Leaving an antiquated industry to grow an AI-native platform to a 7-figure ARR

While building tools for food distribution, Joseph Lee was constantly frustrated by how hard it was to explain his tool to a non-technical customer base. So he built a tool to solve the problem — and marketed it to a more tech-savvy audience.

Today, Supademo is bringing in a mid-7-figure ARR.

Here's Joseph on how he did it. 👇

From seafood to AI-native demo automation

I’m a serial founder who’s been building projects and businesses since the age of 15.

I’m currently the founder and CEO of Supademo, an AI-native demo automation platform. Our platform helps thousands of companies create engaging, interactive product demos for onboarding, sales, marketing, or training.

We're currently mid-7-figures in ARR, with thousands of paying customers and 100k+ founders, marketers, CSMs, and salespeople leveraging our platform.

We’ve been growing really quickly — about 3x so far this year and 8x in 2024.

Before Supademo, I built and scaled Freshline.io, an e-commerce and operations platform for perishables, and Coastline Market, a predictive seafood marketplace, to 7-figure revenues.

Born out of personal frustration

Supademo was born out of personal frustration. Having been a founder in a traditional, antiquated industry — food distribution — I always struggled to explain our product and features to a tech-nascent audience.

With our marketing lingo being ineffective, I experimented with slide decks and videos, only to realize how resource-intensive they were to create, edit, and maintain. Not only that, I noticed that prospects would seldom watch more than 2 minutes of a video monologue. And it was a constant struggle to keep these videos updated with the pace of our product and design changes.

I figured there had to be a better way. So after leaving my last startup, I began ideating and exploring, through which I met my co-founder, Koushik. Together, we began building Supademo, and the rest is history!

You'll learn faster by doing than by analyzing

Supademo is built on

• NextJS

• React + TypeScript

• Node

• PlanetScaleDB

• AWS

And on the AI side, we use our own proprietary agentic AI flows alongside tooling like:

• OpenAI

• ElevenLabs

• Claude

• Gemini

Building the initial product took a lot of trial and error. We began with a simple screenshot-based product, starting with a really narrow ICP and wedge. From there, we began experimenting with adding guided workflows, videos, HTML steps, and more — eventually landing on what is Supademo today.

I think the key lesson here is to get moving ASAP. No amount of planning, business-plan writing, or strategy will prepare you for the lessons that come with real customer interactions.

IMO, you’re always going to learn faster by doing vs. analyzing, especially in the early days.

Supademo's four pillars of growth

We’re a product-led, freemium SaaS platform with a self-serve motion.

Most of our revenue growth comes from virality and word-of-mouth generated by our free plan, alongside expansion from teams adding seats, demos, and advanced features like personalization and analytics.

As far as user growth, we've leveraged:

1. Product craftsmanship — clamping down on feature creep and making the core experience delightful and intuitive regardless of their technical competency. Bias for shipping quickly at established weekly intervals.

2. Lean, product-led experimentation — leveraging our product via ungated experiences, free tools, demo-led SEO, and forced activation during onboarding.

3. SEO and Content:

   1. Programmatic TOFU SEO (piggybacking on common product search terms by creating interactive tutorials), MOFU SEO (target search terms for “{product} demo” by creating pages), and comparison pages.

   2. Using LinkedIn thought leadership to build a founder/brand moat — going from 10k to >500k impressions per quarter.

4. Product-led virality

   1. Made Supademo fully ungated — the ultimate exercise of “try before you buy.”

   2. Free Tools: making subsets of our existing product accessible to adjacent users without creating anything new (i.e., free online screenshot editor).

   3. Implementing a reverse trial with full Scale plan access without signup needed.

   4. Rapidly A/B testing watermarks on Supademos, free tools, trigger moments to drive product-led virality.

Tactics to 5x ARR

This year, our goal was to 5x ARR, and we're not quite there yet. To do that we've been trying to do two other things:

• Graduate from founder-led spray-and-pray growth tactics to hypothesis and experiment-driven growth tactics — and establish at least two consistent, repeatable growth channels.

• Improve product stickiness via expansion (organizational land-and-expand or external virality) by A/B testing, using automated ABM, product-led horizontal seat expansion, and surfacing lateral use cases.

Founder-led distribution with a caveat

Founder-led distribution has been a superpower — sharing raw learnings consistently built trust and created compounding demand.

The one caveat here is that building in public is great, but you should be tailoring your approach for your unique use case and stage.

For instance, X, Product Hunt, Reddit, and Indie Hackers are great channels to build in public if you have an early prototype and want to get feedback and iterate. The same may not apply on LinkedIn, where expectations of quality might be heightened due to the presence of more mature, established companies.

The key here is to deeply understand where your ICP at your stage of venture/idea resides, and tailor the “Build in Public” strategy accordingly based on your needs.

Don’t confuse motion with progress

Here's my advice: Don’t confuse motion with progress.

Get relentless about why you’re doing things. Instead of over-indexing or spending time planning and analyzing, get into the habit of obsessing over selling products or building products — and nothing else.

Ship quickly to put the product in front of real users — measuring only a few core, compounding KPIs. If there’s no painful workaround today, it’s probably not a real problem; find lived friction, show value fast, and let distribution amplify what already works.

What's next?

We’re doubling down on personalization-at-scale and agentic AI—tailoring demos by segment or account with zero extra work, and turning engagement into revenue signals. The goal is to make “show, don’t tell” the default way software is sold, onboarded, and supported.

You can connect with me on LinkedIn or visit supademo.com!