Discovering Your Ideal Users: How Visily Scaled to 100K Signups in 5 Steps

“A picture is worth a thousand words,” as the adage goes. The chart below demonstrates the value of obsessing over your user’s needs, perhaps better than the words that follow. But let me provide some context.

Visily’s new monthly signup growth from October 2022 through April 2023.

I spent the last several years of my career in growth roles at companies like FullStory and Cypress.io. These experiences helped me launch my path toward advising and consulting other companies on their growth strategies—and ultimately led me to my current position at KMS Technology as VP of Growth.

In October of 2022, I started advising the company Visily through KMS Studio. Visily is an AI-powered wire-framing and design tool that helps non-designers rapidly create designs. As a non-designer, it allowed me to transform a screenshot of a UI into an editable mockup that I shared with others, including my wife, an actual designer.

This first introduction to the product blew me away.

So I wasn’t surprised to hear that the Visily team gained serious traction, both in signups and in notable companies using their product in the last few months of 2022 into 2023.

But they also noticed an alarming trend in their strong signup growth: activation and retention among newly acquired users was lagging behind where they should be.

As the team dug into the data, they were unsettled to find that a majority of signups came from geographies far outside their target geographies. Additionally, many of the people signing up had no clear need for the product, based on their answers during onboarding (i.e., “do dentists do a lot of wire-framing?”).

The culprit? New signups that were cheaply acquired through paid channels.

But as the previous graph shows—the purple part, specifically—something equally noteworthy was happening during the ensuing months. Ads were completely turned off, and “organic” (direct, word-of-mouth, referral, etc.) signups flourished.

The graph below shows when new activated users started to climb.

Monthly new activated users from Jan 2023 to April 2023.

So what on earth was going on?!

How do you find your Ideal Customer Profile?

From the beginning, the Visily team had assumptions about who benefited most from their product: non-designers who, despite lacking design skills, wanted to create their own wireframes. However, they never concretely formalized an ICP or specified the context where Visily would be most beneficial (i.e. the “job” in “Jobs to be Done” theory). This lack of clarity had a number of unintended consequences:

• Disjointed messaging: If you’re speaking to everyone, you’re speaking to no one.
• Misaligned marketing channels: We struggled to know where to find the right users for our product, so we relied on the easiest channels to drive signups, irrespective of the quality of those signups.
• Conflicting opinions on product roadmap: Perhaps the biggest challenge the lack of ICP presented was how it impacted the product roadmap. Without a deep understanding of our ideal user, it was impossible to build a cohesive product vision.
• Stymied growth: Many people are oblivious to the relationship between ICP and growth, particularly in a PLG company. A simple way of thinking about it is that if you do not correctly identify who your product serves, you will struggle to build a product that’s relevant for your users. You will struggle to retain users, which means your viral flywheel will never take off—no user invites, no virality, etc.

The 5 steps for discovering your ICP

Worried by these developments, our team crystallized our ICP and the jobs they need Visily to help them perform. That, in turn, helped inform the team how to connect the ICP to Visily’s features, messaging, and positioning. The results from this increased focus on user needs helped Visily cross the 100K user milestone.

Here’s the process we used to align ourselves and build an organic growth machine:

1. Formulate a hypothesis

As a team, we discussed our personal (and previously unspoken) hypotheses about who our product was for:

• “We’re a design tool, so we target designers.”
• “Project managers at agencies like us because they can quickly whip up wireframes.”
• “Founders needing to design their app UI get value from us.”

While none of our assumptions were completely false, they were hardly aligned.

To fix this, we initially removed titles and roles from the discussion. We found that emphasizing the job title of our ICP, at least at first, introduced a number of biases and assumptions about said roles that clouded our discussion. This helped us clearly state what we believed our product helped people accomplish (again influenced by Jobs to Be Done), rather than the title or functional role of the people that would find Visily valuable.

2. Conduct prospective ICP interviews (not just customer interviews)

The goal of a prospective ICP interview is not to talk about your product (at least not at first). Instead, you want to understand how the person or team works and validate that the problem your product solves is indeed important to them.

For example, rather than ask users directly about product features they’d like to see, we ask questions pertaining to the workflow we believe Visily might help. Here are some examples of the questions we might ask during a Visily prospective ICP interview:

• What team or individual typically owns the different steps of the feature development process?
• What are the formal stages or steps adhered to in the product development process?
• Are there points in the process that frequently break down or lack clarity? If so, can you describe an instance where this has occurred? What caused it?
• What are the hand-off points between tools used during this process?
• If a non-designer has a UI idea, how do they convey that idea to the team? Do they work with a designer or work alone to create an artifact of some sort?

Notice that none of these questions attempt to discuss our product or lead the interviewee to a specific answer. Sometimes people confuse customer or user interviews for ICP interviews, but they are not the same.

For one, users are presumably already finding value in your product. By contrast, prospective ICPs most likely have no familiarity with your product but ideally do have familiarity with the problem your product solves.

After interviewing various people involved in the product life-cycle (PMs, UX specialists, designers, developers, etc.), we started to see a trend. Product managers at software companies:

1. universally loathed the design tools foisted upon them internally (“they’re too complicated!“), and
2. struggled to adequately convey (visually) their ideas when discussing them with designers. Bingo.

3. Confirm your findings with segmented usage data

Conferring with your data should be a constant pursuit, full stop. However, doing so after coalescing around an ICP is especially useful.

The reason to hold off on a deep analysis on product usage until after creating an ICP hypothesis and conducting interviews may be controversial, but the logic is sound. The current product usage data is for the product you built before identifying your prospective ICP. In other words, it’s possible that your product currently does not maximally benefit your prospective ICP.

As a result, being overly reliant on current usage data to come up with your Ideal Customer Profile hypothesis may actually obscure your ICP and how they’d leverage your product.

Luckily for Visily, our analysis showed that our proposed ICP had a 30-day retention rate that was almost 2x that of non-ICPs, and this was without making any product or marketing changes.

Had we not first discussed the underlying assumptions of our product, come up with the core “job to be done” that it fulfilled, and interviewed prospective ICPs before deeply diving into the data, we may have come to the wrong conclusion.

4. Align roadmap, messaging, and channels

Visily AI feature usage from March 2022 to March 2023.

Driving alignment across all core functions has been essential for Visily through this journey. When we uncover new insights from ideal customer profile and user interviews, we share notes and recordings throughout the team, ensuring everyone has equal access to them.

This alignment has led us to translate insights into action plans for each functional area, including the re-working of our product roadmap to speak specifically to ICP needs.

A few examples of how Visily translates insights into team-specific actions.

Visily’s new AI panel, which increases AI feature usage among all users.

5. Establish “ICP Confirmation Metrics”

How to know if you’re going after the “right ICP” can perplex many founders. Much like the question of achieving product-market fit, people often sense it but have lingering concerns. This is where something I call “ICP confirmation metrics” comes into play.
As mentioned in step three, segmenting your metrics by ICP versus non-ICP is critical for tracking progress. If you’re a product-led company, there are (at least) two metrics you should watch carefully:

1. Activation: ICP vs. Non-ICP
2. Retention: ICP vs. Non-ICP

How to know your process is working

Combining these simple steps ultimately helped all to align around some very particular use cases and the roles they served, but sometimes we were still left asking ourselves whether things were working or not.

In reality, your PLG flywheel will never spin if you can’t activate and retain your ideal users. This is the reason we established these as our “ICP confirmation metrics.” As you continue to refine and iterate on your message, channels, and product to appeal more to your ICP, you should start to see in your own ICP confirmation metrics these segments diverge more and more (in a positive way). You should also see your ICP segment activating and retaining at a progressively higher rare. That is how you will know things are working.