How To Launch A Product or Feature To Maximize Growth

Written By

Superhuman was founded in 2015 (3ish years ago at the time of writing this). From their landing page, they are building “The Fastest Email Experience Ever Made.” To this day, you can still not sign up and instantly gain access to their product. Yet, they have received more press, word of mouth, and funding than 95%+ of other products. Why? Because they've done the exact opposite of what most do for product and feature launches. Let me explain...

Right now someone, or an entire team, inside your company is probably pouring a ton of energy and thought into a new product or feature launch. That plan might includes things like:

  • Releasing to 100% of user base and make it available to the public

  • Sending an email/notification to the entire marketing or user database

  • Launching on and getting to the top of Product Hunt, Hacker News, etc

  • Lining up Techcrunch and other press pieces

  • Writing a “launch” blog post on Medium announcing the product/feature

This effort is done with the best intentions. The hope is create the spark that leads the product or feature to escape velocity. But, what actually happens in is the opposite. Most launches create an incredible amount of head wind to growing and kills your chances for long term sustainable growth.

In August 2017, a social quiz app aimed at teens and high schoolers named TBH (To Be Honest) skyrocketed to #1 in the app store in just a couple months. Facebook quickly acquired them for $100M in October 2017. Nine months after being acquired, Buzzfeed reported on an internal memo from the TBH founders about how to execute product launches. The key quote is:

It is critical to design a process that allows you to launch vastly different product experiences within specific communities so your product can reach critical mass.”

There is so much wisdom in this quote, that it spurred the idea for my presentation at the ProfitWell Recur conference. In this post I will:

  • Break down the four reasons why most product/feature launches hurt growth rather than help it.

  • Describe a step by step process on how to launch.

  • Go through a couple recent product examples who are using this process.

Why Most Product/Feature Launches Hurt Growth

Reason one: you kill your most valuable+important channel

When evaluating the growth of companies I work on and invest in, I ask - If we turned off all acquisition spend/effort today, would we still grow? We will likely see a dip in the short term, but it should eventually level out and then start to go back up. That growth might be extremely slow, but if we saw that it means that we haveflat retention curves + solid word of mouth. Those two things provide the solid foundation to layer on growth efforts.

There are three common thoughts about Word of Mouth that are incorrect in my opinion:

  • Word of Mouth just happens.

  • You don't have control over Word of Mouth or are unable to influence it.

  • To create Word of Mouth, you just need to build a great product.

Let's break down what drives word of mouth.

Word of Mouth Loop - Brian Balfour.jpeg

As all things compelling in growth, word of mouth is a loop.

  1. A new user/customer starts using the product.

  2. That product greatly exceeds expectations.

  3. As a result, that person tells others.

  4. Loop repeats.

The key in this loop is step 2 - the product greatly exceeds expectations. In other words, there is a large positive delta between your product experience, and the alternative.

Product Expectation vs Alternative - Brian Balfour.jpeg

If that delta is incremental, the user will likely not be compelled to tell others about the experience. One part of creating a non-incremental delta is building a great product. But that is only half the equation. The other half, is who you expose the product too. Expose a great product to an audience it isn't built for, and you end up with an incremental or negative delta. This is exactly what most product/feature launches do.

The issue is that most product/feature launches don't have any controls around who they are exposing the product to, even though that version of the product/feature was hopefully built with a very specific target audience in mind. When this happens you end up with a lot user/customers that it wasn't build for who either have an incremental or negative delta experience.

Rationally, we think that users/customers will know that the product/feature hasn't built for them. But in reality, they don't and just form a negative opinion. In the end, you end up with a greater population of people who have neutral/negative things to say than positive which kills your Word of Mouth loop.

Don't Launch - Recur 2018 - Brian Balfour.019.jpeg

Reason two: You slow down validating your product/feature hypothesis

The goal of any product team is to validate their product or feature hypothesis as quickly as possible. There are two ways to get validation around your hypothesis, qualitative and quantitative.

For qualitative feedback, we might be looking at NPS or other information. Since most product launches are untargeted, we end up with a bunch of noisy responses that are end up being hard to sift through.

On the quantitative side, we are looking for flat retention curves for the product or feature. But once again, the typical un-targeted product launch bites us in the ass. With a much wider audience than what we have built for, our retention curves are likely to show an unhealthy trend.

The point is, if we put random garbage in, we are going to get garbage out. To validate our hypotheses we are going to have to dig through that garbage. That will make our insights slower and ultimately less accurate and effective.

Reason three: You deplete your ability to get your audience's future attention

In the Reforge Growth Series, we spend a week on thinking through the qualitative side of growth - understanding User Psychology. One of the many frameworks we dive deep on is Darius Contractor's (Head of Growth at FB Messenger), “Psych” Framework. A core part of that framework is thinking about your users attention and motivation like a fuel tank from empty to full. There are things that drain the full tank, and things that add fuel to the tank.

This hypothetical fuel tank is more fragile than ever. It is much easier to drain the fuel tank than to fill it. With everyone being over loaded with email, push notifications, etc, we are much quicker to start ignoring. That is exactly what most product or feature launches do. They notify the entire user base to try and jump start adoption, but what you are really doing is napalming your audiences attention. This is going to make it incredibly difficult to get your audience's future attention when you need it.

Reason Four: You fail to set yourself up for continued success

Our ultimate goal is to create long term sustainable growth for our product or feature. That comes from building a sustainable growth model based on compounding loops.

Unfortunately, we tend to think more about the launch than we do the machine that will keep growth going after the launch. Every product and feature needs a kick start (the launch), but most of the energy should be spent on building the engine that will keep on running. The launch is a means to an end, not the end itself.

How to Launch A Product or Feature To Enable Growth

So what to do instead? Let's return to that quote from the TBH founders:

“It is critical to design a process that allows you to launch vastly different product experiences within specific communities so your product can reach critical mass.”

There are are a few components in this quote:

Repeatable Process + Vastly Different Product Experience + Specific Community = The Chance To Reach Critical Mass. This looks like a five step loop.

Repeatable Product Launch Process - Brian Balfour.jpeg

Step one: Scope

The first step is to do the exact opposite of what most product launches do, scope your target audience way down to who you have initially built for. Specifically spell out a hypothesis of who that is. Most product launches start with “What are the places we can get attention?” The better way is to start with, who specifically are we trying to reach with this v1? Then, you can think about where those people “live.”

Step Two: Figure Out Access

Once we have our initial definition, we can then come up with a bunch of ideas on how to access that audience. This will differ based on what the initial hypothesis is, if it is a product vs feature launch, and if we are working with an existing user/customer base or launching something brand new. Email, Paid, Press, Medium, Product Hunt, Hacker News, Referrals, etc. This step doesn't matter as much as step one and step three.

Step Three: Filter

You'll never be able to target your audience hypothesis perfectly with any marketing mechanism. So we need to think about how take initial interest, and filter it to our audience hypothesis. There are a number of ways to do this, but the main ones would be:

  1. Existing Usage Data - If you are working with an existing audience, you should have a good amount of data on them. Not just who they are, but what they have and have not done in your product. This is all valuable data to filter for your hypothesis.

  2. User Submitted - Data submitted by the user. Key is to ask the right questions or collect the right data that will let you filter effectively. (See example in the section below).

  3. Passive Data - There are a lot of tools like Clearbit which help us append data to understand more about who they are.

Step Four: Search for your success signal

As you let people through the filter and use the product , you look for success signals that validate your product or feature hypothesis. I think about them in three levels, each level requiring more volume, data, and time to get:

  1. Qualitative - NPS, Very Disappointed Survey, etc.

  2. Feature Market Fit or Product Market Fit - Healthy retention curves on a product or feature level.

  3. Feature Product Fit - Casey Winters has written about Feature Product Fit. From Casey - “Feature/Product Fit requires the feature to improve retention, engagement and/or monetization for the core product. If it doesn't this means it is cannibalizing another part of the product.”

In a lot of cases, we don't find the success signals on our first try. Thats fine. Assuming we have done our filtering, it will be easier to find out why our hypothesis is wrong which will help inform us of where we should navigate to next.

Step Five: Leverage

Once you find some initial success signals around your hypothesis, it is time to leverage it into the next layer of audience you want to target. Think about it as a layer of concentric circles, starting at the center and expanding from there. At some point, you will have built up enough success signals, successful users with strong word of mouth, and other elements that you can remove all filters and swing the doors open.

Example: Superhuman

Back to where I started this post, Superhuman. Just to be clear, I have no affiliation with Superhuman or know anyone on the team well.

Awhile back (I can't remember exactly when) I tried signing up for Superhuman. Here are the hurdles I had to jump through:

  1. Join The Waitlist - I joined the waitlist by entering my email.

  2. Initial Survey - I then completed a survey giving info like my company, size of company, role, etc.

  3. Long Survey - I then at some point received another survey that was about 15 to 20 questions long asking me about what email client I used, how often I email, what add ons are vital for me, etc.

  4. Follow Up Email Convo - I then received an email from someone on their team, asking me

  5. Manual Onboarding - I then had to set up a time to go through manual onboarding for the product.

Most of you are probably thinking, “Holy Sh*t, this is the worst way to launch.” Before you draw that conclusion, you might want to look at the amount of word of mouth (like this, this, and this) and press (like this, this, and this) they have received because it is more than 95%+ of the product or feature launches I have seen. Not to mention the substantial amount of capital they have have raised over multiple rounds.

Superhuman is following the exact above process. They clearly have a specific hypothesis of who they are targeting with the initial versions of the product. They then have used a bunch of tactics to create a waitlist. They then have deployed a number of filters to make sure they are getting close to their audience hypothesis for their initial users. They are then looking for qualitative success signals via manual on-boarding (and I'm assuming quantitative ones internally as well). They are then leveraging that into the next layer of audience they build for and unlock and repeat the process again.

Note: Somewhere between creating this presentation, and publishing the blog version of this, the founder of Superhuman wrote a post on this process titled How Superhuman Built An Engine To Find Product/Market Fit. It is worth the read, and while targeted for startups launching initial product, the principles apply to any product or feature launch.

I write infrequently, but when I do I try to make it good. Subscribe to my blog here so you don’t miss a post. Or join Reforge with me and a community of other leaders.

Build A Growth Machine

-  Unlock growth for your business, product, or yourself
-  Learn how to never run out of new growth ideas
-  Learn how to rapidly test and analyze experiments
-  Learn the 5 key foundational elements for growth



Superhuman was founded in 2015 (3ish years ago at the time of writing this). From their landing page, they are building “The Fastest Email Experience Ever Made.” To this day, you can still not sign up and instantly gain access to their product. Yet, they have received more press, word of mouth, and funding than 95%+ of other products. Why? Because they've done the exact opposite of what most do for product and feature launches. Let me explain...