A Case Study in the Great Unbundling: Lunchclub

As I have noted in previous posts, it’s my opinion that the most important artifact of your university experience is your network. In this spirit, I have tried various methods of articially improving my network outside of friends and extracurriculars, including cold emails, LinkedIn connections and the very occasional Twitter/Messenger DMs. However, these methods seemed, well, artificial. The connections made were somewhat shallow and trivial and had little basis to build more meaningful relationships. I wasn’t very happy with my progress until I came across Lunchclub. With a friend’s referral, I stepped into the world of AI networking and was pleasantly surprised.

Lunchclub is an excellent case study on some broader trends I have noticed in technology. As always, I want to take a stab at suggesting some product improvements as well business strategy while also exploring broader trends in the technology space.

What is Lunchclub?

Simply put, Lunchclub is a professional matchmaker. Using proprietary artificial intelligence, it uses your goals, interests and other data points to match you with other professionals. It has been lauded for its high quality and relevance of matches, with matches turning into founding teams, investments, close friends and more.

The founding team is comprised of deep technical and operational talent:

• Vladimir Novakovski: previously worked at Quora, Addepar and other financial companies. Most notably, Vladimir started Quora’s ML team, which is known to be an industry bar-raiser.
• Haley Leibson: worked at venture funds and launched the largest female founder network
• Scott Wu: previously worked at Addepar and was ranked as the top competitive programmer in the world in 2014

The team launched their product in a select few cities and quickly started to notice traction. What was especially interesting were some of the metrics in their pitch-deck to a16z for their initial seed round:

• 19% of users at the time were referred by someone else: this is a clear indication of Lunchclub’s ability to capture the network effect
• 42% of users would be very disappointed if Lunchclub went away: this metric indicates that Lunchclub has achieved notable product/market fit
• On average, users have 10 meetings/year: this piqued my interest because it’s a good indicator of the retention of the platform. For a new product, this is a very respectable achievement.

Lunchclub really took off with the COVID-19 pandemic. Users flooded the platform as people tried to find alternatives to the spontaneous interactions that were otherwise present in the world. Coupled with the advances in video communication like Zoom, Lunchclub was poised for a great acceleration in growth. Over the quarantine period, the startup handled a 10x increase in meetings on their platform, with tens of thousands of meetings occuring every week. To top off their incredible success, they raised $24.2Mata$100M valuation in September 2020. Lunchclub is surely taking off!

Key Trends

The growth of Lunchclub exemplifies two trends in technology that have recently come to the limelight

• Unbundling of social media
• Virality of exclusivity

Unbundling of social media

2005-2018 was the golden moment of massive social media companies. Facebook, Instagram and Snapchat had a massive amount of users and LinkedIn had a near-monopoly on professional social networking. However, there has been a noticeable change with new social media companies.

Increasingly, we are seeing social media that has become more specialized and are scaling vertically rather than horizontally. These new social media companies are focusing on specific domains and use cases that mammoth companies have not addressed. Some examples include:

• Nextdoor: a social media for neighborhoods. Creating WhatsApp or Facebook groups for neighborhoods often leads to a jumbled mess. Nextdoor aims to simplify social networking for neighbours and has grown considerably over the pandemic.
• Cocoon: a social media for close family and friends. It has always struck me as odd that 1 platform can work simulatenously to develop close relationships with friends and family while also allowing you to discover new friends. In some ways, platforms like Instagram and Snapchat seem to have a tough time choosing between the two. Cocoon chose the latter and has achieved considerable success.
• Edudo: a nascent social learning platform that acts like TikTok but for education.

Lunchclub fits nicely into this trend. As Reid Hoffman mentions in his book Blitzscaling, LinkedIn chose to go down the path of appeasing multiple users: regular jobseekers, recruiters, companies and more. This has led to a platform that has some major issues in particular use cases. Networking is a great example: meeting new people on LinkedIn feels uncomfortable and unnatural. A lack of focus in any one use case has created a growing distaste with LinkedIn, especially among younger users.

Lunchclub is attempting to unbundle the networking aspect of life from LinkedIn. In fact, the founders have made their ultimate goal very clear: everyone on Earth should use Lunchclub as a means of networking. This hyperfocus on one usecase is emblematic of the general unbundling of social media.

Projecting to 5-10 years from now, we may see cycles of bundling and unbundling. When people feel overwhelmed with the variety of social media, large platforms may have a resurgence. When these platforms fail to meet the demands of particular subsets of users, there will be unbundling. However, this is completely predicated on the general public’s perception of social media; increasingly, it is turning negative. I believe that if this negative perception continues, hyperfocused social media may take the reigns from the giants.

Virality of exclusivity

This is more of a general truth than a trend, but Lunchclub is a perfect example of how exclusive access can paradoxically lead to massive user growth. One thing that stood out in Lunchclub’s founding story was the massive waiting list for their product; this very fact encouraged others to join the bandwagon as people hate to feel like they are missing out.

There are several instances where exclusivity acts as an accelerant of growth rather than a retardant:

• Supreme drops: these drops have become well publicized for the amount of hype it generates. In high school, some of my friends would try to buy Supreme products, only to be disappointed when they sold out within a minute. Exclusive access to Supreme leads to percieved greater social status, which in turn pulls more customers to try Supreme
• Clubhouse: This social media has faced significant growth, partially because of the same tactic that Lunchclub uses. It’s waitlist has generated interest, which again encourages others to join the bandwagon.

What I have realized from this tactic is that exclusivity is acting as a signal. Specifically, it signals to users that their is massive value hidden behind the app, so much so that they need to control access via a waitlist. Whether this is signal or noise is up to users to decide, but no one can argue against its effectiveness in generating hype around products.

Product Recommendations

Users

The first thing we must consider are the users on Lunchclub. Here’s my categorization for the main users on the platform:

I chose to do a breakdown along career stages given Lunchclub’s initial focus on professional networking.

I would like to focus in on 1 user group that satisfies the following criteria:

• Growing segment of total user group
• Vastly underserved by existing solutions
• Can impart long-term value to the platform

With these criteria in mind, I will focus in on students. During my time on Lunchclub, I have noticed more and more students using the platform, indicating that it is a growing segment of the total user group. Furthermore, the Gen Z/Millenial user group is quite underserved with networking solutions; the best alternative is LinkedIn but it has a dying reputation among this group due to the plethora of cringy and boastful posts on the platform. I also believe that if Lunchclub could deliver some outstanding solutions to students, you could create a lock-in effect, where you create students who are users for a much longer period of time. Another reason for choosing students as a user group to focus on is the potential for virality: students are much more likely to share new platforms and encourage others to try Lunchclub out. In short, there is a lot of upside potential if Lunchclub can develop the right solution for students.

Painpoints

To consider the major painpoints of students, let’s first understand why anyone uses Lunchclub in the first place. Here’s my breakdown for the use cases of the platform:

From these use cases, we can derive the following painpoints for students:

1. Providing value to the other person: Networking is a 2-sided relationship, where both parties provide value for each other. For students, it often feels like a 1-sided relationship where the student receives value and the other person gives value. It’s not very obvious how students could provide value to the other person given our inexperience.
2. Building lasting relationships: As I mentioned earlier in this post, the goal is to build meaningful relationships that last for a long time. Although Lunchclub gives the initial spark, it’s up to the individuals that maintain the fire; oftentimes, the fire burns out due to a lack of followup, time and more.
3. Meeting people within a narrow range: As a student, I want to meet very specific people in order to meet my goals. For example, if I am applying to a company for an internship, I would love to get to know someone who already works at the company. With the current Lunchclub algorithm, it is hard to update our preferences for who we want to meet because our interaction with the platform is low (minimum once a week). TikTok is a great example of a company that uses the plethora of interactions to develop niche spaces for your tastes, but that totally depends on the amount of interaction you have with the TikTok app.
4. Flexible timings: A student’s schedule is quite unpredictable. The timings of some of our commitments are hard to determine in advance. Lunchclub has only a few time slots avaliable to meet, making it tough for students who cannnot predict what their weeks will look like

To prioritize these painpoints, I will use two factors:

1. Frequency of pain point
2. Intensity of pain point

Using these two factors, I graphed each painpoint as follows (numbered above):

The high frequency, high intensity quadrant is what I want to focus on. For the remainder of this recommendation, I will prioritize solving the painpoint of building lasting relationships.

Features

I thought of four possible features that Lunchclub could implement in order to tackle the painpoint of building lasting relationships.

1. Follow-up meetings: the key to a lasting relationship resides in the frequency of meeting one another. Currently, Lunchclub resides on the model of introducing new people and putting the onus on the user to follow-up with their past connections. I believe that follow-ups are integral to networking and is the foundation for really strong relationships. The tradeoff that Lunchclub will need to consider with this feature is that one meeting will be ‘wasted’ in the sense that people are not meeting new individuals.
2. Pen pal: Lunchclub could introduce a feature to allow asynchronous messaging between people you have met before. This mitigates the above problem with follow-up meetings as you don’t need a Lunchclub meeting to follow-up with the individual. To encourage people to use this feature, the pen pal could either:
   1. Give interesting prompts for both people
   2. Automate the initial follow-up by telling the other person how their last meeting went
3. Contact detail sharing: For many people, email may not be the best route of communication, but that is what Lunchclub is based around (i.e. the platform connects users by sending each an email introducing each other). The platform could have an after-meeting landing page where participants could share each others contact info and socials.
4. Mutual connection meetups: One of the best ways to network is through warm intros, where a friend of one person can introduce them to another person that they know. This is a proven networking strategy as it provides a common basis, which is the mutual friend. Lunchclub could create a new set of meeting styles where 3-4 people are paired together for a meeting where all of them have met at least 1-2 of of the others in the past meeting. This would provide a sense of novelty and could improve the network effects of the platform as it changes from building 1:1 relationships to 1:many relationships.

To prioritize these features, I will use two factors:

• Engineering effort
• Impact

From my judgement, I thought the following was apt for the above features

I drew circles representing my confidence interval for where the feature could possibly place. For example, the first feature has a low effort variation but has a much higher variation for impact.

From the above, I would prioritize the first feature recommendation as something that can be done within a sprint or 2 (high impact, low effort), but would also develop the last feature as a longer-term goal (high impact, medium effort).

Key Metrics

The primary goal of these features was to solve the painpoint of being unable to build past relationships. There are a couple of ways that I would measure this:

• Number of users who have contacted past contacts with a possible segmentation of when they met in the past (eg. number of users who have contacted people who initially met 2/3/4 months in the past)
• Quality of relationship metric that can be included in occasional surveys

I would also want to measure second order effects with this approach, specifically engagement and retention. My hypothesis is that these features could significantly boost the following metrics

• Engagement: weekly/monthly repeat users, number of weekly meetings, weekly active users
• Retention: churn/week, churn/month, opt-out rate/week

That concludes my product recommendations!

Business Recommendations

All my business recommendations revolve around improving one aspect of Lunchclub that I believe has the highest probability of creating a runaway startup: network effects.

For those reading who don’t know what network effects are, consider why Facebook and Instagram took off. To be honest, it wasn’t the quality of the content on their platform; rather, it was the ability to see and interact with our friends online. As more people get added to the network, the platform becomes more valuable, creating a flywheel for growth.

Lunchclub is in a similar position where it can truly capitalize on network effects. However, there are certain techniques that I believe Lunchclub could use to boost these effects even further.

• Piggyback on existing networks: AirBnB is a classic example of a company that piggybacked on existing networks. To kickstart the AirBnB network, they cross-posted listings on Craigslist, which already had a significant following. Lunchclub could do the same by introducing their services for specific companies, like a Lunchclub version for Amazon. These companies already have network effects and Lunchclub could simply use these network effects to reach more people
• Create Lunchclub meetups: Instagram created two types of networks - networks on their app and in-person networks through Insta Meets (Instagram users meet up and create photos together to post on the platform). The latter is not well known, but it was an integral strategy for building communities around Instagram. Lunchclub could benefit from something similar: networking meetups could be held in-person (when it’s legal) to foster a greater sense of community on the platform. This has asymmetric upside as it could signficantly boost the ‘community’ feel of Lunchclub.

Conclusion

From my cursory research into the company and usage of the platform of the last 3 months, I believe Lunchclub has great potential to succeed. It’s got a winning strategy to address the need of networking where many other companies have failed. Excited to see where the company heads!

Citations

https://medium.com/marketfit-rho/lunchclubs-hayley-leibson-shares-the-6-step-approach-she-used-to-raise-a-4-million-seed-dfc5c52056fe https://www.alexanderjarvis.com/lunchclub-pitch-deck-to-raise-a-4m-seed/ https://www.cnbc.com/2020/09/01/lunchclub-tops-100-million-valuation-amid-quarantine-usage-spike.html