This Email Strategy Changed Everything

Hi from flight #1 traveling from Vietnam -> Taiwan -> NYC! 

I just went on the trip of a lifetime. If you haven’t been to Vietnam, it’s an absolute must. Out of all the places I’ve traveled, this ties for my first against Machu Picchu/Peru. The cities I traveled to ranked by my most favorite to least favorite: 

Last week I wrote about how businesses in Vietnam win when they blend innovation with tradition, and why anchoring your own product in tradition is critical. I received awesome replies to this one, so I’m glad it landed and was helpful. 

This week, I wanted to give you something more tactical to copy.

Below I’ll spill exactly how we turned a flat chart into one that’s aggressively up-and-to-the-right at Citizen.

It’s now driving 40M emails/month (compared to 500K/mo total email sends in Q2 2024) and is a top lever for how we resurrect churned users.

We needed to resurrect users who churned (30 days inactive) and turned off push notifications. Push is Citizen’s main product (our notification engine that sends emergency alerts). When it’s off, we’ve got three channels to connect with people:

We don’t do SMS, so this leaves email as the most direct and cost effective way to reach churned users at scale. We can slide right into their inboxes. 

I watched 2 other team members fail at making email marketing drive significant growth in MAU (monthly active users) or revenue before. I didn’t want a project that was considered a handoff L. I also had logged into Iterable (our Email Service Provider) probably twice in my life. My convo with our Head of Product, Jon, went:

Him: “You should own this.”
Me: “I don’t want this.”
Him: “You’ll crush it, I know you can do it. But I’ll do it if you won’t, and that’s ok.”
Me: “Fine. Fine. Fine. I’ll do it.”

(We laugh about this to this day, you’ll see why. Coworkers who push and believe in you are everything!!!)

This was during Q3. I got to building that day. We started manually sending emails to churned that were simply just email versions of our real-time safety incidents.

Within 4 weeks I had enough stat sig of a clear win based on what I built that we added it to our engineering roadmap and, in Q1 2025, we architected and built an entire product around the work. 

Every strategy I’ve scaled has started the same way: hands-on, very manual, and not necessarily pretty. I test different approaches until there’s just enough stat sig to confidently bring in product, engineering, or contractors to scale it. 

I optimize for volume > perfection in nearly everything I do. 

Shoutout to Mara Aguinis, who helped me implement the MVP of this email resurrection strategy, which was an incredibly aggressively manual concept. She’s a rockstar.

I’ll spill everything about this email strategy below, which now sends 40M emails/month.

We built a system that automatically sends emails when major incidents occur, and has become Citizen’s most consistent growth lever for non-push resurrections (defined as: churned users we can’t send push notifications to) in the past 3 years.

When we slow it down or pause it, the chart dips immediately.

The wins I’ve had at Citizen come from doubling down on the core reason people use the product: to get critical, real-time info on nearby emergencies as fast as possible.

This mindset of doubling down on the ONE thing our users really want is the core reason we scaled our social to 2B+ views in under 2 years (from 250K/month to 100M/month; exact strategy linked).

When it came to resurrecting churned users (those who turned off push) via email, I started from the same principle.

The email setup I inherited was basic: basic lifecycle flows, e.g. welcome, permission nudges, upsells, “we miss you” winbacks. Not enough to move the needle on the volume of users we needed to resurrect.

Instead of sending spammy flows (one of the reasons I hate email marketing as a consumer myself), I started manually sending live incidents to people’s inboxes, as close to real-time as possible. Local first, national only when major (e.g. hurricanes, mass shootings).

I copied exactly what I did on social media that won and did it on email. 

I’d never really built for email before and had probably logged into Iterable (our Email Service Provider) twice. I wasn’t fluent in best practices like subject lines or title strategy, so I just relied on instinct, moved fast and in high volume.

Step-by-step of building the MVP: 

Once we got the rhythm down, Mara and I ended up making and manually sending 15+ emails per day to hundreds of thousands or millions of people. Jordan at Icepop helped us learn how to use Iterable and import/manually send these to validate this MVP.

Challenges of this manual MVP: 

We got immediate feedback from Iterable on things like open rate and CTR, but the only true indicator of non-push resurrections was via a Google Query I built with our Data Engineer, manually running it daily. Core metrics:

I ultimately wanted to see if this was an interesting and hot enough system to bring people back to the product. I’m not saying I picked the most perfect metrics to track, but this is what we narrowed in on and it worked.

I also couldn’t blast out unlimited emails (sadly) due to Iterable contract caps, so we had to be strategic. Early results were murky because of Citizen’s overall user volume (vs. the amount of emails we were physically able to send), but the chart moved just enough that we noticed.

Building is always about tradeoffs.

We made a bet that the stat sig we saw in the metrics moving was enough to allocate resources towards the product. Not everyone on our product team had as high a conviction in this initially, so we definitely had to sell this a bit. 

It turns out the bet was correct.

What Exists Today: 

Trigger-based emails now go out automatically when an incident hits a certain scale, like “citywide”.

If you’ve been one of the people who has reached out after receiving these emails, this is where they came from! We’ve received awesome feedback on this product, including a ton of people organically saying “Oh, I got that via email!”, and have metrics that show it has: 

We are now sending 40M emails/month. This is up from numbers like:

To accommodate for this huge jump in our email volume, I had to renegotiate our Iterable contract that was expiring to accommodate and ended up getting 2.5X+ the amount of email sends for less $.

Of course, we also spent time rebuilding our lifecycle flows like onboarding, upselling, etc. Still, this automated system drives most of our email results.

FYI, we have guardrails on this and carefully ramped to ensure we didn’t mess up our email deliverability and domain. We also haven’t yet hit diminishing returns, and the automated system has been live for almost 6 months.

Because this chart is so up and to the right, we’ve continued to gently iterate on it. For example, we know that our light emails perform better from a CTR (click through rate) and CTOR (click through to open rate) and so we’ve started new designs to reflect that. 

Ultimately, doubling down on the main reason your customers use and love your product is the ultimate growth + resurrection lever. Look to this thesis to move charts!

I’m so proud of what we’ve built. 

I hope you all have a great week. 

Julia