Really interesting article! To your second question, although I think 3D printing could be extremely useful in the jewelry industry I am nervous about its adoption for production, even in mid-tier category. For jewelry and other luxury goods, the perception of the jewelry is sometimes more important than its actual aesthetics, and having something that’s “handmade” will always be valuable, even with flaws. For example with nice dress shoes, having imperfections is a sign of quality because it means they were not machine made. I’m just not sure machine printed jewelry would ever be able to overcome this stigma at the high end, and at the low end it probably doesn’t make economic production sense. That being said, I do think it can replace wax molds. The wax process sounds unnecessarily complicated and it sounds like 3D printing can provide a more capable, efficient process while still allowing for a personalized professional design.
Super interesting article. I love what Airbnb is doing with machine learning. As a consumer too many options is overwhelming and if they can filter down to the 4-5 I’m most likely to select than it significantly improves my experience. One hesitation I have is with how accurate the machine learning actually is. When booking an Airbnb I tend to have different criteria every time depending on who I’m traveling with, if its for work or personal, or where the destination is. If the machine learning systems can’t pick up these differences I worry that it would be keeping certain options from me just because I haven’t selected something similar in the past. It will be interesting to see how Airbnb solves this problem.
To answer your question I think AI and machine learning will be excellent tools to reduce discrimination in the industry. The discrimination found still stems from human users and while it may have been exacerbated by the ML, I think that’s something that can be easily solved for once it is known. Even in other instances where ML has been found to be discriminatory, it still tends to fare better than when humans are left to their own devices. I think Airbnb can actually use ML programs to find hosts who are guilty of this behavior and remove them from the site. Additionally, as the programs get more advanced, AI will be able to find and book places independently, completely removing any identifying characteristics from the process (e.g. race, gender)
Very interesting article! I think what Nike is doing in terms of R&D is very cool, although I do question its ability to scale to the point where it actually makes production more efficient. One thought would be to just have 3D printers in every store which would enable you to have zero inventory, but the technology is definitely not there yet.
To answer your question, I wouldn’t expand the footwear experimentation to other sports until it was perfected, but I would expand to other products in the meantime. Experimenting with one of every product type but for only one sport. Focusing their resources on single product areas would be significantly more efficient and they shouldn’t risk moving resources from their major business lines without knowing for certain the technology was ready. Nike would be risking a lot in terms of quality perception by rolling this out too quickly, and I don’t think its extremely urgent as competitors are still figuring out the technology as well. I also see the biggest benefits in production and not necessarily in the consumer experience, as such the risk is more financial than with consumer demand.
Super interesting article! What Betabrand is doing is very cool. One concern I have is around the true predictability of consumer feedback in fashion. I feel as though new fashion trends usually push the envelope a bit, and that consumers have an initial distaste but then get caught up in the fad as it gains popularity. I would worry that while Betabrand is protecting itself from bad product decisions, they are also missing out on some hits due to this process. For example for movies, regular consumer reactions to scripts/pitches usually have little correlation to how successful the content ends up being, and some of the biggest hits actually got really bad initial feedback.
To answer your question about competition, in order to survive I think Betabrand needs to quickly build its brand image with its core customers. This involves raising awareness, getting lucky with hit items for the next few seasons, and providing more distribution through brick and mortar stores. There’s very few barriers to entry for this business model and Zara and others will definitely incorporate it if they see its success. The only way to protect Betabrand is to build loyalty with both customers and their feedback providers.
Really interesting article! I definitely agree that these D2C startups are a big threat not just because they’re more innovative, but I think they have a natural brand advantage, where millennial consumers are willing to pay more for a “startup” product then one produced by a conglomerate (e.g. Away). This unfortunately erodes Unilever’s scale advantage.
To answer your question, I actually don’t think Unilever can identify promising brands early. In general, corporate VC/incubator arms have lower success rates than a regular VC. One reason for this is a corporate program tends to inhibit the exact innovation they are trying to produce. As a result corporate VCs are generally seen as less attractive investors to a startup, so popular companies with multiple funding options usually go with other investors (excluding top tech cos.). I think if Unilever had the ability to identify these companies, they would also have the ability to produce similar ideas in house.
I actually agree with Unilever’s strategy of acquiring companies in the ~$500M range. At that size they have already established the brand in the eyes of consumers and are large enough to operate fairly independently once acquired. Additionally they are big enough where they can take full advantage of Unilever’s scale in distribution and production. This also provides the influx of outside innovation they need, but at a scale where it can actually be infused within the organization. In combination I would source ideas straight from consumers by holding case competitions or having open submission contests. This brings innovation into the company at the ideation stage where Unilever can take advantage of their own production process. I believe this provides the same upside and risk as a VC/incubator but gives them significantly more control and allows them to lean into their own core capabilities.
Great article! Very interesting points. Two things come to mind when I think about how Comcast’s new product offerings and machine learning development will be affected by their entrenched position as a cable provider and ability to compete with Google and Amazon. One is their status as a high speed internet provider and two is the limited footprint in which they are allowed to operate.
Through Xfinity Comcast is one of the country’s biggest internet provider and I think this gives them a huge advantage. To answer your brand question, the Xfinity brand has been able to distance itself from the low end cable services and is known more for their high end tech solutions which should help them as they expand. Additionally, even as cord cutting has increased, Comcast’s high speed business has grown significantly which helps maintain their footprint and also gives them even better consumer data on which to build machine learning tools. The data from internet services is much more complete and diverse than data from companies where users have to self select into specific services. Controlling a user’s internet also allows them to build better smart home products that integrate more seamlessly than a 3rd party would be able to do. Also, since they are usually the best, if not only, high speed internet provider in a given region, consumers are almost forced to interact with them, giving them an opportunity to bundle smart home products with internet services.
One concern I would have is how Comcast’s internet and cable footprint is set by regulatory agencies and thus they would have difficulty growing it. As smart homes become more connected to communication, the inability to reach everyone in the country will affect how many services Comcast can offer. For this reason I think their best course of action would be to partner with a tech company also competing with Google and Amazon (Microsoft?) and together their data capabilities would allows for better products.