Author Archives: Casey Winters

The Death Spiral of Startup CMOs

March 13th, 2017

I’ve met with a lot of CMO/VP Marketing types at startups. Generally speaking, they do not have a long shelf life. I was curious why so many startup CMOs leave the company after a short period of time, so I dug in a bit to figure out why. I found that it’s partially due to bad hiring practices for startup executives (which I have previously blogged about), but even more common was that big brand traditional marketers have a tough time “scaling down” into the scrappy, digital-first world of startups tactically.

The traditional marketing executive is more likely to be a brand marketer from a large company than an online marketing or growth person from a startup. These marketers have built careers deploying big budgets and leveraging a few channels that really matter to tell the brand story. The main one of course is still TV. TV is great for brand building, but startups cannot wait for multiple years of investment in brand advertising to start to pay off. They typically will run out of cash before they start seeing a return on the investment in TV. (This problem is particularly acute if you’re a startup that is live only in a few cities and gradually expanding across the country — i.e. most marketplaces.)

How does a marketing executive get a quicker read on effectiveness, especially if their startup doesn’t cover the entire country? One way I’ve seen marketers try to solve this problem quickly is to buy local TV brand advertising in a test market and compare to a control test market where they have not bought TV advertising. They measure if brand metrics improve in the paid market during a three month test. If brand metrics improve, it should be leading indicators for sales to improve.

On paper, this makes sense, but in practice, it is the first trigger of the CMO death spiral. The problem is twofold. First, the cost of local TV is 10x more expensive than national advertising. That premium is hard to make up. Second, TV brand advertising is generally not very effective and even harder to drive quick results. With a typical brand-based buy, ads will run at many different times during many different, popular shows. So even though many people will be watching and thus exposed to the brand, the viewers’ focus is on the show and not on a call to action.

In short, big brand marketing executives have used large budgets for a test and gained small bits of exposure for the company. That exposure usually doesn’t transit into brand lift or sales, and the founders quickly lose confidence in the CMO. From there, I’ve seen it is a matter of months before they either get fired or lose their budget, which will make them leave.

But its not all fire and brimstone! There are savvy ways for marketers with big brand backgrounds to be successful at startups. For one thing, start by testing national, direct response TV. This type of ad buy targets people who are watching TV because they are bored — not because they are super engaged with the show. In this scenario, ads run at off-peak times or on less popular networks and only cover a small percentage of the country. You get the data on the exact times and areas where the ads run. Then, you can watch for an immediate response in website/app traffic right after the ad, and track those users to see what they end up doing. Agencies that sell direct response typically offer data science services to measure the impact, and you can augment their reports with tools like Convertro or C3 Metrics that do it.

Grubhub and Seamless are both good on-demand startup case studies before they merged: Seamless bought national, direct response TV ads while the service only covered 20% of the U.S. population. Grubhub did it with only 40% coverage. Seamless then used all the website traffic data outside their coverage areas to determine where to launch next. At Grubhub, we saw a response immediately during some of these ads airing, and could actually calculate not just brand lift surveys, but CPAs due to transactions.

A broader lesson here is that traditional target marketing can be expensive as the people selling advertising have learned that traditional marketers will pay a premium for it. And while that math still works for a CMO of a CPG brand, it won’t typically work for a startup.

What is the the best way for a CMO to be successful and stick with your company for the long term? First, they need to find ways to reach the target market they want to reach in a cost effective way. (I go deeper into this topic on my post on remnant inventory.) Seamless reached an insane amount of people in New York with DRTV ads, so it didn’t matter that some people in Boise saw them, too. Secondly, CMOs need to work on other ways to add value to the business besides telling big brand stories. Depending on the business, that can include understanding customers better, building communities, performance marketing, or product driven growth.

Currently listening to Children of Alice by Children of Alice.

The Real Value of PR for Startups

February 23rd, 2017

Many startups get obsessed with press. Not just CEOs, but also employees, friends of employees, and investors. Press, like paid acquisition, can be a drug though. I’ll talk about some of the ways the startup ecosystem perceives press is bad, what it is actually good for, and some outliers.

Press Abuse
First, let’s talk about some of the abuse of press by startups in the past. Business insider has a great recap of some of the issues Evernote faced in the couple years before their CEO stepped down. I’ll highlight this quote in particular.

“There was a feeling that we were working on the wrong priorities,” a former employee said. “It was clear the motive was to just continually drum up press. They had no idea how to optimize and improve growth.”

From The inside story of how $1 billion Evernote went from Silicon Valley darling to deep trouble

Startups frequently correlate press with growth, and think that they need to create new stories to drive growth. Then they start changing their entire product roadmap to drive press instead of value to customers. Startups should not look for PR to drive growth. PR doesn’t usually drive growth. And when it does, it’s a bump, not an engine. In startups, you want to build growth engines.

This is arguably a better problem than the second way PR drives bad acting in the ecosystem: founders getting addicted to the attention of press. It feels great when there’s a picture of you in the Wall Street Journal and your mom sees it and your dad’s friends. And founders frequently want to create that feeling. It’s as if appearing successful is more important than being successful. This can be a dangerous pattern.

Another problem with press is that it helps competitors understand exactly what you’re doing. I’ve talked in the past of the advantage of being a silent killer and why seeking press attracts unneeded competitor attention. I won’t repeat that here.

What Press Is Good For
A common refrain I tell startups is that press is good for three things: investor interest, employee interest, and links for SEO. I’ll break those down a bit, and talk about why they matter. First, let’s talk about investors. Very few businesses make sense for venture capital, but when they do, they increasingly need larger amounts of it. The venture business has extreme cases of FOMO. One deal can make all the difference. So investors rely on a lot of signals, and the tech press is one important one they look at not only to find new business to research, but to also get comfort in their interest in certain businesses. When Grubhub was raising its Series C from Benchmark, we got written up about our iPhone app on Techcrunch. For Grubhub’s growth, it didn’t matter at all. But it looked great for our fundraising.

Likewise with investors, potential employees rely on press to hear about potential new places of employment and to feel comfortable they are making a good move. Recruiting outreach will have a lower conversion rate if the person has never heard of your company, or if the person’s network hasn’t heard of you when they ask around. This can be overcome, especially if you’re well connected or can tout amazing metrics privately (like WhatsApp), but my guess is WhatsApp always had a harder time recruiting than Snapchat.

Lastly, press is very helpful for SEO. SEO is about relevance and authority, and links from press are a great way to increase authority in the eyes of search engines. Not only can press drive a quantity of links, but the quality of links are probably the highest most startups will receive. So many people link to news publications that they have among the highest domain authority on the internet. Now, getting one of these links won’t make you rank #1, but it’s part of a healthy SEO strategy.

When Can Press Be an Engine of Growth?
Now that I’ve set an argument that press doesn’t usually drive growth directly, I want to talk about some outliers of when it does. Press can drive growth if a certain story about your startup is self-perpetuating. Then, the volume of articles written about you happens in such a regular cadence that readers eventually act on it. I have seen this happen in two major startups over the last ten years in Twitter and Uber. With Twitter, it received such a tremendous uptake from journalists that they kept writing stories about it. This drove their readers to try it, and it became a sustained vehicle for organic acquisition (unfortunately with a low activation rate, but that isn’t press’s fault, but the product). Uber’s growth via press was a little more sculpted. They were able to create a story about the business of Uber fighting against corrupt incumbents and governments that were preventing a better product from succeeding. It was almost a movement. For years this created a headline about Uber almost every week as they faced this resistance in almost every city in which they launched.

I want to highlight that these are anomalies, and it’s unlikely you will be able to create sustained growth directly from press in your company. But that doesn’t mean press is worthless. Just remember what it’s good for, and make it work for you for those interests.

Starting and Scaling a Growth Team

February 13th, 2017

I get a lot of questions about how to start and scale growth teams. Growth teams need to spend as much time thinking about how to scale themselves as they do on scaling the product. In this post, I’ll outline how I’ve seen growth teams start and evolve, define the ideal end state, and describe some of what you work on at each stage along the way.

The first question to ask is why you need to start a growth team. Why not just have traditional product and marketing teams? It’s a good question, and not well understood. Traditionally, product teams make the product, and marketing is in charge of getting people to try and continue using the product. Marketing can include traditional efforts like events, advertising, and PR, and perhaps some branding efforts and some online marketing. Of note, online marketers and traditional marketers have historically been very different, so depending on the manager, all of the budget and team allocation typically went to either online or traditional — not evenly split.

However, the best startups grow super fast not because of traditional marketing or online marketing, but because they tune the product to drive growth. This is why growth teams matter. Marketing teams typically don’t have access to the product roadmap to make changes to things that will impact SEO, conversion, optimization, or virality, nor the expertise to work with engineers and designers on making these changes. And product managers often prioritize new product features over engineering that will drive more people to the existing value. Growth teams are the connective tissue between product, marketing and engineering.

Starting a Growth Team: Focus On One (Easier) Problem
Goal: Improve one area important to the growth of the business
Team: PM, Designer, Engineer(s), Analyst/Data Scientist (sometime PM or engineer assumes this role as well early on)

It can be overwhelming to consider all the problems you need to solve. When you start a growth team, do not to try to solve all those problems at once. A growth team should start by going deep on one carefully chosen problem:

  • Pick a problem the company is not currently working on (to prevent turf wars)
  • Pick one of the easier — not harder — problems to improve (to build up credibility and drive toward early wins)

Since growth teams are typically treated with a healthy dose of skepticism by other teams when they start, carefully choosing your first area of focus will help maximize the growth team’s chances for success. Let’s say a growth team starts with a hard problem: Trying to increase activation of new users. It takes a month to measure increases in activation rates, and it has one of the smaller sample sizes in the traditional growth area. You run an experiment. It doesn’t work. You run a second, and another month goes by, and it still may or may not work. Now two months have gone by without a growth team win. People on the growth team are questioning if this is the right team to be on, and other teams in the company begin to question your purpose. (Hat tip to Andy Johns from whom I am blatantly stealing this example).

There are common growth problems where you can likely show quick progress. One is conversion optimization. Another is returning users via email and notifications. Still another is improving referrals or virality. These areas allow you to run multiple experiments quickly because the metrics will move immediately (they don’t take time to measure like activation), and because they have high sample sizes (all new users or all existing users opted in to communication).

Last note: When starting a growth team, it’s important that they sit together and not with their functional team members. The collaboration between the different skill sets and short feedback loops is how the magic happens.

Growing a Growth Team: Own the Growth Funnel (And Some Real Estate)
Goal: Improve metrics across the growth funnel
Team: PMs, Designers, Engineers, Data Scentists/Analysts

As a growth team grows, the team can start looking at multiple problems. These problems are typically across the AARRR framework by Dave McClure, (however revenue is frequently a separate team in an ad supported business.) The growth team separates into sub-teams who have their own meetings, and entire growth team meetings become less frequent. PMs, designers, and analysts may work on multiple teams. At this point, the growth team assumes ownership of certain areas of the product, including (but not limited to):

  • The logged out experience, which can involve both SEO and conversion
  • All emails and notifications (sometimes coordinating with marketing)
  • The onboarding flow, sharing flows, et al.

Owning these areas creates easy swim lanes between teams and prevents turf was with the core product team. Growth leaders have to determine how to allocate people to work on the most impactful problems across the growth stack, and if a sub-team does not have enough support to make an impact, the growth leader should consider whether to support that problem at all. Growth teams also have to own goal setting, which means understanding historical performance and setting absolute goals.

At this point, teams start to work on both improvements in infrastructure and experiments. For example, at Pinterest, I was managing both the acquisition and retention teams. The retention team was struggling to grow because of an unwieldy email infrastructure. The infrastructure was built to support triggered, social notifications, but our strategy had evolved to more personalized, content-based recommendations. These jobs were taking days to run and would send automatically when complete, even though someone might have received a social notification five minutes before. So we spent nine months rebuilding it to allow us to scale more effectively. The retention team saw a step change in its performance and actually hit its Q1 goal in the middle of January after it was complete.

Evolving a Growth Team: A Special Forces Team Operating Across Borders
Goal: Reduce friction across the product that prevents people from connecting to product’s core value
Team: PMs, Designers, Engineers, Data Scentists/Analysts, Researchers, Marketers, Operations

As the growth team continues to grow, it involves more stakeholders and expands its scope. Instead of having clear areas of focus separate from the core product, the team shifts to analyzing the entire product and trying to figure out the biggest obstacles that are preventing the company from growing faster. At this point, the growth team has built up enough credibility that it doesn’t fight turf wars — it is laser focused on finding and solving the biggest problems that prevent people from connecting to the current value of the product.

Usually, the problems at this stage are deep inside product features and go beyond improving SEO, signup, viral and on-boarding flows, and sending the right emails and notifications (though growth teams still work on those too). The growth team is now identifying the friction that prevents people connecting to the core value elsewhere in the product. A lot of this work is simplifying product flows or building country-specific optimizations for slower growth countries. Growth teams can also become service organizations to marketing initiatives around this time, and I’ll talk about that in another post.

At Pinterest, we migrated to this phase rather recently. As we started to look at friction in the entire product, we saw in qualitative research that new users were getting bombarded with so many concepts it confused them: what a Pin is, where they come from, how to add your own Pins, what a board is, group boards, profiles, etc. The growth team made a list of the core things new people needed to know to get excited and start to build an understanding of the service. Then, we started experiments to remove everything else from the new user experience and only introduce it back once the core concepts were clearly learned. This helped improve activation rates.

What I described above is an amalgam of what I’ve seen in the market that works best. Of course, you should always tweak your approach based on the talent of your team and your company’s culture. I would love to hear what you think of this approach, and any additional insights you think would be useful to share with other growth leaders.

This posted originally appeared on the Greylock blog.

Currently listening to 28 by Aoki Takamasa + Tujiko Noriko.

Three Mistakes Integrating Growth and Design Teams and How to Address Them

January 23rd, 2017

While I encourage growth teams to start with a dedicated designer, that doesn’t usually happen. Usually, growth teams scale with just engineering and product, and as they scale to a certain size start to eventually “earn” a dedicated designer from the organization. This news makes growth teams extremely happy for a short period of time, but pretty quickly starts to create problems and culture clash. I’ll talk about what happens, and some ways I’ve found to solve these issues. Note: You can replace growth with marketing, and just about the same thing happens in these scenarios.


When you hear you’re getting a dedicated designer, you’re happier than Jim Carrey was back when he had a career.

Problem #1: Team Control AKA The Two Scenarios
One of two scenarios usually emerges when designers join a growth team. In the first scenario, the engineers or PM or marketing person starts with, “I’m so glad you’re here. I need this done and this done and this done for an experiment…tomorrow. No, we don’t need research to understand the problem. Don’t worry. I’ll only ship it if the metrics increase.” The designer realizes the growth team didn’t want a designer. They wanted a pixel monkey to just do what they say, not ever use their brain.


I got 99 problems, but a designer using their brain ain’t one.

The other scenario is just as bad. In this scenario, I call it the designers “moving in”. I liken this to the scenario when a significant other moves into your place and brings way more things than you own to move in. “Don’t worry, I got better furniture and books, and, oh, we’re going to have to get rid of those curtains.”

The design version of this is something like “I’m going to completely rethink all our strategies. I need to spend three months just doing research and then at least double that for design concepts. That home page really needs to change though.” Engineering thinks “this isn’t what we need. These people don’t even know what they’re talking about.” Engineering gets kicked out of the process of figuring out what to work on, and since the design process has no deadlines on it, engineering begins to have nothing to work on.


I tried to find a picture of people dressed in all black moving in, but the best I could do was chambray.

In both scenarios, instead of designers, engineers, product managers, and marketers trying to unify to form one team that leverages all of their strengths, one team tries to dominate the direction. What we did at Pinterest to try to solve this problem was design a process where the teams, specifically design and engineering, are jointly responsible for problem definition and solutions. Here’s what it looks like:

Project Kickoff:
Attendees: PM (leader), engineering lead, design lead, engineer and designer that will be working on the project

Goals:

  • Define the problem we’re trying to solve
  • Any existing ideas on how we might solve it
  • Any previous attempts to solve the problem
  • What metrics we need to inform potential solutions
  • How will we tell if we’re successful

Output:

  • Notes emailed to broader sub-team
  • Slack channel created just for the project, which includes notes and all future communication for the project

Project Brainstorming:
Attendees: engineer and designer on the project, PM (optional so as to prevent their calendar from being a bottleneck)

Goals:

  • Produce multiple potential solutions to the defined problem
  • Prepare those for feedback from attendees of kickoff
  • Cover additional measurement needed for directions that have been chosen e.g. how many people click top right login button

Brainstorm Review:
Attendees: PM, engineering lead, design lead, engineer and designer working on the project (leaders)

Goals:

  • Feedback on concepts from brainstorm
  • Choose one or two directions for experiment

Experiment Launch:
Attendees: engineer and designer working on the project (leaders), PM

Output:

  • Experiment doc that team agrees reflects what we’re testing and why
  • QA and approval over Slack from every team member that experiment is ready to launch to 1%
  • Note: Ramp ups of experiment communicated in Slack channel

Experiment Review:
Attendees: PM (facilitator), engineering lead, design lead, engineer and designer that worked on the project (leaders)

Goals:

  • Determine if you gathered data on key questions you needed to answer for the experiment
  • If yes, what does the data say?
  • If not, how do we get the data?

Output:

  • Ship/kill/iterate decision
  • Emailed notes to the entire team of what happened and why
  • Updated experiment doc on what happened
  • Updated Slack channel
  • Designer brings ship experiment candidates to design manager/lead for any feedback before shipping
    • Problem #2: Design Best Practices
      Designers bring with them a history of best practices from working on other design teams and schooling/training. It can be a culture shock for a designer joining a growth team and seeing little of these being followed. Designers will usually respond to this environment by trying to educate on many things the growth team is managing are designed “wrong” and how they need to be changed. These recommendations based on best practices are usually rejected by engineers on the team as it contradicts what they’ve seen in the data.


      A growth engineer after he hears about a marketing or design best practice.

      Design best practices were created with good purpose, but they quickly become inferior to AB testing in an environment where a direction can be put in front of users and its value quantitatively determined in days or weeks. This doesn’t mean you should AB test every design decision, but it does mean a designer can’t put their foot down by saying something is a best practice. In growth, the users being tested on determine what’s best.

      Another issue with design best practices being applied to growth is that design best practices are usually suited entirely for building user value only. The growth team in many cases needs to trade off user value and business value, or short term user value vs. long term user value. For example, someone can come to my site, and I can offer a great experience. The user then leaves and never comes back. Growth might decide to let someone get a preview of that experience, then ask the user to sign up so they can personalize and deliver more value. The result is that some people will not sign up (creating a worse user experience) and some will sign up, allowing the site to create a better user experience that day and over the long term. This accrues value for the business as well.

      Designers usually feel threatened by this environment initially. They see testing as a threat to their expertise. What you have to do is teach them to use it as a tool to get closer to user feedback at scale and be more efficient with their time. Most design work is wasted because it is spent on a direction that later proves to be flawed. With AB testing, a design can get quick feedback on an idea, validate the direction, then spend the time making it amazing once they know it’s worth that effort.

      That said, while this user feedback at scale is good at showing what users do, it’s not great at explaining why. So, along with spending time getting designers comfortable with testing, growth teams need to start doing more regular qualitative research as well. Designers will usually volunteer to do it themselves if there isn’t the right person to staff for it. Some PM’s are comfortable doing it as well. Engineers and engineering managers can be resistant to spending time watching the sessions, so the first couple you schedule have to be on critical areas you feel there is some understanding the team is missing by only looking at the metrics.

      Once designers get a little more seasoned on growth, you should also work with them to create the new best practices for growth to make the design and engineering process move faster. They will look very different than what designers recommended at the beginning, and be backed by data. Along with this process, I think it’s important to start creating user experience goals for your growth team at this time. These will be components that may not affect the metrics, but ensure a quality and consistent user experience. At Pinterest, we made compliance to design spec a requirement to ship after an experiment is validated, and said our top five flows had to be completely audited for quality user experience. This is a worthy compromise with design to show you actually care about the users and not just the business, a common complaint.

      Problem #3: Growth Onboarding
      Sometimes, you want to make sure you just nail the basics. Most growth designers are joining a new team in a discipline they’ve never worked on before, yet they don’t receive any onboarding as to what growth is and why it’s different. This is also an issue with other new growth team members. It’s like being dropped into a foreign country, not knowing the language, having no one help you, and being expected to contribute to the culture. Like France (okay, I’ve actually been to France, and the people were surprisingly welcoming even though they’re not known for being so). What happens is designers get frustrated and churn from the team.

      It’s critical that every new team member, especially designers, go through a growth team onboarding process. The first thing you do is state the purpose for your growth team, why it’s different from other product teams, and why it’s important. At Pinterest, I would say that while other product teams create value or improve the value of the product, growth teams focus on connecting people to the current value of the product and reducing the friction that prevents that connection. This is important because it’s not enough to build a good product. If no one knows about it, it will die. If it’s too hard to uncover its value, it will also die.

      What we did at Pinterest was create a history of major growth projects that were successful, walked through them, and explained why they were successful. That led into some of our team principles. We also spent a lot of time educating new people on the metrics we use and why. You can’t be expected to design winning experiences if you don’t understand the major criterion for success we’ll be using.

      After onboarding, instead of starting designers on large, complicated projects, it’s important to start them on a well scoped, smaller project, preferably with an experienced PM and engineer, hopefully patient ones. These should still be projects that are worth doing, but not large projects. We had a designer start on growth at Pinterest, and we immediately put her on one of the most strategic, long term investments for the team. While she did great work there, after a few months, she did a smaller side project redesigning our mobile web home page. The conversion rate increased, and we shipped it. She said, “Guys, this is the first time my design increased the metrics!” She was beaming. You want to get new growth designers to that moment as soon as possible.

      Lastly, once onboarded, you want designers contributing growth ideas as soon as possible. I like the idea of forcing people to bring one new idea to the team per week. I believe this is a muscle that needs to be developed via practice. New entrants to the team (design or otherwise) will typically propose bad ideas for a while. That’s okay. The trick is to provide a framework for generating ideas that makes new team members think about elements that typically make for good growth ideas, give feedback on the ideas submitted, and have the ideas be submitted as metrics wins or user experience wins. The Pinterest template focused on :

      • How many people would see this experience if it were built? This is usually by far the most important criterion for a successful growth idea. Any experience has to be seen by a lot of people to have a big possible impact on the business.
      • Has the company tried something in this area before? If so what were the results? This helps us make sure we use any previous learnings to avoid making the same mistakes. Just because we tried something before doesn’t mean it’s a bad idea though as growth environments change frequently.
      • How much effort is required for this idea? On growth, we also try to do the least amount of work to validate an idea is worth it.

      There is no reason designers can’t be key contributors to a growth team, but expecting it to happen automatically is usually a recipe for failure. I hope some of these tips can help you create a thriving cross-functional growth team with all the right disciplines involved. If you’ve have any other issues integrating design into your team, I’d love to hear about them in the comments.

      Currently listening to Sorry I Make You Lush by Wagon Christ.

Don’t Become a Victim of One Key Metric

October 12th, 2016

The One Key Metric, North Star Metric, or One Metric That Matters has become standard operating procedure in startups as a way to manage a growing business. Pick a metric that correlates the most to success, and make sure it is an activity metric, not a vanity metric. In principle, this solves a lot of problems. It has people chasing problems that affect user engagement instead of top line metrics that look nice for the business. I have seen it abused multiple times though, and I’ll point to a few examples of how it can go wrong.

Let’s start with Pinterest. Pinterest is a complicated ecosystem. It involves content creators (the people who make the content we link to), content curators (the people who bring the content into Pinterest), and content consumers (the people who view and save that content). Similar to a marketplace, all of these have to work in concert to create a strong product. If no new content comes in, there are less new things to save or consume, leading to a less engaging experience. Pinterest has tried various times over the years to optimize this complex ecosystem using one key metric. At first, it was MAUs. Then it became clear that the company could optimize usage on the margin, instead of very engaged users. So, the company then thought about what metric really showed a person got value out of what Pinterest showed them. This led to the creation of a WARC, a weekly active repinner or clicker. A repin is a save of content already on Pinterest. A click is a clickthrough to the source of the content from Pinterest. Both indicate Pinterest showed you something interesting. A weekly event made it impossible to optimize for marginal activity.

There are two issues at play here. The first is the combination of two actions: a repin and a click. This creates what our head of product calls false rigor. You can do an experiment that increases WARCs that might actually trade off repins for clicks or vice versa and not even realize it because the combined metric increased. Take that to the extreme, and the algorithm optimizes clickbait images instead of really interesting content, and the metrics make it appear that engagement is increasing. It might be, but it is an empty calorie form that will affect engagement in a very negative way over the long term.

The second issue is how it ignores the supply side of the network entirely. No team wants to spend time on increasing unique content or surfacing new content more often when there is tried and true content that we know drives clicks and repins. This will cause content recycling and stale content for a service that wants to provide new ideas. Obviously, Pinterest doesn’t use WARCs anymore as its one key metric, but the search for one key metric at all for a complex ecosystem like Pinterest over-simplifies how the ecosystem works and prevents anyone from focusing on understanding the different elements of that ecosystem. You want the opposite to be true. You want everyone focused on understanding how different elements work together in this ecosystem. The one key metric can make you think that is not important.

Another example is Grubhub vs. Seamless. These were very similar businesses with different key metrics. Grubhub never subscribed entirely to the one key metric philosophy, so we always looked at quite a few metrics to analyze the health of the business. But if we were forced to boil it down to one, it would be revenue. Seamless used gross merchandise volume. On the surface, these two appear to be the same. If you break the metrics down though, you’ll notice one difference, and it had a profound impact on how the businesses ran.

One way to think of it is that revenue is a subset of GMV, therefore GMV is a better metric to focus on. Another way to think of it is the reverse. Revenue equals GMV multiplied by a commission rate for the marketplace. So, what did they do differently because of this change? Well, Seamless optimized for orders and order size, as that increased GMV. Grubhub optimized for orders, order size, and average commission rate. So, while Seamless would show restaurants in alphabetical order in their search results, Grubhub sorted restaurants by the average commission we made from their orders. Later on, Grubhub had the opportunity to test average commission of a restaurant along with its conversion rate, to maximize that an order would happen and maximize its commission for the business. When GrubHub and Seamless became one company, this was one of the first changes that was made to the Seamless model as it would drastically increase revenue for the business even though it didn’t affect GMV.

This is not to say that revenue is a great one key metric. It may be better than GMV, but it’s not a good one. Homejoy, a service for cleaners, optimized for revenue. Their team found it was easier to optimize for revenue by driving first time use instead of repeat engagement. As a result, their retention rates were terrible, and they eventually shut down.

Startups are complicated businesses. Fooling anyone at the company that only one metric matters oversimplifies what is important to work on, and can create tradeoffs that companies don’t realize they are making. Figure out the portfolio of metrics that matter for a business and track them all religiously. You will always have to make tradeoffs between metrics in business, but they should be done explicitly and not hide opportunities.

Currently listening to A Mineral Love by Bibio.

The Right Way To Set Goals for Growth

August 16th, 2016

Many people know growth teams experiment with their product to drive growth. But how should growth teams set goals? At Pinterest, we’ve experimented with how we set goals too. I’ll walk you through where we started, some learning along the way, and the way we try to set goals now.

Mistake #1: Seasonality
Flash back to early 2014. I started product managing the SEO team at Pinterest. The goal we set was a 30% improvement in SEO traffic. Two weeks in we hit the goal. Time to celebrate, right? No. The team didn’t do anything. We saw a huge seasonal lift that raised the traffic without team interference. Teams should take credit for what they do, not for what happens naturally. What happens when seasonality drops traffic 20%? Does the team get blamed for that? (We did, the first time.) So, we built an SEO experiment framework to actually track our contribution.

Mistake #2: Only Using Experiment Data
So, we then started setting goals that were entirely based on experiment results. Our key result would look something like “Increase traffic 20% in Germany Q/Q non-seasonally as measured by experiments” with a raw number representing what a 20% was. Let’s say it was 100K. Around this time though, we also started investing a lot in infrastructure as a growth team. For example, we created local landing pages from scratch. We had our local teams fix a lot of linking issues. You can’t create an experiment on new pages. The control is zero. So, at the end of the quarter, we looked at our experiments, and only saw a lift of around 60K. When we look at our German site though, traffic was up 300%. The landing pages started accruing a lot of local traffic not accounted for in our experiment data.

Mistake #3: Not Factoring In Mix Shift
In that same quarter, we beat our traffic and conversion goals, but came up short on company goal for signups. Why did that happen? Well, one of the major factors was mix shift. What does mix shift mean? Well, if you grow traffic to a lower converting country (Germany) away from a higher converting country (U.S.), you will hit traffic goals, but not signup goals. Also, if you end up switching between page types you drive traffic to/convert (Pin pages convert worse than boards, for example), or if you switch between platform (start driving more mobile traffic, which converts lower, but has higher activation), you will miss goals.

Mistake #4: Setting Percent Change Goals
On our Activation team, we set goals that looked like “10% improvement in activation rate.” That sounds like a lofty goal, right? Well, let’s do the math. Let’s say you have an activation rate of 20%. What most people read when they see a 10% improvement is “oh, you’re going to move it to 30%.” But that’s not what that means. It’s a relative percent change, meaning the goal is a 2% improvement. With this tactic, you can set goals that look impressive that don’t actually move the business forward.

Mistake #5: Goaling on Rates
Speaking again about that activation goal, there’s actually two issues. The last paragraph talked about the first part, the percent change. The rate is also an issue. An activation rate is two numbers: activated users / total users. There are two ways to move that metric in either direction: change activated users or change total users. What happens when you goal on rates is you have an activation team that wants less users so they hit their rate goals. So, if the traffic or conversion teams identify ways to bring in more users at slightly lower activation rates, the activation team misses their goal.

Best Approach: Set Absolute Goals
What you really care about for a business like Pinterest is increasing the total number of activated users. At real scale, you also care about decreasing churned users because for many business re-acquiring churned users is harder than acquiring someone for the first time. So, those should be the goals: active users and decrease in churned users. Absolute numbers are what matter in growth. What we do now at Pinterest is set absolute goals, and we make sure we account for seasonality, mix shift, experiment data as well as infrastructure work to hit those goals.

Currently listening to RY30 Trax by µ-ziq.

When Experimenting, Know If You’re Optimizing or Diverging

June 21st, 2016

When I first joined the growth team at Pinterest, in one of the early syncs a fellow PM presented a new onboarding experience for iOS. The experience was well-reasoned and an obvious improvement to how we were currently educating new users on how to get value out of the product. After the presentation, someone on the team asked about how the experiment for it was going to work. At this point, the meeting devolved into debate about how they would break out all these new experiences into chunks and what each experience needed to measure. It was liking watching a car wreck take place. I knew something went seriously wrong, but I didn’t know how to change it. The team had an obviously superior product experience, but couldn’t get to a point where they were ready to put it in front of users because of a lack of experimentation plan to measure what would cause the improvement in performance.

After I took over a new team in growth, I was asked to present a product strategy for it to the CEO, and the heads of product, eng, and design. Instead of writing about the product strategy, I wrote an operational strategy first. It was just as important for the team and this group of executives to know how we operated as it was to know what we planned for our operation. One of these components of the operational plan was a mix of optimization vs. divergent thinking. Growth frequently suffers from a local maxima problem, where it finds a playbook that works and optimizes it until it receives more and more marginal gains, making the team less effective, and hiding new opportunities not based on the original tactic. Playbooks are great, and when you have one, you should optimize every component of it for maximum success. But optimization shouldn’t blind teams to new opportunities. In the operational strategy, the first section was about questioning assumptions. If an experiment direction showed three straight results of more and more marginal improvements, the team had to trigger a project in the same area with a totally divergent approach.

This is where our previous growth meeting ran into trouble. When tackling a divergent approach as an experiment, you will either start in one of two directions. One is that the team steps back and comes up with a much better solution than what it was optimizing before, and are pretty confident it will beat the control. The other is the team determines a few different approaches, and they are not sure which ones if any will work. In both cases, it’s impossible to just change one variable from the control to isolate the impact of changes in the experience.

In this scenario, the right solution is not to isolate tons of variables in the new experience vs. the control group. It is to ship an MVP of the totally new experience vs. the control. You do a little bit more development work, but get much closer to validation or invalidation of the new approach. Then, if that new approach is successful, you can look at your data to see which interactions might be driving the lift, and take away components to isolate the variables after you have a winning holistic design.

If you’re optimizing an existing direction, isolating variables is key to learning what impacts your goals. When going in new directions though, isolating variables sets back learning time and may actually prevent you from finding the winning experience. So, when performing experiments, designate whether you are in optimization mode or divergent mode, and pick the appropriate experimentation plan for each.

Currently listening to Oh No by Jessy Lanza.

The Present and Future of Growth

June 7th, 2016

Quite a few people ask me about the future of growth. The idea of having a team dedicated the growth in usage of a product is still a fairly new construct to organizations. More junior folks or people less involved with growth always ask about the split between marketing and growth. More senior folks always ask about the split between growth and core product. Growth butts heads with both sides.

Why do more senior folks tend to turn to the difference between core product and growth than marketing? For this I’ll take a step beck. Now, I’m a marketer by trade. I have an undergraduate degree in marketing and an MBA with a concentration in marketing. So I consider everything marketing: product, growth, research, and I’ve written about that. I used to see what was happening in tech as marketing’s death by a thousand cuts. I now more so see it as marketing’s definition has gotten so broad and each individual component so complicated that it can by no means be managed by one group in a company.

So if marketing is being split into different, more focused functions, growth teams aren’t really butting heads with the remaining functions that are still called marketing over responsibilities like branding. They are butting heads with the core product team over the allocation of resources and real estate for the product.

So how do growth team and core product teams split those work streams today, and what does the future look like? The best definition I can give to that split for most companies today is that growth teams focus on getting the maximum amount of users to experience the current value of the product or removing the friction that prevents people from experiencing current value, and core product teams focus on increasing the value of the product. So, when products are just forming, there is no growth team, because the product is just beginning to try to create value for users. During the growth phase, introducing more people to the current value of the product becomes more important and plays in parallel with improving the value of that core product. For late stage companies, core product teams need to introduce totally new value into the product so that growth isn’t saturated.

My hope is that in the future, this tradeoff between connecting people to current value, improving current value, and creating totally new value is all managed deftly by one product team. That team can either have product people naturally managing the tradeoffs between these three pillars, or three separate teams that ebb and flow in size depending on the strategic priorities of the organization. All three of these initiatives – connecting people to current value, improving current value, and creating new value – are important to creating a successful company, but at different stages of a company, one or two tend to be more important than another.

We should evolve into product organizations that can detect which of these three functions adds the most value at a particular point in time naturally, fund them appropriately, and socialize the reasons for that into the organization so these different functions don’t butt heads in the future. I believe that is the product team of the future. I now believe this is more likely than marketers evolving to manage branding, research, performance marketing, and product effectively under one organization.

Currently listening to Good Luck And Do Your Best by Gold Panda.

If It Ain’t Fixed, Don’t Break It

April 11th, 2016

Frequently, products achieve popularity out of nowhere. People don’t realize why or how a product got so popular, but it did. Now, much of the time, this is from years of hard work no one ever saw. As our co-founder at GrubHub put it, “we were an overnight success seven years in the making.” But sometimes, it really just does happen without people, inside or outside the company, knowing why. Especially with social products, sometimes things just take off. When you’re in one of these situations, you can do a couple of things to your product: not change it until you understand why it’s successful now, or try to harness what you understand into something better that fits your vision. This second approach can be a killer for startups, and I’ve seen it happen multiple times.

Let’s take two examples in the same space: Reddit and Digg. Both launched within six months of each other with missions to curate the best stories across the internet. Both became popular in sensational, but somewhat different ways, but Digg was clearly in breakout mode.

What happened after the end of that graph is a pretty interesting AB test. Digg kept changing things up, launching redesigns and changing policies. Some of these might have been experiments that showed positive metric increases even. Reddit kept the same design and the same features, allowing new “features” to come from the community via subreddits, like AMA. By the launch of Digg’s major redesign in August of 2010 (intended to take on elements from Twitter), Reddit exploded ahead of Digg.

This is what the long term result of these two strategies look like. Digg is a footnote of the internet, and Reddit is now a major force.

Now, neither of these companies are ideal scenarios. The best option in the situations these companies found themselves in is to deeply understand the value their product provides and to which customers, and to completely devote your team to increasing and expanding that value over time. But, if you can’t figure out exactly why something is working, it is better to do nothing then to start messing with your product in a way that may adversely affect the user experience. This has become one of my unintuitive laws of startups: if ain’t fixed, don’t break it. If you don’t know why something is working (meaning it’s fixed and not a variable), do nothing else but explore why the ecosystem works, and don’t change it until you do. If you can’t figure it out, it’s better to change nothing like Reddit and Craigslist than to take a shot in the dark like Digg.

Currently listening to Sisters by Odd Nosdam.