Lately I’ve been thinking about how craft is the antidote to ubiquitous AI slop. Quality still matters, and it might even matter more than before when it’s easy to make something mid and hard to make something special. It’s a core part of how I approach this newsletter.

I’ve found a software company that embodies this ethos and the results speak for themselves. It’s Mangomint, booking and management software for salons and spas. They compete in a crowded category to say the least, and nearly all of their customers rip-and-replace an existing product when they adopt Mangomint. 

Yet they recently passed $25 million ARR and are doubling year-on-year. Logo retention is 88% and net revenue retention is 110%, unusually high for a product that starts at $165 per month. Their NPS is 72. 

The unusual ways Mangomint got there via co-founder and CEO Daniel Lang:

Let’s get into it.

👋 Hi, it’s Kyle Poyar and welcome to Growth Unhinged, my weekly newsletter exploring the hidden playbooks behind the fastest-growing startups. Today’s edition is free. Subscribe or upgrade to support Growth Unhinged (and unlock new Unhinged Perks).

Mangomint started inbound-only. 60% of their lead volume still comes from organic, non-paid sources.

“We don’t really know where they came from and they don’t really know, either,” Daniel admitted. 

Daniel’s best guess is that there’s a combination of word of mouth and a viral coefficient. Mangomint spas have visible “Powered by Mangomint” branding, which puts Mangomint’s name in front of more than a million people booking appointments each month (and more than $1B in annual transaction volume). This works because Mangomint spent real time perfecting the consumer experience around the product and then makes their product easy to buy.

This made for an efficient acquisition motion, but definitionally constrained Mangomint’s growth. “I always felt like we could’ve grown faster if we did a lot more outbound. In vertical SaaS SMB that usually means cold calling. We’ve never done that and still aren’t doing it today.”

The company doesn’t cold call, but they’re increasingly investing in outbound marketing. Daniel sees outbound as one touchpoint in what will end up being a multi-touch, long-term play. “Rather than cold calling into businesses, our first touch is a highly personalized gift that shows up in direct mail,” he told me.

This is automated through data that Mangomint has tirelessly collected. They tried Clay, Apollo, and other third-party data tools, but concluded that none of them work for Mangomint’s hyper-specific SMB vertical. 

Part of the issue: Mangomint has a nuanced ideal customer profile (ICP) that’s more qualitative than quantitative. “It’s not about how many people they have or where they’re located. It’s about how good their website is and whether the business feels premium.”

Instead, the team invested in their own data infrastructure. They’ve built a Snowflake database that pulls in over a million businesses. These are selectively enriched with AI agents. This gets used to prioritize the best-fit businesses and to personalize marketing campaigns like direct mail: a hair removal business is going to respond to very different campaigns than a med spa. 

| Related: How to use AI agents for marketing

Mangomint now ships hundreds (if not thousands) of items each month as part of their outbound marketing. They ship so much they have a literal warehouse with three warehouse workers who are Mangomint employees.

What they’re shipping: “It’s not the expensive high value items. It’s often the thoughtful items that might even feel too little value.”

Mangomint initially tried a campaign where they shipped med spas “vibrators but it’s for your face”. (These are real things.) These were custom manufactured and included Mangomint branding. It backfired; prospects thought Mangomint was a vibrator company.

They pivoted to Apple AirTags, the bluetooth tracking devices. This didn’t work, either. “Some people booked demos because they felt guilty, not because they really wanted a demo.”

What did work: sending out wax sticks and other low-cost, industry-specific swag to hair removal businesses. “We had funny stickers and pins. People loved those. We signed up a ton of businesses and got a lot of unboxing videos on Instagram and TikTok.”

In the past few years big software companies have bought up niche media and community sites. Outreach acquired Sales Hacker, the destination for B2B sales pros. HubSpot bought The Hustle, which reached two million people at the time. Semrush paired up with Backlinko, a newsletter reaching 170,000 marketers. Pendo grabbed Mind the Product. 

The rationale is clear: third-party media companies have cracked the code on attracting a large and highly engaged audience for very little cost. Software companies have cracked the code on monetizing an audience. But these media acquisitions have almost always been the purview of extremely well-financed unicorns or public companies. 

Mangomint spent $2 million on The Salon Business, a media destination for salon business owners. At the time Mangomint had a mere $6 million in cash. Talk about risky!

“It has paid off many times over,” Daniel shared. “It’s an evergreen source of leads. When the LLMs came out it exploded in value.”

AI search, and ChatGPT specifically, look for in-depth content from what they see as independent and authoritative sources. On that front a media company fits the bill exceptionally well. 

The Salon Business publishes evergreen comparative content, including guides to the best salon software and how to price services. This gets further segmented by vertical including guides for medical spas, tanning salons, barbershops, and massage therapy. This is a powerful way to win AI search that almost nobody’s talking about.

At the beginning of the year I probably wouldn’t have predicted content-related tactics being the best place to double down. Elena Verna wrote in March that “company blogs are no longer worth the investment.” The reported SEO decline of HubSpot’s blog went viral.

The real story isn’t about decline. It’s about a platform shift as B2B buyers turn to ChatGPT for recommendations about which vendors to consider and which products to buy. AI discovery now accounts for 12.7% of Docebo’s high-intent leads (up 429% year-on-year) and 10% of Webflow’s signups (up 4x year-on-year).

Some positive news: this isn’t hard pivot. Late-funnel SEO is the foundation for great answer engine optimization (AEO).

Mangomint’s average customer spends about $7,000 per year. This is a notorious no man’s land for product-led growth. It’s typically too expensive (and complex) for fully self-onboarding and too cheap to fund a complex rip-and-replace of a legacy system.

100% of Mangomint is PLG and there’s no way to buy it without going through a trial. Customers don’t need to sign a contract. And customers can export all their data at any time in a universally accessible format. 

Onboarding (with a human onboarding manager) happens during the free trial, not after someone starts paying. “We flipped it because we know that once someone’s onboarding their chance of converting is way higher,” Daniel emphasized. Their win rate from lead-to-paying-customer: 36%.

If you’ve read this far, you’d think Mangomint would be bullish on customer success. The reality: they don’t have a single CSM whose job is to make customers happy.

Instead Mangomint has onboarding managers who help before someone pays. And they have a support team who helps after. It’s that simple.

The learning: this elevates the role of customer support and helps provide a career path inside the team. Support isn’t relegated to cleaning up when things go wrong; they’re the main point of contact with the customer. 

What about upsell? It’s mostly product-led. (More on that later.) 

Daniel has strong opinions (clearly). One of them: Mangomint has chosen not to use AI for customer support. They don’t even use third-party support software; they cancelled Zendesk and invested in their own learning center.

“Everyone on our support team gets paid 30% more than at comparable companies,” Daniel said. “We don’t use AI. We don’t use automations. We have a text box, and a human reads it and responds directly. And we still have high gross margins.”

Daniel’s advice: Try your own customer service as a customer. You’d be shocked at how bad it is.

Mangomint doesn’t have customer success managers. Yet their net revenue retention is 110% (quite high for an SMB sale) and the average customer now adopts 3.3 products. More than half purchase Forms and nearly as many buy Connect.

How they do it: they always try to make add-ons native to the user experience. It’s a product-led growth experience, applied to cross-selling new products. “Anytime we don’t do that it’s very cumbersome,” Daniel told me. 

Here’s a taste of what that looks like for the Connect product, which lets customers call, text, and chat with clients.

Some products like payrolling processing don’t fit naturally with PLG. Customers generally need a personal interaction to talk about benefits, insurance, compliance, and so forth. Even here, Mangomint includes a contextual call to action (CTA) in the product, which triggers a product qualified lead (PQL) for the customer support team. Support sets up a call to onboard the customer and owns the upsell (yes, you heard that right). More than 20% of Mangomint’s customers attach payroll, a product that was launched in the past year.

What stands out to me about Daniel is his thoughtfulness and attention to detail around every aspect of the customer experience, and how he believes this is central to Mangomint’s long-term growth. It’s a refreshing antidote to the AI slop that’s increasingly filling our feeds.

Daniel also joined the Mostly Growth podcast where he shared more behind-the-scenes insights about Mangomint’s growth strategy as well as his vision for bringing AI to Main Street businesses.

📺 Watch on YouTube | 🎧 Listen on Apple Podcasts | 🎧 Listen on Spotify

What we covered:

Invite your friends and earn rewards