released simultaneously, the family could just as easily pay $10 for a rental fee split amongst 4 people and while the family has appreciated the movie all the same, Disney loses $50. Overall, for most of Disney’s theater content, it is likely not in Disney’s best interest to release a movie in theaters and online at the same day. What would make more sense for Disney is to control more of the distribution of its product at different stages of a movie’s lifecycle in market. For instance, a movie might be in theaters for fewer weeks (as earlier weeks send a higher percentage of box office back to the studio than later weeks), may then go to Disney’s OTT service in a buy-only mode, and may then appear for rent.

One additional thing to consider. As Disney contemplates moving into OTT streaming, it will likely need to re-evaluate the content that it creates for the small screen. As you noted, technology is threatening the movie industry, but there are inherent differences in how a movie can be experienced in a theater vs television vs laptop vs smart phone. I would expect Disney to keep generating spectacle-driven entertainment for a theater which can capture a lot of margin for them and perhaps experiment with more Netflix-like binge television shows and even perhaps mobile content (which no one seems to have a good handle on yet).