We recently launched what was literally the most requested feature we've ever had: mail merge with automatic follow-up sequences in Gmail. It lets you send personalized mass emails with automatic follow-ups to a list of recipients to help with outreach, customer education, or hiring.

Why did we even build this?

For the longest time automated sequences has been a #1 feature request of Streak users. We always hesitated to build it because we figured there were specialized tools that could do it better and that users should integrate those with Streak.

Two things changed our mind and made us want to build it ourselves:

1. Most of the email sequence products out there kind of sucked for small agile teams. We tried all the products out there and couldn’t send an email within the first 30 minutes of using them. They were (rightfully for them) built for Fortune 500 companies and had a pretty awful UX for both the sender and recipient. 
2. We realized that we were asking too much of our customers when we told them to integrate Streak with these other tools. It was too cumbersome to send an email with another tool and then use Streak to track the replies and conversations. Our customers wanted a deeper automation, like changing the stage in the CRM when a recipient replied. Asking them to piece that together with Zapier or our API was just too painful.

Creating a new product development process to incorporate user feedback

We knew this would be a big undertaking. We didn’t want to build something that no one used, so we tried out a new product development process to incorporate user feedback into the early stages of product development. It resulted in us building this product a little differently than usual, and the results were amazing.

I’ve learned that this is the best way to build anything that takes longer than a few weeks. Follow this product development process and you’ll:

• Never build something users don’t want
• Keep your team motivated with regular launches AND early usage from customers
• Always know what to build next because your customers will tell you
• Have a community of people rooting for you on launch day

Here’s how we did it, and how you can try this product development method, too.

Flywheel #1: Interview to Prototype Loop

We started our product development process with some basic customer interviews asking about the problems customers were facing and how their business worked. Next we’d show them some high fidelity mocks of the product we were thinking of building.

At first these mockups weren’t nearly what users wanted. But showing them helped the interviews get more concrete so we could understand what users really wanted.

For example, we learned that most customers wanted to send mass emails with automatic follow-up to get responses, not necessarily meetings, from the leads and opportunities in their pipelines. It became clear that we needed to optimize our mail merges for getting responses. 

Next, we’d iterate on these high fidelity mockups based on what we learned during the previous customer interview. We’d use this new version of the prototype in the next interview. And then we just kept doing this - interview, learning, iterating on the prototype – then repeating.

Slowly the changes between interviews became more and more minor till we eventually started getting the  "hey, so can I use this right now?!" type feedback in our sessions. That’s when we knew we had something great.

Flywheel #2: Building in public 

We've had huge death-march product releases in the past and we wanted to avoid that, so when we started building out this feature we set a goal of having something that at least 1 user (me) could use in the first 6 weeks. 

It was barely functional, but again, we built a fast iteration loop.

Slack channel for beta users

Once we moved into the building stage, we set up a Slack channel for beta users so we could let them know about every update we made. This also meant they could easily share feedback directly with the builders on our team.

Because this was such a requested feature, we got tons of feedback. This super engaged community led to our second feedback loop:‍

We'd add users to the beta, see what was missing and most important to them, build it out in 3 weeks, then kept repeating 🔄. ‍

We just kept doing that over and over: add users, respond to feedback.

Deciding what to work on next was so easy, the feedback was overwhelmingly obvious. This kept our team’s momentum and morale high – every time we launched something new we had customers use it right away.

Sharing our process on Twitter

We also expanded the feedback loop beyond our beta users by building in public on Twitter. I shared the project details, posted pictures of early mockups, and even shared challenges and progress updates along the way.

It was super honest and candid - and we got great responses. People were able to share what they wanted to see in our mail merge and what hasn’t worked in other products. 

For example, one person shared, “I never want to email someone I’ve emailed before unless I okay it.” This feedback directly led to us adding a feature that alerts you if you or anyone on your team have already emailed someone in your recipient list.

People were really excited about this feature and loved getting a “behind the scenes” look at how it was built and our thought process along the way.

Beyond all of the feedback we got, sharing our product development process on Twitter was a great way to build connections with our customers and get them excited about the launch. 

Launch rollout

Since we got so much customer feedback throughout the entire product development process, it was super clear when we were ready to launch: When our beta testers loved it, and the product was 🔥. 

It's a leap beyond our previous mail merge functionality, more stable, and has the guard rails in place to make sure users don't exceed Gmail's tricky quota limitations.

You can try out mail merge for free by installing Streak in the upper-right corner, and learn more about our product development process and behind the scenes at Streak on my Twitter @aloo.