Experimentation and a Singular Focus

When 100% of your sales come from online subscriptions, the marketing department needs to mix proven methods and new ideas, so Copy.AI CMO Blake Emal takes experimentation seriously. In this interview, Blake discusses how to manage experimentation in a way that effectively increases company revenue.

Copy.ai provides artificial intelligence that helps companies create blogs quickly, and the company has grown extremely rapidly. Blake was the company’s first hire, so there was no onboarding process--he had to establish the procedures for hiring the dozen team members that have joined over the last year. Although his official title is CMO, Blake functions as a Chief Revenue Officer, because all of the company’s revenue comes from online sales--that is, all marketing. 

Blake goes through how he keeps his team creative while staying focused on one primary outcome. 

In this article, you’ll learn:

• How to create a formal process for vetting ideas
• How to simplify measurement of outcomes
• Asynchronous processes that keep the team in flow

💡Ideas that contribute to the bottom line

Because the Copy.ai website gets a good amount of traffic--more than 1000 sign-ups per day--the team has a lot of room to try out new ideas to improve the site’s performance. At the same time, there’s only so much that the team can do. After finding himself overwhelmed with too many experiments, Blake created a process that all team members need to follow before they implement any new idea.

“It’s easy to come up with a million ideas, but if you just write your ideas on a list, you’re probably not going to do them,” says Blake. “I’ve developed a formal process that creates quite a bit of friction up front and requires you to do some work before starting any experimentation.”

The company has two templates in Notion for experimentation: proposal and post-mortem. The proposal template takes about an hour to complete and includes the following information:

• Precedent: Source of the idea. Usually ideas come from practices the team saw elsewhere.
• Sanity check: The team member is encouraged to speak to a few other team members to make sure the idea makes sense.
• Description: Brief description of the idea.
• Hypothesis: The hypothesis describes why the experiment is expected to work, and what behavior is expected.
• Development stage: Status of the idea or where it fits into the overall process
• Role: What team or role the idea belongs to (product, marketing, retention, etc.)
• Channel
• Timeline
• Core metrics: Specific metrics by which the success or failure of the experiment will be measured. 
• Expected metric input: The resources that will be needed to perform the experiment.
• Prioritization score: Team member suggestions get a prioritization to determine what to do next.
• Approaches to carrying out the experiment: The idea may have one or more possible modes for execution of the experiment.
• Milestones and timeline.

{{vital="/blog-inserts"}}

Blake says that once a person takes the time to put all of this information down on paper, they are more committed to actually following through on the experiment. The document doesn’t have to be perfect--if a third or a quarter of the experiments succeed, the company will see increases in Monthly Recurring Revenue (MFRR).

“What’s most important is to keep innovating,” says Blake, “but we want to avoid the mistake I made when I first came to the company, trying to do everything all the time.” By thinking through all aspects of the experiment, the team members end up executing on the best ones on a regular basis. 

To track everything that’s happening in the marketing department, Blake maintains one central page where all the experiments are listed. “I’ve created this experimentation table that’s the one page that, if you look at it every day, you know that you haven’t missed anything major in the department,” he says. “Even if you end up ignoring 99% of the other pages on Notion, you have one source of truth where people can see exactly what we’re working on, why it’s impactful and how we expect it to go.” Rather than just a list of links, Blake thinks of this as a “workstreams” page that allows everyone in the department to stay on track and not get bogged down in the tremendous amount of content that builds up in the team communications portal. 

📈One Metric to Rule them All

“I measure marketing success by one metric: Monthly Recurring Revenue,” says Blake. “We don’t have a sales team; we’re bottom-up SaaS. If there’s anything I do that doesn’t feed back into MRR, I need to question that.”

The myopic focus on revenue has clearly increased the company’s growth tremendously. 

⚖️ Combining asynchronous and synchronous meetings

The Copy.AI team successfully integrates synchronous and asynchronous communications, alternating weekly one-on-ones between sync and async. 

For the synchronous meetings, Blake lets the team member lead the conversation in terms of the agenda, updates, and blockers where they might need help. For asynchronous meetings, Blake uses a simple template:

• Opening with a “more human” prompt and check-in.
• Three core points, usually about the experiments and execution of that specific employee.
• Closing survey. The survey changes so it doesn’t get too boring. Typical questions would be how effective the session was, how organized they are feeling, how the leadership is communicating, etc.
  

Sticking to three or four points keeps the meetings brief and focused on the department’s priorities. By giving the employee the ability to lead the call every other week, Blake ensures nothing gets missed.

Surprisingly, Blake has gotten more in-depth and honest feedback through the asynchronous meetings than face-to-face.

“I've gone through some async interviews where that the people have written so much that it takes 30 minutes just to go through it. But it’s a very healthy exercise. I get way more honest feedback and asynchronous one on ones than I do in zoom calls, because people are more willing to say what they truly mean when they can write it and not have to look at you while they're saying it."

{{go-further-with-hugo="/blog-inserts"}}