How Morning Brew’s referral program built an audience of 1.5 million subscribers
The inner workings of one of the most powerful growth tools in email.
I joined Morning Brew in 2017, back when our daily newsletter had a modest 100,000 subscribers. Since then, our audience has grown to 1.5 million — in just 18 months. I wanted to share how we did it — both to tell our story, but also because we’ve benefitted from others who have shared their story with us along the way.
For those who haven’t heard of Morning Brew, we’re a newsletter-first media company that offers young professionals engaging, entertaining business news.
For the past two plus years, I’ve been working in a role that’s best described as a hybrid of engineering, product, and growth positions. That’s partly out of necessity (as employee #2, you have to wear a lot of hats), and partly out of interest (I’m obsessed with growth, but also love software development and creating product roadmaps).
Back in 2017…
My goal was simply to build anything that would grow the audience.
First, I created webpages that allowed Morning Brew’s content (the full newsletter and the individual stories that comprised it) to be archived online. That made it possible for readers to share the individual newsletter stories on their social media accounts directly from the newsletter. The idea was that people who saw these posts would read the stories and sign up for the newsletter while visiting Morning Brew’s website.
Up until that point, all of our growth had come from word-of-mouth and the nonstop hustle of the Brew’s 2 co-founders, Alex Lieberman and Austin Rief.
After a few months, I turned my attention to maximize the value of the 100,000 existing and passionate subscribers. The goal was simple: to create the most badass and effective referral tool in the world. (Ambitious and dramatic, sure…but I would rather aim too high than not high enough).
While the referral program obviously doesn’t account for all 1.5 million subscribers we have today (we began pursuing paid acquisition in early 2018), it does account for over 30% of our total subscribers and is the “secret sauce” that makes our growth flywheel spin. It’s helped turn readers into evangelists and evangelists into walking advertisements. It’s the ultimate 1 + 1 = 3 scenario that makes all of our acquisition channels X times more effective.
Over the past few years, one of the most common replies we get each morning is some variation of, “what software do you use for your referral program?” The answer is we don’t use any third party applications. It’s entirely custom built, and after putting in a ton of time speaking to people about the program, I decided it’s time to put these thoughts on paper.
Whether you are in the media space or have a physical consumer product, I believe the strategies here apply across industries and I hope you find tremendous value in this post.
Rule #1: Start with the product
When looking to maximize growth, there are two essential criteria to consider:
- Is the product valuable to the end user/customer?
- How can you build growth mechanisms natively into the product?
Morning Brew grew organically via word-of-mouth for years before an official referral program was ever put into place. When your product is good enough that people instinctively refer others without any sort of incentive, that’s a strong indicator that you’re ready to build a referral program. On the flip side: trying to optimize a referral program for an incomplete (or useless) product is the equivalent of trying to fill a bucket with a hole in the bottom. Analogy ✓.
As for the second criteria, I can think of countless examples.
- HQ Trivia provided extra lives when users referred someone else to play.
- Dropbox gave free additional storage to those who referred others.
- Harry’s Razors created a milestone-based referral program providing free product before they even launched.
- Wealthfront has an Invite Program that gives both the referrer and referee an additional $5,000 managed for free. And it works really well (you can use my referral code here).
Morning Brew has implemented a milestone-based referral program, which has a dedicated “Share The Brew” section baked into every single newsletter we send.
I’m fortunate to have an incredibly talented editorial team that makes receiving an email every morning enjoyable. Our referral triggers are strategically implemented into every newsletter and allow readers the opportunity to earn rewards from sharing.
Our referral program is entirely custom-built using Ruby on Rails and a few integrations with our ESP (email service provider), Sailthru. Here’s how this works:
- Every subscriber has a unique referral code associated with their account (mine is d8aabc).
- Using this referral code, Morning Brew generates a unique referral link for you to share with others (mine is https://www.morningbrew.com/daily/r/?kid=d8aabc).
- When someone (whom we’ll refer to as subscriber X) signs up using your unique referral link, they are forever linked to your account. You will always be their “referrer.”
- Morning Brew uses a double opt-in process to confirm every email is legitimate (nice try firstname.lastname@example.org). After subscriber X signs up, they’ll be prompted to confirm their email.
- Clicking the link to confirm their email triggers an action on Sailthru to send an API request to our database.
- That request will tell us that subscriber X has just confirmed their email, meaning that you (as the referrer) should receive credit for that transaction.
- My referral count will be incremented by 1, and you may or may not receive a transactional email alerting you of the update (depending on how many referrals you have…but more on that in a bit).
This process took a long time to develop. When I first joined the company, our philosophy was to grow at all costs. We initially didn’t require referrals to double opt-in and confirm their email because it creates an additional barrier to signing up. However, to appease the email gods — I decided it was in the best interest of the company to make the switch. Since then, we’ve seen a conversion rate of over 85% on referral signups going on to confirm their email.
While that effectively means we decreased our conversion rate 15% on referral signups, we’ve also ensured higher quality subscribers join our list. When building a referral program, it’s important to consider all sorts of tradeoffs — whether it’s profits, time, or something else specific to your business.
It’s also important to build safeguards and understand your program’s vulnerabilities. Providing free rewards and incentives always invites nefarious actors. In addition to requiring referrals to double opt-in and confirm their emails, I’ve implemented a few defensive measures:
- Using a third party email validation software to validate every email address that enters our ecosystem.
- Receiving an email every time someone hits a milestone that displays all of their referrals’ emails (it’s very easy to identify fake and/or disposable emails).
- Having a list of 600 (and growing) fake email domains that we blacklist.
- Blacklisting said nefarious actors from being able to refer any additional people.
- Blocking specific IP addresses from submitting emails at any time.
I’ve also built an integration with Forest Admin to create an admin panel with custom functions that allows our lovely interns the ability to provide some basic level of “customer service.” With thousands of people referring others every single day, there are plenty of opportunities for readers to get confused or potentially fall through the cracks.
Our admin panel makes it simple to adjust readers’ referral counts and ensure they are getting the proper rewards. It’s extremely important that you’re able to effectively manage the referral program and take care of those who are going out of their way to share your product with others.
Over the years, we’ve been able to take advantage of the intimacy of email and garner tons of tangible feedback from our readers — they just have to hit “reply.” We also leverage the passionate community of Morning Brew Insiders (more on that in a moment) conduct polls and facilitate discussions.
The result: a milestone-based rewards program that was extremely calculated and demand-driven.
Our first reward at just 3 referrals is Light Roast, an exclusive Sunday newsletter.
Light Roast is effective for two main reasons:
- It’s likely that additional “exclusive” content is aligned with the interests of a reader who already enjoys receiving the Brew.
- It’s completely free for us. We don’t need to purchase raw goods, ship products, etc. You share the Brew three times, and we’ll add you to an additional email list at no real cost to the business.
While the daily newsletter is sent Monday through Saturday, Light Roast is currently sent to over 75,000 subscribers each Sunday. That is, over 75,000 people have shared Morning Brew with at least 3 people.
5 referrals: we mail readers Morning Brew stickers. Bulk ordering 10,000 die cut stickers from StickerMule breaks down to a unit cost of $0.20/sticker and our pre-stamped envelopes cost roughly $0.65/envelope. That totals roughly $1.25 in cost, or a $0.25 CPA (cost per acquisition) for 5 new subscribers to the Brew.
For reference, the CPA for us on Facebook/Instagram, our largest paid acquisition channels, is typically between $3 and $5.
10 referrals: readers gain access to our exclusive “Insider” community. The private Facebook group, which is quickly approaching 10,000 members, is the place to discuss the latest stories, trends, and events in business, pursue career opportunities, and network with other like-minded Brew readers. Again, this reward comes at no real cost to Morning Brew.
To this point: In exchange for 10 referrals, we have provided value in the form of premium content, an exclusive community, and swag…all at the cost of $1.25.
15 referrals: we mail readers a custom Morning Brew silicone phone wallet. The cost of the phone wallet is $1.50, and the cost of the pre-stamped envelopes is $0.65/envelope.
When someone refers 15 people, the CPA on those referred is $0.23 ((cost of phone wallet + stickers)/15). We’ve actually managed to spend less per subscriber as someone continues to refer additional people.
25 referrals: we ship readers a super soft Morning Brew tri-blend t-shirt. We initially went as cheap as possible and you could tell — they were uncomfortable and no one really wanted to wear them. About a year ago we decided to pony up and purchase the nicest custom t-shirts we could find. Now, the reaction to the shirts is overwhelmingly positive and you can find no fewer than 4 people wearing them in the office on any given day.
50 referrals: we ship readers a Morning Brew “Rise and Grind” coffee mug (a fan favorite).
100 referrals: we ship a “Brewneck” (Morning Brew crewneck sweatshirt). As with the t-shirt, the Brewneck is legitimately one of the more comfortable articles of clothing I own. Not only do you want those who refer 100 people to feel appreciated in receiving quality swag, but when it doubles as a comfortable product…people actually prefer to wear it.
At the end of the day, the rewards program is effective in incentivizing our most engaged readers to share the Brew with others. In exchange, they’ll get content, community, and swag. You can also make the argument that the swag doubles as an indirect marketing channel, meaning that when someone is walking around wearing a Morning Brew t-shirt or drinking from a Morning Brew coffee mug in the office, it’s increasing our brand’s exposure.
I also think that the referral program actually boosts engagement for the person who refers others. If someone goes out of their way to share a product or service with their friends, classmates, co-workers, etc., I think the likelihood they continue to engage with that product increases. No one wants to look foolish by abandoning a product they’ve so vehemently recommended so soon after making that recommendation
With the tech built and the rewards in place, the next step was creating a central hub where users can learn about the rewards, track their progress, and utilize tools to help them share the Brew seamlessly. Said in broader terms — the page’s main goal is to educate, motivate, and assist the end user to continue to share our product with others.
People first need to know you have a rewards program (and how it works) before it can become effective.
The top of the page explicitly states the value proposition to the user, i.e. if you share Morning Brew, you’ll get free rewards. That’s followed by a beautiful image that displays each reward clearly and the associated referral milestone required to receive it.
When building a referral program, you have to put yourself in the mindset of your users. We know our readers are busy and a majority typically read the newsletter while getting ready in the morning, commuting to work, or sitting at their desks to start their day. Making the rewards as clear, visual, and repetitive as possible is crucial.
We motivate users by providing a real-time counter that tracks how many referrals someone has, along with some encouragement: “You’re only X referrals away from receiving Y!”
We previously experimented with a progress bar (shown below) to convey a similar message. The decision to move from a progress bar to a numeric counter was actually less data-driven than it was aesthetic; however, we haven’t seen much change in effectiveness. The goal is ultimately the same: update the user with their progress and encourage them to hit the next milestone.
This is where you do all you can to make actually sharing your product as easy as possible. The ultimate goal here is to remove any bit of friction.
First, we provide users with their unique referral link along with explicit directions on how to use it. We also added a “Copy Link” button to make it even easier to copy to your clipboard. This option educates the user about their unique referral link and gives them flexibility on the places they can share it. Once it’s copied to their clipboard, they can do with it as they please…whether it’s posting it on Reddit, sharing it with classmates, or texting it to a friend.
Next, we allow users to refer others via email. Users can manually enter in people’s emails or access their contacts via a 3rd party integration with Cloudsponge, which allows them to import contacts from Gmail, Yahoo, or Outlook. Again, not everyone knows their colleagues’ emails by heart, so we make it simple to import contacts with a few clicks.
Within the form for the email content, we provide a pre-drafted “blurb” explaining the Brew and its value proposition. This prevents the user from wasting their time coming up with a way to explain it to a friend, and quite frankly — we probably can explain it a bit better. However it is editable, so if the user so chooses, they can add their own flare to the message or start from scratch.
Lastly, we provide a section dedicated to sharing on social platforms (and SMS, WhatsApp, and Messenger on mobile). By pressing any of these sharing buttons, a user’s unique referral link is appended to a pre-drafted message waiting to be shared with their network.
When it comes to providing options, sometimes less is more.
For example, people who are referred via email have a conversion rate of 75% on our landing page. That’s significantly higher than the 35% conversion rate for all referrals who did not come from an email invite.
Given that data, perhaps it’s best to only provide the option to invite via email, although that could have an adverse effect. It’s possible only a subset of people would ever utilize the email invite feature, vs. a larger pool of people who are more comfortable sharing on social media. We’re continuing to test these hypotheses.
We also found that sharing via WhatsApp and SMS leads to 10x more signups than LinkedIn, 5x more signups than Twitter, and 2x more signups than Facebook. For example, every five people who share via WhatsApp results in one confirmed subscriber, whereas it takes ~25 people sharing via Twitter to result in a single subscriber.
The data tells us that 1:1 channels like SMS and WhatsApp are significantly more effective than the 1:many networks of Facebook, Twitter, and LinkedIn. Does that mean we should hide those social networks to users who are on this page via a mobile device? Again, something we are continuing to test.
Let’s continue to work our way down the funnel. After a reader opts to share their unique referral link with someone, the goal now is to make sure that someone becomes a subscriber.
Our landing pages are designed to be minimalist and straightforward. You either sign up or you don’t. That being said, there are still several variables on the landing page that can be tested, which include the layout of the page, header text, subheader text, text on the button, style of the form, color of the button, additional images, testimonials, etc.
Depending on the source, site visitors respond to different messaging. Facebook ads, blog posts, and in this case, referrals, are all driving traffic to our site. Each of these user buckets responds differently to the aesthetic, language, and functionality of our webpages.
We use Google Optimize to test every aspect of our landing page, and that’s especially important for optimizing incoming traffic from our referral program. We split our testing audiences into two main cohorts: referrals via email invites and all other referrals.
Starting with the latter, we increased the conversion rate by over 4% through a few iterations involving the header, subheader text, button text, and button color, and we’re still continuing to test new hypotheses. That 4% increase in conversion rate leads to more than an additional 4,500 subscribers per month via the referral program alone.
For example, we tested our original landing page against a new design with copy that called out that your friend recommended you sign up for this.
The new variation beat out the original by 3.1%.
By running the same test of original landing page vs. new designs, we were able to increase the conversion rate of traffic via email invites by over 7%.
As you can see, the conversion rate of traffic from email invites is significantly higher than other referral traffic (from social media, SMS, etc.). This is why we prioritize the “Share via email” section on the referral hub, placing it above the social sharing buttons.
Newsletter share section
If I broke the entire referral process into a funnel, it would look like this.
Now that you’re familiar with the nuts and bolts of the referral program, let’s talk about the top of funnel.
We have a dedicated “Share The Brew” section included in every single newsletter we send. This section’s purpose is to inform the casual reader about the referral program. For the more engaged reader who has already referred a few friends, this section’s purpose is to incite more action.
To emphasize the importance of this section:
- We send our daily email to 1.5M people every morning
- ~45% of people open the newsletter on any given day
- ~X% of these people click the “Click to Share” button (which links to their unique referral hub) on any given day
It is only after step #3 where everything I’ve discussed up until this point plays a role. The different features within the referral hub, the rewards we offer, and the conversion rate of our landing pages don’t even matter if readers don’t engage with the “Share the Brew” section in our newsletter (that is, we need to optimize the bolded X in step #3 above).
There are a few different ways in which we accomplish that. Our generic share section accomplishes similar goals to the referral hub. We provide a value proposition at the top, give the reader an update on their current referral count, and outline all of the rewards in a visual and easy-to-digest manner.
We also aim to remove as much friction as possible, so we provide the reader with two options on how to proceed to share the Brew. They can “Click to Share” (which takes them to their unique referral hub), or they can “copy & paste” their unique referral link. All of this should sound fairly similar to the referral hub section above…because it is. It’s essentially a poor man’s referral hub (or at least as much as we modify it within the constraints of email).
It gets more interesting when we utilize Zephyr, which is Sailthru’s scripting language that allows us to add some conditional logic into our emails. For example, if you have fewer than 3 referrals, we know you don’t have access to Light Roast. Therefore we can show you a share section dedicated exclusively to Light Roast (as opposed to someone with 20 referrals who would likely see something about being 5 referrals away from an awesome Brew t-shirt).
To ensure this section doesn’t get too stale, we are constantly changing what a reader sees on any given day based on their referral count. We’ll rotate between animated gifs and showcasing specific rewards.
For example, if you have 0 referrals: on Monday you’ll see a plug about Light Roast, Tuesday you’ll see something about stickers, Wednesday a t-shirt, Thursday a coffee mug, etc.
Different people are motivated by different rewards, so we try to switch it up as much as possible. It’s also an opportunity to educate readers about our rewards as much as it is an opportunity to encourage them to take action.
Aside from our daily email, which is considered a “campaign” or “bulk email” because it’s sent to a significant number of people all at once, we also send many one-off “transactional” emails to our readers. The primary objective of the emails is to boost the effectiveness of the referral program, and it can be broken down into three main categories: the initial ask, milestone, and nudge.
The Initial Ask
The reality is that the vast majority of our readers have 0 referrals. Convincing someone to go out of their way to share our product (or any product) is a challenge, so most of our focus goes into enticing them to get from 0 to 1 referral. Once a reader gets their very first referral, it’s apparent that they:
- Acknowledge the existence of the rewards program
- Are incentivized by the rewards
- Understand how to share their unique link
- Have received some sort of gratification for doing so
Despite the existence of the “Share the Brew” section in the newsletter, there’s no guarantee they a new subscriber knows about or understands the rewards program. To force the issue, we send a dedicated email to all subscribers who have 0 referrals exactly 7 days after they subscribe.
Our hypothesis: After reading the newsletter for a week, readers are at their peak excitement about being a new subscriber and digesting business news in a fun and entertaining manner. It’s at this point that they are most likely to talk about it and refer others to join.
I think there’s a strong correlation between novelty of product and the willingness to share it with others. I love my iPhoneX, but haven’t really told anyone about how much I enjoy using it since a few weeks after I bought it.
Because the biggest obstacle is getting a user to share for the first time, we’ve tested this initial “7 day email” from dozens of different angles. We’ve tested it by incentivizing Light Roast vs. stickers vs. showcasing all of the rewards vs. a genuine friendly message from our CEO vs. offering Starbucks gift cards, etc.
Using Sailthru, we’re able to split test the email 5 different ways at any given time. We’ll then take the winning email, think a little bit more about why it won, and set up another 5-way split test to optimize the conversion rate even further.
Below is a chart displaying the number of referrals a user gets by days since they signed up for the newsletter. The result is what you would expect.
Just by looking at a small cohort of users, you can see the noticeable impact this one-off email has in accelerating the referral flywheel early on in the user’s lifecycle. With just a 1% increase in conversion rate on this email, we can increase the number of referrals by thousands each month.
These emails are triggered after your first successful referral and at each subsequent referral milestone in which you unlock some sort of reward (5, 10, 15, 25, 50, 100, 1,000). The purpose of these emails: to acknowledge the reader’s accomplishment, show them how to redeem their reward (if necessary), and to motivate them to hit the next milestone.
This entire process is what we call “the referral pipeline.” Ultimately, our goal once you get your very first referral is to push to get your 3rd, 5th, 10th, etc. We’re trying to maximize the amount of value (in terms of growth) users can contribute to our company.
Again, with each of these automated emails sent along a user’s journey, we test them five different ways to optimize the CTR (click through rate). For the first email you receive after you get your first referral, we’ve tested a congratulatory message, a sincere “thank you,” a locker room speech to get you to run through a wall, and everything in between.
The message that’s consistent in each of these emails: “Congrats for hitting this milestone and thanks for sharing the Brew…but did you know you’re only X referrals away from unlocking the next reward Y? Why on earth would you stop now?”
For milestones where we need to ship swag to readers, we provide both a form (via Typeform) to collect subscribers’ addresses and relevant info and a message to encourage them to hit the next milestone. In all, we have eight milestone emails. Each is being tested multiple ways to optimize the number of recipients who take action and click “Share the Brew” (which takes them right back to their referral hub).
The lowest hanging fruit in the pipeline is what I call “nudge emails.” Essentially, readers who have been stuck for two weeks just one referral away from a milestone (2, 4, 9) receive an email to nudge them to get that one last referral.
For example, if you’ve been stuck at four referrals for two weeks, you’ll receive an email saying, “Hey, congrats on getting four referrals and thanks so much for sharing the Brew with others…but did you know you’re only one referral away from getting stickers? You’ve come this far, might as well just get one more, right?”
As with everything else in our pipeline, we split test the messaging of this email several ways to optimize CTR.
If you’ve been a subscriber of Morning Brew for more than a few months, you’ve probably noticed that we do a big giveaway every 4–5 weeks. It’s nothing revolutionary, but it’s a massive stimulant to the referral program.
The messaging is simple: For the next X hours, every person you refer who confirms their email is an entry into the giveaway. The more people you refer, the more entries you have.
Given that the biggest challenge is getting a user from 0 to 1 referral, we’ve found that providing a strong incentive via a giveaway converts thousands of new subscribers who previously had 0 referrals to refer. In other words, people who have seen the “Share the Brew” section dozens of times and chosen to ignore it now have an incentive strong enough to click the button taking them to their referral hub.
Once there, they see their dashboard and how simple it is to share the Brew with others. With just a few clicks, they are able to share it with tons of friends, receive a few emails congratulating them for doing so, and receive that instant gratification to boost that positive feedback loop. While the odds of them winning may be very low, we were able to “teach” them the behavior of sharing the newsletter with others.
We’ve tested giving away MacBook Pros, iPhones and Samsung Galaxys, $2,000 cash, trips to Singapore, and trips to NYC to hangout with Brew employees and take an exclusive tour of the Nasdaq. We try to keep the giveaways fresh and a true value-add to the reader so it doesn’t get annoying.
Over 225,000 people have referred at least one person to subscribe to Morning Brew. During the summer of 2018, we had months where we averaged over 1,000 referrals per day (it’s obviously increased since, but that’s all I’m going to give you here).
While the referral program itself has been extremely effective, there’s always room to improve. A few things we’re continuing to evaluate:
- What other rewards could we introduce or substitute that would increase demand while remaining cost-effective?
- How can we continue to optimize the language around the referral program to entice a larger portion of our audience to participate?
- Should we introduce additional sharing functionality to provide more options for readers to share, or in contrast, remove some features to simplify the process?
- Can we coincide this rewards program with organic social media campaigns to showcase subscribers reading the Brew and wearing our swag?
- Are there parts of the pipeline where additional nudges are needed, or are there redundancies that can be removed?
A few percent optimization anywhere along the funnel can compound into thousands of additional referrals per month. I’m just fortunate to be surrounded by an incredible team that makes each additional percent that much more enjoyable.
And if you’ve made it this far and aren’t already a subscriber—feel free to subscribe for free here.
A few very special shoutouts: to Neal Freyman, Kinsey Grant, Taylor Johnson, Bryan Wish, and Tommy Johnson for editing and helping me finalize this piece. To Austin Rief and Alex Lieberman who both took a chance in hiring me 2 years ago and have since worked closely with me to build Morning Brew to be the best it can be. To Jenny Rothenberg who has worked side by side with me the past 10 months in helping to optimize this very referral program (and also helped edit this piece). To Francis Scialabba for making the awesome header image. And to our incredible intern Betsally “swag daddy” Falcones who does the dirty work behind the scenes of managing the fulfillment of the thousands of rewards we ship to readers each week.