There are four key factors that shape every marketplace:<ul><li>network effects;</li><li>type of supply;</li><li>incentives;</li><li>size and frequency of interaction.</li></ul> Understanding those factors — and structuring your platform accordingly — can make the difference between a wildly successful marketplace and one that never gets off the ground.

Four key factors that shape every marketplace

Simplistically, a marketplace is a place where buyers and sellers meet to conduct commerce. In the online setting, that definition still holds true — online marketplaces give buyers a choice in what they would like to buy and sellers a variety of people to sell to. I classify Airbnb as a search marketplace, where the guest selects where they will stay, choosing what to “purchase” in the marketplace.In recent years, the definition of a marketplace has stretched to include on-demand models that are considered marketplaces because of the independent contractor relationship on the seller side.

Definition of a marketplace

Advantages:<ul><li>They involve low capital costs, as inventory is brought to the market from the suppliers.</li><li>The market can self-correct by offering more of the good or service as buyers demand it more, making each marketplace function like a mini economy.</li></ul> Disadvantages:<ul><li>Because marketplaces need a certain amount of supply from day one, they are super hard to start.</li><li>It is hard to control the quality of inventory.</li></ul>

Advantages and Disadvantages of marketplaces

All marketplaces exhibit a form of network effects, where each additional user on either the demand side or the supply side enhances the utility of the network for all users. Airbnb has global network effects because travel is global. Companies like DoorDash and TaskRabbit have root density network effects - each market grows independently of another. Supply gains little value from demand growth in another market, and demand gains little from supply developing in another market. These businesses grew from being the first mover in each market, then aggressively building supply, city by city.

1. Type of Network Effects: How will your marketplace grow?

Another factor that dictates the dynamics of the marketplace tremendously is the uniqueness of your supply.Airbnb offers heterogeneous supply, meaning that each listing is unique. This creates a competitive moat — it is hard to replicate that supply — and an inherent advantage to Airbnb. Heterogenous supply generates cognitive load. Many other marketplaces offer homogeneous supply, meaning that one supplier’s goods are roughly indistinguishable from another’s. Unfortunately, that makes it easier to replicate the business, as the supply is not unique and can be quickly copied (think Uber)

2. Homogeneous vs. Heterogeneous Supply: Is your startup a search marketplace or a matching marketplace?

Heterogeneous - search marketplaces, guests are responsible for finding their desired listing. Once you have unique supply, it is up to the product team to build a compelling demand-side product experience that simplifies search, creating as much of a one-click purchasing experience as possible. Homogeneous - matching marketplaces, where demand doesn’t care which supply they receive as long as it is above the standard.Homogeneous marketplaces compete on other aspects of the product offering, like price, greater loyalty, brand distinction, or what values the marketplace ultimately stands for

Homogeneous vs. Heterogeneous - continued

There are two broad incentives to keep transactions on the platform: trust and convenience. Technology can bridge trust by offering:<ul><li>Identity Verification —Can I trust them?</li><li>Reviews — What do other have to say about it?</li><li>Assurance — If something happens, can I get help?</li></ul>Technology can offer greater convenience by facilitating:<ul><li>Payments — Will I get my money or the product/serivce?</li><li>Messaging — Can I easily communicate and reach out to the other party?</li><li>Transactions — Did I buy the good or service? Did it reduce the burden of going back and forth multiple times?</li></ul>

3. Two-Sided Incentives: How will you keep both buyers and sellers in the marketplace?

The size and frequency of interaction in a marketplace are paramount to its health. Frequency of interaction determines how much liquidity there is on the platform. . Larger-dollar transactions are better for the platform because they generate more economic activity. Some businesses are high dollar but low frequency. Airbnb fits squarely in this category. Every marketplace should strive to drive higher-frequency usage. Liquidity is key. Making sure that transaction volume is thriving is more important to a new marketplace than getting the largest revenue per transaction.

4. Size and Frequency of Interaction: What are the unit economics of your marketplace?

The six years I spent building products at Airbnb, scaling the marketplace over 100X, transformed how I understand the work of starting a marketplace platform. No startup is immune to hiccups and…

entrepreneurship

Entrepreneurship

startup

Startups

product

Product & Design

Four Questions Every Marketplace Startup Should Be Able to Answer

The photos I take, the movies I buy, the music I rent, the emails I write and receive, the messages I send - all of these are captive. They exist within a database controlled by Google, Netflix, Spotify, Gmail, WhatsApp. If these services disappear, so does my ownership/rentership record. In a future where digital assets are worth not $10 or $20, but hundreds or thousands of dollars, ownership that survives a company becomes an essential substrate of commerce.

A permanent ownership record that exists beyond and outside a company

Internet hegemons have decimated entire spaces: social networks, advertising technology, video streaming and rental, paid email, infrastructure. Their economies of scale and network effects mount significant barriers to competition. What’s a startup to do? Compete on a different axis: reward users with tokens. A social network rewards its most valuable users in the coin of the realm. As the network becomes more valuable, so does the user’s stake in the company. Web3 companies employ tokens to reward their customers for providing value, giving startups the upper hand.

Paying customers in “equity.”

Twenty years ago, startups IPOed after 4 years. Today, it’s 12, driven by several factors but regulatory costs present the principal one. Crypto companies access pools of capital web2 companies cannot because the regulation doesn’t exist. Where nascent rules are present, the regulation isn’t yet a warren of legalese- yet. Within this freedom to maneuver, Defi protocols invent new financial instruments. The <a href="https://research.paradigm.xyz/cartoon-guide-to-perps">perp</a>, a perpetual swap, has become the most traded crypto derivative). Perps don’t exist outside crypto.

Regulatory arbitrage.

Open-source software is arguably the most important motive force powering innovation in technology. Open-source software powers every server and most software. Crypto empowers open-source projects to monetize their innovation in a way web2 never achieved. DAOs bind the loose community of open-source software contributors through tokens. Tokens provide an ongoing financial link that aligns long-term incentives. Foundations with significant token stakes fuel the ongoing protocol development, balancing the capitalism needed to fuel progress with the benevolence of the open-source ethos.

DAO/Foundations:

The fifth cluster of innovation benefits central banks, not startups.CBDCs (Central Bank Digital Coins) are programmable money: central bank digital coins enable a government to focus stimulus as granularly or as broadly as they like. Imagine a reserve bank stimulating the electric bicycle industry in the Kansas City metropolitan area by providing residents new US digital dollars that are only valid within bike shops in Kansas City.

Bonus: great for central banks

As I’ve explored web3 I’ve been forming a mental model of the major innovations powering this enormous wave of innovation. At the moment, I see four clusters:
 A permanent ownership record that exists beyond and outside a company. The photos I take, the movies I buy, the music I rent, the emails I write and receive, the messages I send - all of these are captive. They exist within a database controlled by Google, Netflix, Spotify, Gmail, WhatsApp.

crypto

Cryptocurrency

technologyandthefuture

Technology & The Future

Four Fundamental Innovations of Web3 that Will Upend Web2 Incumbents by @ttunguz 

Some neural networks are very expensive to train. This led to the popularization of an approach known as pre-training, whereby a neural network is first trained on a large general-purpose dataset using significant amounts of computational resources, and then fine-tuned for the task at hand using a much smaller amount of data and compute resources.

Expensive to train?

The use of pre-trained networks allows a startup, for example, to build a product with much less data and compute resources than would otherwise be needed if starting from scratch. This approach is also becoming popular in academia, where researchers can quickly fine-tune a pre-trained network for a new task, and then publish the results.

Pre-trained networks give smaller teams a leg up


Pre-training has continued to evolve with the emergence of foundation models such as <a href="https://ai.googleblog.com/2018/11/open-sourcing-bert-state-of-art-pre.html">BERT</a>, GPT, <a href="https://openai.com/blog/dall-e/">DALL-E</a>, <a href="https://openai.com/blog/clip/">CLIP</a>, and others. These models are pre-trained on large general-purpose datasets (often in the order of billions of training examples) and are being released as open source by well-funded AI labs such as the ones at Google, Microsoft, and OpenAI.  They allow startups, researchers, and others to quickly get up to speed on the latest machine learning approaches without having to spend the time and resources needed to train these models from scratch.

Foundational models

One of the risks associated with foundation models is their ever-increasing scale. Neural networks such as <a href="https://github.com/google-research/text-to-text-transfer-transformer">Google’s T5-11b</a> (open sourced in 2019) already require a cluster of expensive GPUs simply to load and make predictions. Fine-tuning these systems requires even more resources. More recent models created in 2021-2022 by Google/Microsoft/OpenAI are often so large that these companies are not releasing them as open source – they now require tens of millions of dollars to create and are increasingly viewed as significant IP investments even for these large companies.

The risks of foundation models: Size & Cost

Dataset alignment can also be a challenge for those using foundation models. Pre-training on a large general-purpose dataset is no guarantee that the network will be able to perform a new task on proprietary data. The network may be so lacking in context or biased based on its pre-training, that even fine-tuning may not readily resolve the issue.  Any startup leveraging foundation models in its machine learning efforts should pay close attention to these types of issues.

The risks of foundation models: Outsourced innovation

The opportunities and risk around using hosted and pre-trained models has led many companies to leverage cloud APIs in the “experimentation phase” to kickstart product development. ﻿Once a company has determined it has a product-market fit, it often transitions to self-hosted or self-trained models in order to gain more control over data, process, and intellectual property. This transition can be difficult, as the company needs to be able to scale its infrastructure to match the demands of the model, as well as manage the costs associated with data collection, annotation, and storage.

The tradeoff 

Foundation models and cloud APIs bring opportunities, risks, and limitations for startups that want to innovate in AI and machine learning.

artificialintelligence

Artificial Intelligence

How to Use Massive AI Models (Like GPT-3) in Your Startup

There’s another sorta, kinda moon in a sorta, kinda orbit around Earth that was discovered only in 2016. The quasi-moon - named Kamo’oalewa, after a Hawaiian word that refers to a moving celestial object - is not much to speak of, measuring less than 50 m (164 ft) across. It circles the Earth in a repeating corkscrew-like trajectory that brings it no closer than 40 to 100 times the 384,000 km (239,000 mi.) distance of our more familiar moon. Its odd flight path is caused by the competing gravitational pulls of the Earth and the sun, which continually bend and torque the moonlet’s motions.

One more moon, kind-of

And according to a <a href="https://www.nature.com/articles/s43247-021-00303-7">new study in Nature</a>, we may at last know how it was formed.The moon’s been getting bombarded by space rocks for billions of years, resulting in all manner of lunar debris getting ejected into space (nearly 500 bits of which have made it to the surface of the Earth as meteorites). Kamo’oalewa is one such piece of lunar rubble that spiraled away from the moon.But rather than landing on Earth or simply tumbling off into the void, it found itself a quasi-satellite in its own right.

The origin of Kamo’oalewa

Kamo’oalewa won’t stick around all that long, as its current trajectory is not entirely stable. According to estimates from Sharkey (researcher) and others, the object will remain an earthly companion for only about 300 more years - nothing at all on the cosmic clock - after which it will break free of its current gravitational chains and twirl off into the void. Originally a part of the moon, then a companion of Earth, it will spend the rest of its long life traveling on its own.

Not here for long

I never knew about this quasi-moon. If you are into space and space exploration I highly recommend Jeffrey Kluger's weekly newsletter from Time.com.

Scientists have learned where Earth's 'second moon' came from—and how much longer it'll stick around.

scienceandnature

Science & Nature

Earth Has a Second Moon—For Another 300 Years, At Least

Only drug dealers and software companies call their customers ‘users’

Focus on people

What is the product a customer buys immediately before they buy yours? Usually, customers, especially during a process of digital transformation, are on a journey.Examples: <ul><li>modernizing the sales infrastructure from simply Salesforce to a broader suite of call recording, marketing attribution, and sales content enablement </li><li>migrating to a Kubernetes stack to power a microservices architecture with distributed tracing</li></ul>The typical <a href="https://www.blissfully.com/saas-trends/2020-annual-report/">mid-market company purchases 137 SaaS products, and the figure grows 30% annually</a>.

Customers are on a journey

Which is the rung of the software ladder immediately beneath yours? Whatever it is, those are the companies to partner with, develop technical integrations to accelerate adoption, co-sell and generally go-to-market alongside. After a prospect has adopted the software rung below yours, their eyes will rise to yours, and that’s <a href="https://tomtunguz.com/category-creation-speed-strategy/">your opportunity to win the great game of Risk played in software.</a>

Where are you positioned?

What is the product a customer or a prospect buys immediately before they buy yours?
We’ve all heard the words digital transformation. Now that I’ve given you time to unroll your eyes and let the sarcastic comments dissipate, please give me a chance to make something useful of this idea.
As a customer progresses along their journey, they evolve from a previous inferior state to a new superior one. The journey might be a migration from an on-premises data stack to a modern cloud data pipeline.

marketingandsales

Marketing & Sales

What is the Product the Customer Buys Before They Buy Yours? by @ttunguz

Never ASSUME, because when you ASSUME, you make an ASS of U and ME

In the <a href="https://en.wikipedia.org/wiki/TV_Land" target="_self">TV Land</a> 2006 documentary The 100 Greatest TV Quotes and Catchphrases, <a href="https://en.wikipedia.org/wiki/Lowell_Ganz" target="_self">Lowell Ganz</a> credits Belson with including the catchphrase in the script of the season 3 <a href="https://en.wikipedia.org/wiki/The_Odd_Couple_(1970_TV_series)" target="_self">Odd Couple</a> episode "My Strife in Court" (originally aired Friday, February 16, 1973).&nbsp; Ganz noted that Belson had heard it used years ago by a teacher in a typewriter repair class.

Origin of the quote

movies

Movies & Shows

About assumptions

Madhavan Ramanujam, board member and Partner at Simon-Kucher & Partners (a global strategy consulting firm) answered this question during Office Hours at Redpoint in Aug 2021. He is the author of Monetizing Innovation. Massive companies have been built using both pricing structures: Salesforce and Adobe bill per seat while Snowflake and Twilio charge per use. Deciding which to use involves considering factors such as customer preference and competition.

Price per seat or user?

If your customers demand predictable bills, then per seat pricing is the way to go. The question is do they prefer it or do they demand it? Most customers will prefer predictability, but won’t necessarily demand it. Would they switch if the pricing weren’t predictable? That’s a question worth examining in your pricing research. If you would like to create switching costs, per seat pricing with annual contracts establishes some lock-in. Usage pricing provides more flexibility to customers to try alternatives.

When to use per seat pricing

If your costs scale with usage then usage based pricing aligns your costs with your customers' spend. This prevents very large customers from being your worst customers, by generating lots of revenue, but costing you money because the account is gross-margin negative. If you are pursuing a two step go-to-market strategy with which the first user has a low willingness to pay, but the ultimate buyer has a larger budget, consider usage pricing. This is a land-and-expand motion.  If customers use the product intermittently and the value is intermittent, usage based pricing works well.

When to use usage based pricing

Madhavan tells that the most common pricing mistake in a land and expand motion is to cede too much value in the free/lesser product, scuttling any expansion opportunity.

The most common pricing mistake

There are many companies who employ a two-part tariff: a base platform fee and an ongoing usage fee to capture positive aspects of both types of pricing strategies. Segment is a good example of this. The platform fee establishes a stable relationship and the usage pricing enables the customer to scale up or down as a function of their traffic which might vary throughout the year.

The hybrid approach

Should you price your SaaS per seat or per use? That was one of the salient questions Madhavan Ramanujam answered in last week’s Office Hours at Redpoint. Madhavan provided an excellent framework for answering this question, in addition to a multitude of other insights.
The first thing to state is that massive companies have been built using both pricing structures: Salesforce and Adobe bill per seat while Snowflake and Twilio charge per use.

business

Business

A Framework for Evaluating the Right Strategy for Your Startup by @ttunguz

Often board members give entrepreneurs two bits of advice regarding scale:<ol><li>Get a mentor.</li><li>Find some “been there, done that” executives who already know how to scale</li></ol>These answers, while fine as far as they go have some important limitations. First, if you don’t know anything about scaling an organization, then it will be very difficult for you to evaluate people for that job. Second, many investor-board members don’t know anything about scaling a company either.

Time to scale a company

The following things that cause no trouble when you are small become big challenges as you grow:<ul><li>Communication</li><li>Common knowledge</li><li>Decision making</li></ul>Specialization, organizational structure, and process all complicate things quite a bit and implementing them will feel like you are moving away from common knowledge and quality communication. It is very much like the offensive lineman taking a step backwards. You will lose ground, but you will prevent your company from descending into chaos.

Give Ground Grudgingly

There are 3 things that we need to take care of when scaling a company:<ul><li>Specialization is needed because everybody cannot know everything. As you scale you will need people focused on bits of the org. Specialization creates complexity and the need for organization design.</li><li>Organizational design tells how the organization talks between the different components.</li><li>Process defines how communication actually happens.</li></ul>

How to Scale 

The first rule of organizational design is that all organizational designs are bad. With any design, you will optimize communication among some parts of the organization at the expense of other parts. Your goal is to choose the least of all evils. The basic steps to organizational design:<ol><li>Figure out what needs to be communicated;</li><li>Figure out what needs to be decided </li><li>Prioritize the most important; communication and decision paths;</li><li>Decide who’s going to run each group;</li><li>Identify the paths that you did not optimize;</li><li>Build a plan for mitigating the issues identified in step 5.</li></ol>

Organizational design

A process is a formal, well-structured communication vehicle. It can be a heavily engineered six-sigma process or it can be a well-structured regular meeting. The size of the process should be scaled up or down to meet the needs of the communication challenge that it facilitates. For new companies, here are a few things to keep in mind:<ul><li>Focus on the output first: what should the process produce? </li><li>Figure out how you’ll know if you are getting what you want at each step:</li><li>Engineer accountability into the system – Which organization and which individual is responsible for each step? </li></ul>

Process

remotework

Remote Work

corporateculture

Corporate Culture

strategy

Strategy

humanresources

Human Resources

Taking the Mystery out of Scaling a Company

They are just one-two lines of text but short is difficult, conciese ever more. So I decided to look over some examples of what a few of the great companies out there came up with.

Value propositions are deceptively hard

If I am to speak ten minutes, I need a week for preparation; if fifteen minutes, three days; if half an hour, two days; if an hour, I am ready now

Tame your work, organize your lifeRemember everything and tackle any project with your notes, tasks, and schedule all in one place.

Evernote

A note-taking tool for networked thought.As easy to use as a document. As powerful as a graph database. Roam helps you organize your research for the long haul

Roam

Convey ideas with context & clarity Meet Gamma, the Modern MemoWrite like a doc, present like a deck, discuss live or async

Gamma

Simplify how you workSecure collaboration with anyone, anywhere, on any device

Competitive intelligence that wins business.

Klue

Grow Your Revenue with GongGain critical insights into what’s happening with your remote sales team, your deals, and your market with Revenue Intelligence from Gong.

Gong

Trello helps teams move work forward.Collaborate, manage projects, and reach new productivity peaks. From high rises to the home office, the way your team works is unique—accomplish it all with Trello.

Trello

Work on big ideas, without the busywork.From the small stuff to the big picture, Asana organizes work so teams know what to do, why it matters, and how to get it done.

Asana

Where the world builds softwareMillions of developers and companies build, ship, and maintain their software on GitHub—the largest and most advanced development platform in the world.

Github

Power the Modern, Agile EnterpriseCrush your 2021 goals and keep moving forward with Wrike's work management platform.

Wrike

Google Photos

Nothing great is made aloneFigma connects everyone in the design process so teams can deliver better products, faster.

Figma

Keep life organised and work moving – all in one placeDropbox gives you secure access to all of your files. Collaborate with friends, family and colleagues from any device.

Dropbox

Stunning presentation made together.Pitch combines the best parts of productivity, design, and presentation software to help fast-moving teams create beautifully designed decks in minutes.

Pitch

All-in-one workspaceOne tool for your whole team. Write, plan, and get organized.

Notion

Great Writing, SimplifiedCompose bold, clear, mistake-free writing with Grammarly’s AI-powered writing assistant.

Grammarly

Grow Your Business with Unbounce Conversion IntelligenceGet More Conversions with Unbounce Landing Pages & Conversion Intelligence Tools

Four Questions Every Marketplace Startup Should Be Able to Answer

Four Fundamental Innovations of Web3 that Will Upend Web2 Incumbents by @ttunguz

How to Use Massive AI Models (Like GPT-3) in Your Startup

Earth Has a Second Moon—For Another 300 Years, At Least

Focus on people

Liviu Lica