AI Dev Cases
← Back to list
ToneAdapt

A 21-Year-Old Student Turned His Own Guitar-Tone Frustration Into an App — ToneAdapt Hit $25K/Month in 5 Months on Zero Ad Spend and 100K Users

A guitar/bass ‘tone-matching’ app built by SDSU computer-science student Kyan Santiago-Calling: pick any song’s tone and it translates the settings to your own gear. Shipped as a web app in December 2025, it reached 100K users in 100 days and ~$25K/month in about five months — with zero paid ads, purely organic TikTok/Instagram content.

Published: Jul 23, 20264 min readPrimary-source verified · 4
Monthly (est.)
$25k/mo
Time to grow
5 months
Users
100K
Launched
2025
KyanKyan@kyanbuildsA 21-Year-Old Student Turned His Own Guitar-Tone Frustration Into an App — ToneAdapt Hit $25K/Month in 5 Months on Zero Ad Spend and 100K Users

Key takeaways

  • Choose a niche where you’re personally the sufferer, not one by market size (first-hand pain fuels both product intuition and content credibility)
  • Pick a topic whose demo is entertainment — the product’s core experience *is* the short-form video (choose a distribution-friendly topic first, don’t grind distribution later)
  • Validate demand and revenue on cheap web before investing in native (don’t melt money on native + ads up front)

The pain point, and how they found it

Dialing in guitar tone is a swamp that frustrates even experienced players: you want the sound of a specific song, but EQ curves, presets and pickups differ by brand, so copying someone else’s ‘knob positions’ produces a different sound on your rig. Kyan hit this universal ‘tone doesn’t transfer across gear’ pain himself after getting into ’80s metal — and that frustration was the market entry point. ToneAdapt translates the *tonal character*, not the knob positions, for your specific gear.

Background & product

ToneAdapt translates a song’s guitar/bass tone into settings for *your* specific gear. Name a song (text-to-tone) or upload audio (music-to-tone), and it returns an estimate of the original gear plus custom settings adapted to your amp, guitar and pedals. Behind it sits a library of 50,000+ songs and a gear database of thousands of items — pitched as ‘the largest gear database on the internet,’ which is the core of its differentiation.

It was built by Kyan Santiago-Calling, 21, a computer-science student at San Diego State University (SDSU) with prior machine-learning experience at Amazon. He picked up guitar in 2024, fell down the ’80s-metal rabbit hole, and started the app as a ‘resume project’ out of his own pain of ‘I can’t get that sound.’ The fact that the builder is himself a user stuck in that swamp is exactly what powered the growth later.

The launch sequencing is clever. He first shipped a *web* app on December 8, 2025 (January revenue was a humble $750), validated demand and revenue over the winter, then in March expanded the gear database from 200 to 4,000+ — thickening the data moat — before shipping the iOS app in April. Rather than investing in a native app up front, he validated on cheap web first, then went native: a sequence indie builders can copy.

Pricing is freemium, with Pro at $9.99/month. Guitarists happily pay a small subscription for ‘a tool they open every time they pick up the instrument’ — and indeed, of 100K+ cumulative users, ~73K keep coming back. What worked most was the *shape* of the growth. Guitar tone is something whose value lands the instant you show, in a 15-second clip, ‘here’s that famous song’s sound, on *your* gear.’ In other words, the product itself is the content. Kyan posted those clips to TikTok/Instagram and racked up millions of views. In his own words he ‘still doesn’t understand or run TikTok or Meta ads’ — he has grown this far on zero paid advertising.

From the founder (primary source)

ToneAdapt growth channels and tech stack

The repeatable playbook

  1. 1Choose a niche where you’re personally the sufferer, not one by market size (first-hand pain fuels both product intuition and content credibility)
  2. 2Pick a topic whose demo is entertainment — the product’s core experience *is* the short-form video (choose a distribution-friendly topic first, don’t grind distribution later)
  3. 3Validate demand and revenue on cheap web before investing in native (don’t melt money on native + ads up front)
  4. 4Put the moat in data, not AI — asset-ize the ‘niche mapping table’ only you can accumulate, not the AI anyone can call
  5. 5Price for a hobby wallet ($9.99/mo) and lean on ‘opened every time’ frequency so retention earns back the low ARPU
  6. 6Zero paid ads: concentrate on build-in-public (X) and organic short-form (TikTok/Instagram) to create early momentum

The hard parts

The limits of reproducibility, honestly. The builder, Kyan, is a CS student with machine-learning experience at Amazon — not a true non-engineer; the agentic-AI tone engine takes real implementation chops. The organic growth also worked largely because the topic is inherently demo-friendly *and* he was embedded in the community as a genuine guitarist; a passionless third party mass-producing the same clips would struggle to earn the same trust. The numbers, too, span $17K–$35K across sources (independent tracker TrustMRR shows $17,261 over the last 30 days), and a single hobby niche carries a TAM ceiling plus the structural reality that views and downloads slow the moment the content stops — a stamina game that should be discounted accordingly.

Deep dive

【Deep dive】ToneAdapt’s growth isn’t luck — it’s a stack of repeatable design decisions. Let’s break them down in order.

■ Pick a niche you actually live in (first-hand pain over big TAM). Kyan didn’t choose his topic by market size. He chose ‘guitar tone,’ something he was personally obsessed with and personally stuck on. This pays off two ways: (1) first-hand experience tells you exactly what the real pain is, so you don’t miss the product’s crux; and (2) your content carries the credibility of a real guitarist. Being a genuine insider in a small niche buys early trust faster than chasing a big market you entered from the outside. The ‘boring niche’ the giants ignore is, for an indie, actually a moat.

■ Product = content (pick a topic whose demo is entertainment). His biggest weapon is that guitar tone is inherently satisfying to watch and hear. ‘Here’s that famous song’s sound, on *your* gear’ is a finished 15-second hook. So without running ads, simply turning the product’s core experience into short clips earned millions of organic views. The lesson isn’t ‘engineer viral videos’ — it’s ‘choose a product whose demo is entertaining in the first place.’ Choosing a topic where distribution works beats grinding on distribution after the fact.

■ Validate on cheap web → thicken the data moat → then native (the order of investment). Rather than building an iOS app first, he shipped a web app in December 2025 and checked demand and revenue ($750 in January → ~$9,600 by February). Then in March he expanded the gear database from 200 to 4,000+, thickening the ‘largest gear database on the internet’ differentiation, before shipping iOS in April. Reverse the order — invest in native before validating, pour on ads with a thin moat — and you melt time and money on something that may not land. Validate cheap and fast, then make the heavy investment once the winning path is visible: for an unfunded indie, that sequence is a lifeline.

■ The moat is data, not AI (build the interface, accumulate the asset). The entry point for tone generation is agentic AI (text-to-tone / music-to-tone), but that’s not the real differentiation. The moat is the proprietary ‘gear × song’ database that correctly maps original gear and settings across thousands of pieces of equipment. Anyone can call an AI; that mapping table you have to accumulate yourself. To add one step to Cal AI’s ‘don’t build frontier AI, translate it’ — ‘own the *dictionary* you translate with.’ Deep data in a narrow niche is a barrier the incumbents can’t casually match.

■ Price for a hobby wallet, and usage frequency that means ‘opened every time.’ Pro is $9.99/month. Rather than pushing ARPU up, he set it where a guitarist pays without feeling it. What matters is frequency: a tool you open ‘every time you pick up the instrument’ resists churn — ~73K of the 100K+ users keep coming back, and that high return rate is the foundation under the subscription. Hobby markets have low ARPU, but if you embed into a daily ritual, retention earns it back.

■ The real numbers (read them conservatively). The headline is the founder’s own $25K/month ($10K/mo within 30 days of launch, with a peak ‘$35K last month’). Meanwhile independent tracker TrustMRR shows $17,261 over the last 30 days; between web vs iOS and annual-plan accounting, published figures span $17K–$35K. Indie ‘single-month peaks’ tend to run hot, so it’s safer to make decisions off the low end. Even so, $25K/month in five months from a 21-year-old’s first app on zero ad spend reads as the payoff of the sequence above clicking into place: niche selection → demo-as-content → validation first → a data moat → the right price.

Cross-case “growth playbook” report (coming soon)

We're building a paid report that aggregates every case in this database: which acquisition tactics worked, in which categories, and how well — insights you can't see from a single story. Get notified first when it launches.

Related cases