My views on the market, tech, and everything else

A Dozen Lessons on Building a Business from Sarah Tavel


  1. “Ultimately when evaluating an early stage company, I say it’s a combination of art and science. The art is understanding how products work, the science is knowing how to measure it. The earlier the company, the more it is about art, which in this case is assessing what I think of the product and the use case.” 

Tavel is a great fit for Benchmark since the firm has always believed what they do is a craft. I wrote about this in my blog posts on Andy Rachleff, Peter Fenton and Bill Gurley. Fenton has said:

“because we love the day-to-day work with the entrepreneurs, [it] prevents us from scaling. We don’t have an ability to offload any part of our relationship in the way we practice it, to anyone other than ourselves. So, there’s no associates, no principals, there is really nothing beyond the group of people here and our assistants who keep our lives sane. That’s a strategy.”

What Benchmark partners like to do is invest early and work shoulder-to-shoulder with entrepreneurs. The Benchmark partners would use that approach even if it was not the best way to optimize their financial return since it is what they enjoy most about the process. The “art” of venture capital Tavel is referring to is often found in pattern recognition and that is highly related to good judgment which often comes from first party bad judgment or what Will Rogers once called watching other people pee on an electric fence (third party band judgement). It is important to note that a critically important part of the pattern recognition in venture capital is finding the exception to previous patterns that is different. Part of the pattern is: some rule that other people preciously believed was important is being broken by the best startups, but some rules that other people thought were important are being followed. It is a bit like the spot the difference game, except it must be a very important difference that delivers significant new core value to a really big market.

I have written about the science part of the entrepreneurial process quite a bit lately. Generating growth in a startup is accelerated by great data science because it allows you to measure results and apply the scientific method to growth experiments. When founders get access to great data science they have a greater ability to scale their output and most importantly make their creative contribution enduring. Data science does not eliminate the need for creative sparks, but when used effectively it facilitates creativity by enabling rigorous experimentation and increases the impact and growth of the business.    

  1. “For seed investments, it’s always first team, and second, believing in what they’re doing (as early as it is). I’ve passed on opportunities that had amazing teams, but I just couldn’t get behind what they were doing, for whatever reason. It doesn’t feel authentic to me to just make an investment on #1, if I’ll spend our conversations together trying to convince them that they should change #2. So really both need to line up for me.”

One of the reasons I write about so many different venture capitalists is to make the point by example that there is no cookie cutter way to be successful investing in startups. There are many similarities and common elements, but many differences too. My view is that there are different pools of alpha and different people search for that alpha in different ways. Of course, some investors are more successful than others. Typically the investors that are most successful chose an approach that is most consistent with their nature and unique skills. Benchmark co-founder Bruce Dunlevie puts the strength of the team at the top of the early stage investing hierarchy. Another Benchmark co-founder Andy Rachleff puts more emphasis on the existence of a very large addressable market. Of course, this is a matter of emphasis as both the right team and the right market are important. That Benchmark has adopted a craft rather than a platform approach to their business does not mean that other approaches by other venture capitalists do not work. It is just the right approach for them as individuals and gives then a unique service to offer founders.  Some founders will be more comfortable with a full service venture capital platforms. Some will not. Vive la différence. Strategy is what you do differently than your competitors. You must be different and you must be right about that difference if the strategy is going to be successful. 

  1. “Spend money very very carefully until you have product market fit. You want as lean a team as possible before you get there. There is no point of hiring more than the bare minimum team (usually just the co-founders) before you figure out what users want. Then you scale. Companies that hire before that waste runway, and that’s a shame. Once you have product market fit, you still need to be careful with hiring. 

The critical point Tavel is making is that it make no sense to pursue a growth hypothesis before the company has solved the value hypothesis. Why would you try to get more people to use a product that does not have core product value? It is not only hard to create core product value but the discovery process (if it happens) inevitably requires time and experimentation. The shorter the financial runway of the business the less opportunity exists for the startup to discover product market fit and solidify its value hypothesis. Tavel also notes that hiring is something to be approached with care even after the value hypothesis is proven. Early hires are particularly important as they create the business’ DNA place much more than later hires. Company culture is far more a determinant of success than people imagine. 

  1. “Another thing I see is scaling too quickly, particularly for local businesses. I see a lot a company launch something in one city, and before they’ve figured out the playbook, launch a bunch of other cities which burns through their cash, without really figuring out how to make it work in any one city.” 

Benchmark’s Bill Gurley has said: “We like to say that ‘more startups die of indigestion than starvation.’” One cause of death by indigestion is premature scaling of a growth hypothesis that has not been proven. Putting fuel in a rocket which does not have an attractive target and a working guidance system is a bad idea. Even if there is product market fit, there is a right time to spending big on generating growth and that time if after a sound growth hypothesis has been proven. 

  1. “I’d say that companies that have the greatest chances of becoming sustainable unicorns have incredibly strong network effects or economies of scale. More often than not, you find this with platforms/marketplaces, but not always. e.g., companies like Amazon or big banks, have very strong economies of scale, that make it very difficult to compete. But they take an enormous amount of capital to build.” 

Tavel is pointing out that the primary means of defending a business against competitors today is network effects. My blog post on network effects is here. Other approaches that can create a barrier to entry like economies of scale, regulatory expertise, intellectual property and brand still matter, but they are relatively less important than they once were, especially in the case of software startups. Economies of scale are desirable and complement network effects in many cases, but Tavel is pointing out that they are not “capital light.” One of the big changes in the business world that Charlie Munger and Warren Buffett talked about at the recent Berkshire meeting was the appearance of technology businesses that do not require much capital.

  1. Growing users without growing users completing the core action is the empty calories of growth. It feels good, but it’s not good for you. What is a core action? The action that is the very foundation and essence of your product. Pinterest would not exist without pinning. Twitter would not exist without tweeting.” 

There are a number of terms used to describe the factors that can create sustainable growth. One term that I like is the concept core product value, which represents a solution a real and significant problem that is valuable enough to cause people to want to pay for a product. Core product value is first recognized when the customer connects with the product in what is known as a “magic moment.” A magic moment is a realization or reaffirmation by a customer that a product has core product value. Tavel is saying that that a “core action” is what the customer does to realize core value.

There are different types of magic moments. My taxonomy is as follows:

  1. Hook magic moment – a realization of core product value that occurs within the first 15 seconds of customer on boarding that creates retention. There should be as little friction as possible in getting the visitor to the Hook magic moment as fast as possible.  Delivering the core experience quickly without overwhelming the visitor with features and sign up requirements is critical. For example, Alex Schultz of Facebook explains:

“[if you want] people to buy an item on eBay, should I land them on the registration page or the search results page?” You’d probably land them on the search results page, and that would get them to that magic moment faster. What is the moment when you used AirBnB? What  is the moment when you used DoorDash, when you used Slack, when you used Facebook,  LinkedIn, WhatsApp? Whatever the product, what is that moment when you went, ‘Yes. I’ve got  it.’ On Facebook it’s friends. Getting  that first person reserve your house, staying at that first property, those are our magic  moments. eBay, buying your first item from a stranger online.”

Scott Belsky describes the objective:

“Just one example: At Behance, in the sign-up process for our service, we used to ask new Behance members to select their top three creative fields. New users took an average of 120 seconds to browse the list and select their top fields. We lost around 10% of new members at this particular step in the sign-up process. And so, we removed it from the sign-up process and resolved ourselves to capture this information later on during active use of the website. As a result, sign-ups went up. This is true for every online service or store. In the first 15 seconds, your visitors are lazy in the sense that they have no extra time to invest in something they don’t know. They are vain in that they want to look good quickly using your product. And they’re selfish in that, despite the big picture potential and purpose of what your service.”

  1. Critical mass magic moment–  a realization of core product value that is strong enough to ensure that customers are retained over the long term. To determine whether retention has reached a critical mass. A Bain study concluded:

“Depending on which study you believe, and what industry you’re in, acquiring a new customer is anywhere from five to 25 times more expensive than retaining an existing one. It makes sense: you don’t have to spend time and resources going out and finding a new client — you just have to keep the one you have happy. If you’re not convinced that retaining customers is so valuable, consider research done by Frederick Reichheld of Bain & Company (the inventor of the net promoter score) that shows increasing customer retention rates by 5% increases profits by 25% to 95%.”

  1. Reinforcing magic moments–  customers experiencing “core product value” reinforcement while experiencing the product. An example of this approach happens when Facebook sends you a message that someone has tagged you in a photo. The intent is to harness your curiosity to drive you back to core product value. For Facebook, reinforcing magic moments include adding a friend, liking or sharing a post, or updating your status within 24 hours of downloading the mobile app. Chamath Palihapitiya believes you can create loops that expose that core product value over and over again. he says:  “You have to work backwards from ‘what is the thing that people are here to do?’ ‘What is the A-ha moment that they want?” Schultz cautions:There’s a really fine line between removing friction and duping users. Tricking users hurts users. Adding friction hurts users.”
  2. Reactivation magic moments–  experiences that cause a customer to returns to a service after being inactive. “We missed you” emails are intended to trigger this type of Magic Moment.

How do you discover magic moments for your service? The objective is to look for correlations between usage, demographics and other behavior and retention. Examples of magic moments discovered by some businesses include adding seven friends in ten days or sending several thousand messages.

  1. “Virtuous loops are the flywheels that covert your users’ engagement into fuel to power your company forward. The strongest virtual loop is a network effect. Virtuous loops are really hard to create. Most products don’t have them.”

Virtuous circles and vicious circles are processes which can reinforce themselves through a feedback loop. Virtuous circles produce feedback that results in an increasingly positive outcome and vicious circles the inverse of that. One famous picture of a flywheel at work that you see a lot is this one below which describes one aspect of Amazon’s business:

AMZN loop

  1. “To get the flywheel spinning for a marketplace, you can’t sit in front of your computer and code. You need to pound the pavement more often than not and do things ‘that don’t scale’ to get the liquidity to help you start scaling.” 

Y Combinator co-founder Paul Graham describes what a business must do to start a business from a standing start better that just about anyone I have seen: 

The most common unscalable thing founders have to do at the start is to recruit users manually. Nearly all startups have to. You can’t wait for users to come to you. You have to go out and get them. A good metaphor would be the cranks that car engines had before they got electric starters. Once the engine was going, it would keep going, but there was a separate and laborious process to get it going.”

There is a big difference between doing things that don’t scale and being in a business that won’t eventually scale. Scalability is an important goal for any business but almost always in the beginning efforts  will be require that don’t scale well.  CEOs will go on sales calls, products will be created by hand, and many manual processes will be used at first. This info-graphic captures some non-scalable approaches that startups have used to bootstrap their flywheels past the cold stat problem.



  1. In non-transactional products, real value will be created when you create accruing benefits.  A product has accruing benefits if a user would say the more I use the product, the better it gets.’” 

Tavel uses examples from her experience at Pinterest to make this point:


10. “Mounting loss happens as a product becomes something you depend on, part of your identity, or a product in which you’ve accrued value of some sort (e.g., a following). ‘I’d have a lot to lose if I left this product’ is the claim to test.”


11. “The clearest way to understand a company’s engagement is to look at cohorts: number of weekly users completing the core action and percentage of weekly active users completing the action.”

There are an endless number of ways you can think about cohorts. One simple approach to illustrate what can be done is to look at a cohort’s behavior graphically: the number of days from acquisition is on the x axis, on the y axis is the percent of people who are still monthly active. With each day that passes some members of the cohort will stop using the service but the critical mass objective is to get the curve to “flat line” in a way that asymptotes with the x axis. What causes the line to stop dropping is often the network effects of more people being on the network creating a critical mass of value for the customer.

A retention curve that does not flatten and instead drops to zero may be caused by a product market fit problem or it may be the nature of the product (e.g. some video games). The single most important factor that drives growth is retention.


12. “Blogging is venture capital’s freemium model.”

There are at least three reasons that venture capitalist’s write for public consumption.

  1. It helps the writer think things through and find new solutions and ways to communicate ideas.
  2. It is way to give back- teaching is a way of giving back.
  3. It is way to use content marketing to source new deal flow.

It’s always nice when you get three things for the same amount of effort. It is even better if you can do well financially by doing something good for other people. Regardless of whether someone is first starting out or even if they are a big success writing is a way to turn thinking, words and effort into a substitute for capital. Content marketing is both more capital efficient and produces better results than traditional marketing. 

Writing is my way of giving back. Revenue is negative. It makes me feel good though, which is highly underrated.


The Hierarchy of Engagement, expanded https://medium.com/@sarahtavel/the-hierarchy-of-engagement-expanded-648329d60804

How To Create A Sticky Product Like Facebook and Evernote https://www.linkedin.com/pulse/accruing-benefits-mounting-loss-sarah-tavel 

How to build an enduring, multi-billion dollar business https://medium.com/@sarahtavel/how-to-build-an-enduring-multi-billion-dollar-business-hint-create-a-10x-product-recast-3527df2b8fcb 

Engagement Hierarchy: Core Actions  https://medium.com/@sarahtavel/engagement-hierarchy-core-actions-dd4f72042100 

Venture Capital’s Freemium Model http://www.adventurista.com/2009_11_01_archive.html 

Growth Hackers AMA https://growthhackers.com/amas/live-jun-14-ama-with-sarah-tavel-vc-at-greylock-partners

Lesson from Scaling Pinterest https://medium.com/@sarahtavel/five-more-lessons-from-scaling-pinterest-9c10fe97d325

The Mitochondria of Startups https://www.linkedin.com/pulse/mitochondria-startups-sarah-tavel

20 Minute VC https://www.producthunt.com/posts/the-twenty-minute-vc-sarah-tavel-partner-greylock-partners

Graham- Do Things that Don’t Scale http://paulgraham.com/ds.html

Bain:  https://hbr.org/2014/10/the-value-of-keeping-the-right-customers

Belsky: https://medium.com/positiveslope/the-first-15-seconds-9590d7dabc

Alex Schultz: https://venturebeat.com/2014/08/06/facebook-growth-chief-you-lose-users-if-you-try-to-trick-them/

Quora:  https://www.quora.com/Growth-Hacking/Growth-Hacking-How-do-you-find-insights-like-Facebooks-7-friends-in-10-days-to-grow-your-product-faster-1?utm_content=buffer2afcc&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer#!n=12

Categories: Uncategorized