🌿 Palmstreet: From Plant ID App To 9-figure Live Shopping Empire

The best consumer founders know the game. 35K+ of them read Consumer Startups every week.

Stay ahead. Get the playbook behind today's breakout startups.

Read time: 6 mins 08 seconds

​​This founder built 30+ apps before finding the one that took him to $1M ARR in <6 months.

He then pivoted to pursue something even bigger. 

Meet Chen Li, the founder of Palmstreet, the live shopping app that's quietly dominating niche collectibles, from $500 rare tropical plants to designer toys. 

Palmstreet grew 50x in the last two years to $200M+ GMV, and is nearly profitable at the Series A stage (almost unheard of for a marketplace).

The path there involved a graveyard of failed pivots, an Instagram Live hack, and a growth play where the founder literally bought plants from sellers just to get them to talk to him.

Excited to share the winning playbook from one of the fastest-growing marketplace startups with you all 👇.

A word from our partner

A lot of teams are building based on vibes.

I used Listen Labs to get real signal for Pikes AI in the creative AI space: what users complain about, what they’ll actually pay for, and what’s just noise. I set it up in minutes, and the output looked like what you’d normally wait weeks (and spend a lot) to get.

Listen recruits from a 30M+ network, runs video interviews with real follow-ups, and turns it into a clean report with themes + highlight clips. Use it for churn diagnosis, concept testing, messaging, or “which feature matters” debates.

Microsoft, Sweetgreen, and Anthropic are already on it. 1M+ interviews done. $100M+ raised from Sequoia and Ribbit. Check it out!

Chen is part of the first batch of iOS developers, a real OG. He worked at Apple, Instagram, and then Firework (a TikTok competitor in 2018).

He loves building apps. He has shipped 30+ apps in his career, each one born from a personal pain point.

Palmstreet started as a simple plant identification app for his daughter during the pandemic.

His daughter kept asking "Daddy, what's that plant?" on their walks. He wanted to be more knowledgeable, so he gathered a team and built an AI-powered app called Plant Story that could identify any plant from a photo.

Different from other plant ID apps in the market, Chen made the app super simple — no sign-up walls, no lengthy onboarding questionnaire, no data collection screens. You open it, you point your camera at a plant, you get an answer. He posted it in a few Facebook groups … and Plant Story took off!

The app started getting 5,000+ organic downloads a day, converting to paid subscription at ~5%.

Plant Story reached $1M ARR in < 6 months.

Most founders would double down on growing the same product after reaching this level of traction, but Chen didn't.

He saw something bigger when building Plant Story.

Users weren't using it on trails like he expected. They were using it in their backyards and home gardens. The pandemic had millions of people investing in their gardens for the first time, and they needed to know what was growing in their own soil.

Americans spend $20 billion+ a year in their gardens, and his users were already in those gardens.

The next two years turned out to be a painful series of pivots.

❌ Attempt 1: watering reminders. Plant ID is a one-time action. Maybe users needed ongoing value like reminders to water their plants. However, when plants are in a garden outdoors, they don't need manual watering. There are devices and irrigation systems for that.

❌ Attempt 2: social media for plants. Maybe there was a play to build a better version of Facebook plant groups. They did it and it was better. BUT it was only about 50% better than Facebook groups. And 50% better is not enough to get someone to leave a platform they already use, where their friends already are, and switch to a new app.

❌ Attempt 3: Etsy-style marketplace. They went full commerce and started building an Etsy-style marketplace for buying and selling plants, including payments, authentication, product listings, and all types of commerce features. Unfortunately, the same story happened. They were slightly better than Etsy and Facebook Marketplace, but not different enough. People weren't willing to leave platforms they already trusted.

It was clear that building an incrementally better product was not going to work. Chen and his team needed something a lot more innovative. 

Chen went back to first principles and tried to identify the biggest pain points in the plant buying journey. He first segmented the plant market into two buckets:

 Common plants ($5-$20): This is Home Depot territory. People would rather drive 10 minutes and carry a plant home than pay $10 shipping on a $20 fern. There's no way to out-compete Home Depot here. The economics don't work for an online marketplace.

Rare plants ($200-$500): This is where things get interesting. Shipping costs become trivial relative to the price. However, there was a massive trust problem. These plants were being sold in Facebook groups, and scammers were rampant. You'd send $400 to a stranger and hope your rare Monstera shows up alive.

One idea he had was to use live shopping to address the trust problem in the rare plant buying process. If a seller is on camera, holding up the exact plant you’re buying, showing you every leaf, there’s no reason to scam when your face is exposed.

There were precedents in other countries. In China, live shopping was already a massive category on Douyin (TikTok China), but no one had built a dedicated app for it in the US.

This live shopping for plants idea was faced with plenty of doubts from his other smart friends.

If you can’t even make it work with a regular marketplace, how do you expect it to work with a novel model that’s never been validated in the US. 

He decided to give it a try regardless, but this time without building a product first. 

One of his employees, based in Pennsylvania, bought rare plants locally and did an Instagram Live sale. Viewers could comment to indicate their interest in buying the plants shown in the stream, and the team would DM each buyer with a custom Stripe link for purchase.

The friction was absurd, but 70% of buyers still completed payment after the stream. 

They ran three streams and kept iterating based on feedback from each previous stream. Revenue grew every time, starting at $300, then $500, and $700.

Most importantly, they started noticing repeat buyers on the second and third stream. Those three streams gave Chen enough signal to pull the trigger and go all in.

He got an Airbnb for his entire team, two engineers, one designer, and one PM. They shipped the first version of Palmstreet after 8 weeks of Airbnb war room. 

It was a simple live shopping app that also had a native purchasing flow.

Getting initial sellers was the first challenge.

Multiple attempts failed in the beginning. They first reached out to brick-and-mortar plant shops, but shop owners were reluctant to invest in this uncertain technology. They then started cold-messaging sellers on FB groups, but were often ignored.

Then, they finally found the hack. Chen started buying from those plant sellers. When you buy someone's $300 plant, they have to talk to you because you're their customer. After the purchase, Chen would say: "By the way, I loved buying from you, but I actually run this live shopping platform called Palmstreet. I want to pay you to try selling on it."

They took the offer and tried it. Many of them started sticking around after seeing how much easier and more engaging it was than Instagram Live. They wanted to keep selling because the format worked and started getting them repeat buyers and more revenue.

In the rare plant community, Palmstreet was the exciting new thing — live shopping without the painful workarounds of Instagram or Facebook Live.

Fun alone isn’t a moat. Chen and his Palmstreet squad started rolling out deep, plant-specific operational features that no other horizontal marketplace would bother with:

Features like these have made Palmstreet a 10x better platform than Etsy and other horizontal marketplaces. 

This product superiority also led to explosive organic growth. Even today, 90%+ of Palmstreet's users are organic. Within that paid bucket, a big chunk is influencer-driven and user-referral-driven rather than traditional ads. The product itself is the growth engine.

Starting with rare tropical plants, Palmstreet has since expanded into other types of rare plants, aquatics and reptiles, and niche crafts and toys categories.

The category selection framework is both push and pull:

The key strategy behind category expansion is to always expand with a unique operational edge, which often goes beyond software.

Take the aquatics category, for example. You can't ship live fish to someone's doorstep. Palmstreet solved this by integrating FedEx hub pickup into the app. When a buyer purchases a fish, the app asks for their home address, surfaces nearby FedEx hub locations, and lets the buyer choose a pickup spot. The seller confirms, ships overnight, and the buyer picks up the next day at a nearby FedEx hub. The app even asks which day the buyer wants to pick up, because if they're on vacation and miss the window, the animal dies. The entire UX shifts from "ship to my door" to a coordinated pickup flow.

Palmstreet has grown 50x over the past 2 years, reaching 2M+ downloads and $200M+ GMV. It raised $25M from a16z, Craft, Headline, and more.

1/ Being 50% better doesn't create switching behavior. Palmstreet tried reminders, social feeds, and an Etsy-style marketplace. All slightly better than what existed. None worked. The breakthrough came from a major innovation — live video solved trust in a way no incremental feature improvement could.

2/ Cold-start a marketplace by becoming a customer first. When Facebook group sellers ignored cold outreach, Chen bought from them. Once he was their customer, they had to engage.

3/ The best product differentiation can be operational, not technical. Weather-based shipping holds. Tiered dead-plant refund policies. FedEx hub pickup coordination for live animals. None of these are hard engineering problems. They're hard knowledge problems. You have to deeply understand the category to even know they matter.

4/ Segment your market before choosing where to play. Chen didn't try to compete across all plants. He split the market into common ($5-$20) and rare ($200-$500) and recognized that only the rare segment had economics that worked online.

5/ Know when to walk away from "good enough." The plant ID app hit ~$1M ARR in 6 months. Most founders would optimize from there. Chen walked away because he saw the ceiling. The much bigger opportunity required burning down the thing that was already working.

—

See you next Tuesday!

Leo

Follow me on X and LinkedIn

📥️ Want to advertise in Consumer Startups? Learn more.