Lessons Learned from 2008

Original Tweetstorm: https://twitter.com/avichal/status/1238298350077669377

I’m sharing lessons from running a startup through 2008. I tweeted this as well and cleaned up a bit for the blog. Hope it’s useful to founders.👇

1/ Most important — prioritize your health and family. It’s clearly stressful as the world feels like it’s falling apart. If you are healthy and have your loved ones, you will be fine long term. 

2/ Duration — this will last a lot longer than you expect. From peak S&P in 2008 to bottom was 500 days. We are only 20 days in to this correction. Maybe it’s different this time…more likely we have a long way to go. And it will feel a lot worse than it does right now. 

3/ Cut costs faster and deeper than your intuition suggests — cutting costs is hard but probably the best way to extend your runway. Letting people go is *very* hard. But if you are forced to, cut once + deeply so you never cut again and no one else is worried about their job. Make it clear to your remaining team members that you have cut so deeply so that you never have to come again.

4/ Be patient on hiring — Hire slowly. Many talented people may be available in a year and companies that have strong cash positions will aggregate amazing people, which will be a long term competitive advantage. The best people will want to work with the best people, especially in a downturn where talent aggregates in a small number of places.

5/ Revenue often falls as a second order effect — It takes time for the second order effects of unemployment, leverage leaving the system, businesses cutting back, governments losing tax revenue, etc. to materialize. Your revenue may be fine now but may not be in 6 months. 

6/ Help people stay long-term oriented — spouses will lose jobs, friends will worry about layoffs, more people will be sick bc they are stressed, etc. Help people think about focusing on creating long term value. If you have cut deeply enough, you should be able to survive. This will give you emotional energy to support others.

7/ Reputations are made or lost in times like these — people will talk about these crazy times in a decade. Your peers, your investors, and your employees will remember how you behaved on the way down and in tough times. Never lose your integrity or make rash decisions. 

8/ Negotiate — you’d be surprised how much leverage a macro turn will give you. You can probably negotiate a lot of things in your business life. Do not negotiate at the expense of long term relationships and integrity. But if you need something, ask. 

9/ Deals & contracts will fall apart — Contracts are only as good as the people signing them. Investors may pull out after giving you a terms sheet. Partnership agreements may be cast aside. Landlords may change terms. Be ready for people to back out of signed contracts. 

10/ Fundraising will slow down and pick up in 6 mo — Late stage doesn’t want to catch a falling knife and early stage investors know that only companies that have to raise are raising now. Both drive down valuations. But eventually investors have to invest, so they will. In my experience, uncertainty drives people to pull back and be more conservative, and investors know startups that can avoid fundraising in uncertain times will wait. Those who cannot wait are the ones raising now, so investors have more leverage in the short term.

11/ There will be monster businesses built soon! Many of the best businesses scaled in 2008 or 2001. Google, Facebook, Airbnb, etc. A correction accelerates behaviors, i.e. share economy scaled in 2008 b/c people needed extra cash. Remote work & crypto are my bets for this one 

12/ Remember: Few events are as good as they seem in the moment and few as bad as they seems in the moment.

Heuristics to Generate Startup Ideas

I regularly help pre-seed entrepreneurs identify and evaluate potential startup opportunities. The following is a set of heuristics I’ve developed and collected over the years that might of use.

These heuristics are just a starting point to identify interesting, unsolved problems in the world that may turn in to great startups. Entrepreneurs must still go deep on understanding their customers, the business model, have a clear answer to Why Now? there is a big shift in the market, and focus on identifying problems rather than generating ideas that sound good.

  1. Tools inside a big company — High growth companies often first experience a problem other companies will soon face and solve it internally. Bringing these tools to everyone else can work well, e.g. Asana was inspired by the Facebook internal tasks tool, Cloudera was inspired by Facebook’s internal data infrastructure.
  2. Consumer generation shift — Think about a consumer company that started ~7 years ago. Generations shift in about 7-10 years, thus tastes evolve, and platforms shift. Thus there is an opening to acquire customers, e.g. Snapchat started 2011 while Facebook started 2004
  3. Big company acquisitions — If a company is acquired for $5B+, consider a competitor. This may indicate product maturity, a departure of talent after earnouts, or becoming focused on profits vs. product investment. Now may be a great time to re-imaginge LinkedIn or Github
  4. Market validation from other startups — Find a company that recently raised a series A and approach solving the problem differently. Often the first company to identify an opportunity will not win but does indicate customers have a real problem they want solved.
  5. Revisit ideas that were too early — Look back at ideas that were hot 10 years ago because entrepreneurs have a tendency to be too early. See if any might make sense now, e.g. Instacart vs. Webvan.
  6. Invert a successful company’s core competency — Take the core strength of a company and flip it on its head to see what a product experience may be if you did the opposite, e.g. Snapchat was based on ephemerality and Facebook based on permanence of identity, or Southwest was based on a point-to-point model with only one type of plane whereas other airlines use a hub and spoke model with multiple types of planes.
  7. Extend consumer behaviors to businesses — Take a product used by consumers and think about how it might apply to prosumers, enterprises, and small businesses, e.g. LinkedIn vs Facebook.
  8. Turn open source projects in to SAAS businesses — Find open source projects that are very popular and turn these in to out of the box services for enterprises, e.g. PagerDuty is like Nagios.
  9. Compete with old world industries using software — Find an industry that is resistant to using software and build a vertically integrated version to compete with that industry, e.g. Atrium vs. law firms.
  10. Low NPS, fragmented industries — Identify a large, a highly-fragmented industry with terrible net-promoter-score and build a modern, software-first version of their product, e.g. OpenDoor (credit to Keith Rabois for this one)
  11. Make things programmable — Take a desktop utility, move it online, and make it more powerful by making it more programmable, e.g. Airtable and Notion (similar to Elad Gil’s idea around devsumers)
  12. Unbundle a company — Look at the functions of modern companies and unbundle them to make them available for companies of all sizes, e.g. WeWork is the unbundled facilities group at most big companies
  13. Look to engineers’ hobbies — Figure out where hobbyist engineers are spending their spare time and how to productize that (Chris Dixon’s idea)
  14. Understand how teens communicate — Figure out where teenagers are spending their time to see what the communication products of the next decade may be. Teens are unencumbered and very creative, so seek to understand the motivations instead of judge.
  15. Travel and observe non-software solutions — Travel to new places to discover non-software solutions that other countries/people/markets have come up with to various problems, e.g. I was in Moscow in 2006 and people could just hail any stranger on the street to drive them somewhere like a cab. It may not have been a big leap if you saw this behavior to Uber/Lyft on your phone.
  16. Look at behaviors in certain geographies — Smaller, homogenous culture countries (Sweden, Korea, Japan) are early adopters of many technologies. Spend time there to understand what they are doing that isn’t globally popular yet, e.g. watching other people play video games or mobile payments took off in these markets very early relative to the US.
  17. Unbundle the government — Think about what services the government offers and how you might build a profitable business model around these, perhaps even by offering these services in a more cost-effective way back to the government, e.g. SpaceX is a private NASA, FedEx is a private postal service.
  18. Verticalize an already successful business — Find a massively successful business such as Salesforce and target it for a vertical that is traditionally challenging to access, e.g. Veeva is a pharma industry specific CRM worth $15B+ and raised only $7M in funding.
  19. Build something for a distribution channel — Identify a new distribution channel that would enable access to a new customer, and then think through what that customer wants that the distribution channel does not want to provide, e.g. Pinterest and Yelp built on SEO, Paypal built on eBay, YouTube embedded inside MySpace before MySpace had a video product.
  20. Draft on a regulatory shift — Build behind a big regulatory shift to compete with large incumbents who will be slow to move and have regulatory + technical baggage, e.g. Oscar Health after Obamacare. I’m sure some companies will spin up to help people transition to a post-GDPR world.

I’m sure there are other good heuristics to use as a starting point. If anyone sends me others, I’ll add them to my blog post as a reference for others.

The Fat Torso

The Internet’s potential to enable new businesses by building for customers in the Long Tail is well-documented. However, for many Internet businesses, the Fat Torso is now more important.

As the Internet has scaled, we have seen the emergence of markets with Fat Torso and successful businesses that first target the Fat Torso of customers.

The Shape of Markets

Markets can have a big head, fat torso, or long tail. Offline businesses tend to have a fat head because there are real economies of scale. For example, the top five automotive manufacturers own about 50% of the global market and the top 15 own about 90% of the market. Internet businesses such as YouTube have both a fat head and a long tail uniquely enabled by the Internet.

As the Internet has matured, many markets have grown to also have “fat torsos.” These Fat Torso markets are often fantastic markets for startups to enter and the right entry point is through the Fat Torso.

What is the Torso?

Customers in the torso are between the head and the tail. They make decisions quickly and generate significant revenue or engagement per customer. Customers in the torso generate more value for your business (revenue or engagement) than tail customers, i.e. the area under the curve is larger. Unlike customers in the head who often make unreasonable demands because they are used to being catered to and having market dominance, torso companies will make feature requests that apply to many customers. Torso companies are hungry to compete with head companies and will try new solutions in order to gain an edge against their bigger competitors.

What is a Fat Torso?

Not all markets have “fat” torsos. If the torso is “fat,” there are many customers who meet the torso criteria (who spend lots of money and move quickly). Thus you can scale your business quickly: lots of customers * high revenue per customer (and they move quickly).

Go To Market: Torso First

The Fat Torso is the best place for startups to prove their business model. Often the go-to-market:

  1. Prove the model with the fat torso customers – torso customers move quickly, offer significant value to your business, are eager to scale their businesses, and do not have significant leverage over you to make unreasonable demands.
  2. Scale to support the head – now that you have some scale, you can negotiate with larger customers or partners. Your company has the runway to wait out the long time horizons that large customers require.
  3. Build self-service tools to serve the long tail – these customers offer significant revenue at scale but require a significant investment of self-service tools. They serve as a moat around your business if you can get them onboarded. Now that you have the customers they aspire to be (torso customers)


  • Online Ads — target sophisticated online advertisers such as game developers who scale ad spend quickly to $10 million+. Targeting smaller businesses in the tail requires hand-holding, self-service tools, and they will not scale ad spend over time. After you have the torso advertisers, convince large advertisers such as brand advertisers to use test budget and build out self-service tools for the tail.
  • SAAS — start by targeting companies large enough to give their employees autonomy to make purchasing decisions (say 100-1000 people). Avoid Fortune 1000 companies and avoid two-person startups. After you have many companies from 100-1000 people using your product, you can start moving up stream to enterprises and then build self-serve tools for startups.
  • E-Commerce Marketplace — find sellers who can use your platform as part of their existing full time business. At the same time, find buyers who will buy more frequently on your platform not just once per year. Etsy is a good example. It has many suppliers supplement their income substantially via Etsy and Etsy’s customer base appears to have a fat torso with 60% of customers buying only once per year. Over time you may be able to scale to head sellers like Nike or Dell. This is the path eBay took after scaling its core business by first working with sellers who made a living on the eBay platform.

The Torso First approach not the right fit for every market since not all markets have a Fat Torso. However, it is an increasingly common theme in businesses that scale to $100M+ in revenue, both because the markets that have Fat Torsos are great markets and because businesses that scale often use this strategy to scale quickly.

We discussed very similar ideas in some meetings at Facebook. I’m not sure who came up with these insights exactly, so thanks to everyone involved in helping form the ideas: Andrew “Boz” Bosworth, Alex Himel, Pratiti Raychoudhary, and Jon Lax.

Why now?

One of the most important questions for an entrepreneur is “Why is now the right time for this idea? Didn’t others try previously and fail?”

There are rarely new ideas in startups. If you have an idea, someone probably already tried it and failed — maybe they mis-executed, there was something counter-intuitive about the customer or business model, or they were too early (correlated with the market not being big enough yet). If you assume other people are smart, which is generally a good practice, then they executed well and there wasn’t anything counter-intuitive going on.

Earlier startups were likely just too early. So if you can’t point to an obvious misstep in prior attempts, “why now?” is often the critical question to answer.

There are at least 5 classes of good answers to why now:

  1. new technology allows products that simply weren’t possible before, e.g. battery tech and electric cars
  2. new regulation, e.g. Obamacare
  3. new business model has emerged, e.g. advertising could support free content online
  4. new user acquisition channels, e.g. search/SEO, FB Platform v1
  5. customer behavior has shifted, e.g. a desire for ephemerality once people understood the consequences of searchable, permanent identity

1 to 4 are strong answers and generally clear cut.

5 is the trickiest. Behavior change is the most common answer I hear startup founders assert around “why now” for their company. If 5 is true, then you win BIG — these are the “It wasn’t true until it was true” sorts of startups. Facebook, Uber/Lyft, and Snap fall in to bucket 5.

The challenge is 5 is the hardest to know looking forward. You can assert it’s true, but it’s easiest to know behaviors/preferences/attitudes changed looking backward.


What this means

If you can find novel solutions to problems because of new technology, regulation changes, new business models, or a new customer acquisition channel, then you can often win big and win quickly.

However, the vast majority of startups assert they can succeed because behaviors have shifted somehow. But the vast majority of these assertions are incorrect. However, the ones who are correct will win big. You can only know if the behavior shifts are real in hindsight, but you have to make educated guesses about these shifts looking forward. And that in a nutshell is one of the hardest challenges of being an entrepreneur or angel investor.


There are only four types of metrics – quality, scale, engagement, and revenue. Effective metrics track real value, are sensitive, and easy to understand. When communicating metrics, you should call out how the operational metrics that measure progress against your current tactics will drive the topline metrics everyone at your company understands.

Categories of Metrics

Product metrics fall in to four categories:

  1. Quality – How good is your product at doing what it says it does?
  2. Scale – How many people use your product?
  3. Engagement – How deeply do people engage with your product?
  4. Revenue – How much money do you make?

These categories are deeply interconnected — scaling your userbase may drive down your overall engagement rates as new, less engaged users come in. Putting ads in to a product may drive down retention and engagement. Driving up engagement is often one of the best ways to drive growth in the long term, etc.

In general, you strive to make your product great for a small number of people, get more people to use this valuable product, deepen engagement with those people, then figure out how to make money from them, and loop back around to figure out how to build a high quality experience for an expanded set of people.

Characteristics of a Good Metric

  1. Represent real value – The more closely a metric represents the actual value to users and customers the more likely it is to guide a company in the right direction.
  2. Sensitive – a good metric will move up/down based on work you do more than exogenous factors.
  3. Simple – simple counts and time series are surprisingly effective.

For example, “number of people using our product on a given day” aka DAU is real value (more people using your product is good), sensitive (if you send a lot of notifications or if you are in the press, this will go up), and simple because you can look at the graph and quickly know if things are going well.

“The ratio of photos to videos uploaded per person” is not clearly valuable (why does this ratio matter?), sensitive (the ratio may not change a lot unless you fundamentally change your product in a big way), and is not simple (is the above shape good? is it going up because people are uploading more photos or the video upload flow is broken? what would you expect to be a “good shape” to that curve?)

Communicating with Metrics

It’s easy to lose context on why certain metrics are a priority and how all of a company’s metrics fit together. As an employee, this loss of context can feel like there is no strategy. It can lead to teams that struggle to communicate because they do not have context on why other teams’ metrics matter.

The reason this happens is that most people conflate topline and operational metrics.

Topline metrics measure progress towards your mission. They are high level and stay consistently tracked for years. These are the metrics your board of directors want to know about, such as daily active users (DAU) or revenue. You will have only one or two topline metrics for each of the four types of metrics (quality, scale, engagement, revenue).

Operational metrics measure the progress of strategies and tactics. They are in service of a topline metric. For example, using search ads to acquire customers is a tactic, measured by cost of customer acquisition using search ads. The overall strategy you are using may be to pay for new users, measured by an overall blended CAC. This strategy is in pursuit of monthly active users (a topline growth metric).

Operational metrics will change as tactics or strategies evolve. However if you articulate how operational metrics roll in to topline metrics, this will not feel jarring because people will understand how the old and new operational metrics were in pursuit of the topline metric. You can/should also reinforce that topline metrics are proxies for the value you want to create in the world, not the end unto themselves, to keep everyone grounded and focused.

Example 1: Google Search

Here is a simplified example using Google Search. There are many possible operational metrics for each topline metric, and the operational metrics you focus on will change over time.


Example 2: eBay

This is a more complicated example because you have a two-sided marketplace with buyers and sellers, and it’s important to measure the health of both sides of the marketplace.



The biggest risk in creating a metrics informed culture is that over time, people conflate metrics and goals. Metrics are a proxy for value and an abstraction to help you measure progress. Don’t lose sight of the real goals: to create real value for customers. If you lose sight of the value you are ultimately creating, you can move metrics for the sake of moving metrics, picking incorrect metrics targets, and ultimately will be disrupted by another company that actually creates real value for customers.


The right way to network

“Networking” and “connecting” as most people attempt it is a waste of time. [1] As most people with large networks know, focusing on helping others is the best strategy. The more people who ask you for help, the more connected you actually are, and the more able you will be to use your network to accomplish your goals in the long term.


Why is it important to have a network?

Networks create tremendous value through access. All of the best things that have happened in my life have happened because of access to the right people. Whether it is finding a job, starting a company, selling a company, angel investing, or finding my future wife — all of the best things in my life have been a result of having invested in a robust group of friends and colleagues over the years.

None of these outcomes were my goal. They happened as a side-effect of helping others in a large and valuable network.

What is the right networking strategy?

Too often people focus on trying to “connect” with others in the hopes of being able to accomplish their ends through a connection. This is a flawed strategy.

The right way to “network” is to focus on helping others in your network. Generating value for others makes you worth being connected to, which creates greater and stronger inbound connections, and a greater willingness for others to help create value for you.

One simple way to think about influence in a network is PageRank. A webpage is more important if more webpages link to it. And a page is more important if other important webpages link to it. Thus the measure of authority and influence is a function of both quality and quantity of inbound links to a webpage.

The same applies to human networks. If people route requests to you, you are more influential in the network because you are more capable of helping others get things done. And the more influential the people routing requests to you, the more influential you must be because of your ability to help these more influence people. Then, when you make an outbound request you will have more influence (PageRank) that flows out over your outbound request, which increases the likelihood your network will help you.

Where do you start?

1. Figure out how you are uniquely able to help someone. Call this your “unique asset.” Everyone is an expert at something, has proprietary access to some group of people, or has time they can turn in to value for others. For example, if you are a student, companies want to recruit on campus so you can be a conduit to the campus.

2. Find people you can help and actually help them. Don’t just ask how you can help — figure out how to help. Don’t just offer to help — actually help. This is effectively finding product-market fit for your unique asset.

3. If/when people you’ve helped ask how they can repay your kindness, tell them to pay it forward. You now have someone in your network with a different unique asset than your own. This broadens the set of people you can help to those whom you could help directly + those you could refer to someone in your network.

Rinse and repeat until you have a massive, useful network that will help you and everyone else you know.

[1] – The way most people “network” and “connect” feels extractive to me. This is different than asking for help with a clear ask. Asking for help is great and more entrepreneurs should do it.

Onboarding a New Product Manager

Once a startup has 8-10 engineers a company often needs to bring on its first product manager. Below are some of the best practices I’ve learned over the years for ramping up a new PM.

Why is it hard to onboard a new PM?

Ramping up a new PM is challenging for two reasons:

  1. Product Context – to help a team make the right decisions, a PM has to help a team synthesize qualitative information, data, technical architecture, design decisions, go-to-market strategy, etc. This synthesis will take a long time.
  2. People Context – to ship products and features, a PM has to work with a wide variety of people inside a company, likely none of whom report to her.

How to Help: Product Context

  1. Set aside time at the end of every day for some fixed period of time (say two weeks) to answer any questions she has from that day.
  2. Give her access to all of the research, data, past product specs, presentations, sales material, blogs that she should read regularly, etc. and ask her to read as much of it as possible as quickly as possible. Let her know that she may have to read it two or three times in the first month to really understand it.
  3. Have her get as much context as quickly as possible on the other parts of the business — have her sit in on a budget meeting, go out to meet customers, sit in on sales calls. Whatever it takes to get a holistic understanding of how the business works outside of the product.
  4. Help her behind the scenes by helping her do the work without anybody knowing. For example, if the next deliverable is a spec, help her co-author it, edit it, and make it high-quality. Then when it goes to the engineering team she will establish credibility very quickly because it will come in a form that they expect and that a quality bar that they’ve come to expect from you. This will make her successful more quickly and in the long term create the least amount of pain for you because you’ll have landed her well with the team.
  5. Let her know what the ramp up will be very explicitly along with what the major milestones are for your involvement. For example a hypothetical timeline might be: for the first month you will help her edit the specs, can help her draft any emails or answer product questions for her 24/7, and will be in product team meetings. At the second month you will stop doing editing specs, she is expected to run meetings, and you will back out of those meetings. In the third month, she needs to be able to have enough context to make product decisions on her own and you will only get involved if the team is not hitting its ship dates or missing metrics targets. That way there’s no ambiguity about whether you are involved because she is not doing well, if you’re micromanaging, or if she is still in the ramp-up phase.

How to Help: People Context

  1. Let her shadow you for a few days to understand how you run meetings and what the various teams expect from a product manager. Ask other teams (engineering, design, sales, marketing, etc.) on her behalf if she can shadow them or sit in on their meetings to get up to speed.
  2. Give her a list of everyone she needs to have a great relationship with, explain what they do, what motivates them, and what they expect from her. Help her build good working relationships with everyone she needs to have good relationships with. In part this is innate, and in part it’s helping her understand who these people are, what motivates them, what they expect from their PM, what they don’t like, how best to build those relationships, etc.
  3. Help the team understand what she’s responsible for and what she’s not. They have come to expect certain things from you and she may not be able to give them all of that immediately, so they need to have expectations set appropriately up front too.
  4. Often as the former PM, new PM manager, or PM turned CEO of a startup, people will let you know in subtle ways if she is not delivering something the team expected. Help her interpret these signals and understand there is no ambiguity between what she expects and what the team expects. Reinforce this is not about micromanaging or a lack of trust between teammates. You simply have more context on the people to read subtle signals and are passing her this context based on your experience working with this team.

Be patient — often the new PM is stepping in for someone who lived and breathed the product for many years and who helped build out the team around them. These are big shoes to fill and becoming an effective PM takes an uncomfortably long time.