EP77 Garmin Buys TrainingPeaks To Build AI Coaches (ft. AI Insights)
The Deep Dive Podcast: Sports Tech & Performance for Endurance AthletesShow Notes
Garmin Buys TrainingPeaks: The $100M Data Masterstroke (ft. AI Insights)Garmin just bought TrainingPeaks. Your training vocabulary — TSS, normalised power — is now legally owned by their biggest hardware rival's new parent.
On 22 July 2026, Garmin acquired TrainingPeaks and TrainHeroic from Peaks Ware Holdings. The official press release calls it a coaching expansion. The structural analysis calls it something else entirely: a defensive manoeuvre to hoard 25 years of irreplaceable human coaching data before Strava's $3 billion IPO war chest could get there first.
Key questions:
• Why did Peaks Ware keep the music education businesses and sell only the sports platforms — and what does that reveal about AI disruption? • Garmin now owns the trademarks to TSS, normalised power and intensity factor. What does that mean for Wahoo, Hammerhead and Trainer Road? • How does the five-runner estimate a $75–130 million purchase price from public pricing data alone — and why do the two valuation methods disagree by $40 million? • What is compliance data, why is it worth more than raw GPS data, and why can Garmin only buy it once? • Garmin promised neutrality when they bought Firstbeat in 2020. What actually happened — and should non-Garmin coaches trust the same promise today? • If Garmin uses this archive to train the ultimate AI coach, will the human coaches who built TrainingPeaks have provided the data to replace themselves?
Verdict: This is not a $100 million software purchase. It is a four-acquisition monopolistic data stack — Tacx in 2019, Firstbeat in 2020, MyLaps in 2025, TrainingPeaks in 2026 — that captures an athlete's complete journey from daily training prescription to race day finish line without the data ever leaving a Garmin server. Brilliant, ruthless, and almost certainly irreversible.
— CHAPTERS —
0:00 Waking up to find Garmin owns your training vocabulary 0:38 What actually changed hands on 22 July 2026 2:20 Why Peaks Ware kept music education and sold the sports assets 3:38 The asset haul: TrainingPeaks, WKO, TrainHeroic, Training Peaks Virtual 4:19 Why Garmin needed a virtual cycling platform for their Tacx hardware 5:11 The trademark bombshell: TSS, normalised power, intensity factor 5:49 Firstbeat versus TrainingPeaks: two completely different fatigue models 7:24 Why Garmin stayed silent on the purchase price 7:48 The revenue method: $36.5 million annual revenue, $114.5 million valuation 9:15 The profit method: 120 Colorado staff, $73 million valuation 10:10 The likely price: $75–130 million, central estimate $100 million 10:27 When the real figure will surface: Q3 10-Q and the 2026 annual report 11:05 The official press release: what Garmin said — and what they omitted 12:08 Why Garmin did not buy TrainingPeaks for the software features 13:00 Compliance data: the difference between the report card and the teacher's notes 14:37 The one asset Garmin could only buy once 15:10 Four acquisitions, one closed loop: Tacx, Firstbeat, MyLaps, TrainingPeaks 16:23 The Strava threat: $3 billion IPO and a cross-brand coaching stack 17:30 How Garmin starved Strava of the asset they needed 17:50 The neutrality promise — and the Firstbeat precedent 19:21 The coach dilemma: migrate now or trust the promise? 20:03 Every Wahoo and Coros workout now feeds data to Garmin 21:20 The final implication: did coaches build the data to replace themselves?
— SOURCES —
the5krunner.com — strategic analysis: what changes with the acquisition the5krunner.com — acquisition price estimate and valuation methodology Garmin Newsroom — official press release DC Rainmaker — community fallout and non-Garmin user impact Cycling Weekly — what this means for non-Garmin users
— MORE FROM THE5KRUNNER —
the5krunner.com Sign up for The Deep Dive Digest newsletter Subscribe to the5krunner
More from The Deep Dive Podcast: Sports Tech & Performance for Endurance Athletes
Can a blood pressure cuff fit inside a watch strap? The Huawei Watch D3 says yes. THE5KRUNNER puts the science under the microscope.
Chapters: 00:00 Ecosystem defender, not Whoop killer: the convergence point across three reviews 03:27 The sensor paradox: Gen 4.1 in a $199.99 device, and why Garmin's explanation doesn't hold 06:15 Six structured accuracy tests: where wrist placement works and where it fails 09:07 The biceps fix: repeat interval test results after moving off the wrist 10:24 DCRainmaker's cold weather finding and the auto-detect failure mode 12:10 15 nights of sleep data: resting heart rate, HRV, and time awake 15:04 Sleep stages: why TechRadar and The5KRunner reach opposite conclusions 16:56 Buyer verdict: who should buy, who should not, and what the data says Sources: the5krunner.com: Garmin CIRQA Band Review TechRadar: Garmin CIRQA Review — Matt Evans DCRainmaker: Garmin CIRQA Accuracy Deep-Dive — MORE FROM THE5KRUNNER — the5krunner.com Sign up for The Deep Dive Digest newsletter Subscribe to the5krunner
Garmin CIRQA at $199.99: Whoop annoyer or genuine rival? THE5KRUNNER unpacks the launch day reality.
Your smartwatch recovery algorithm thinks you're a 25-year-old man. How wearables fail female physiology, VO2 max and HRV. The interpretation layer inside Garmin, WHOOP and Oura was built on male physiology. The sensors gather accurate raw data, but the algorithms treat the menstrual cycle, pregnancy and menopause as noise. We break down why your luteal phase reads as overtraining, why VO2 max estimates penalise the fittest women, how skin tone and wrist size degrade optical heart rate, and the clinical danger of missing RED-S behind a "normal" recovery score. Key Questions We Answer: • Why does a 28-day hormonal cycle break training software built on a 7-day calendar week? • How does progesterone's thermogenic effect crater your body battery and HRV during the luteal phase? • Why do Garmin algorithms underestimate VO2 max for women and overestimate it for men? • Why does the Forerunner 245 penalise highly trained female athletes by 6.3 ml/kg/min? • How do melanin and small wrist anatomy compromise optical heart rate accuracy? • What early RED-S warning signs can wearables flag weeks before injury? • Why does a regular monthly bleed not confirm hormonal health? The Verdict: Wearable data is useful to female athletes only as a personalised baseline, never as objective diagnostic truth. The fittest your data looks, the day of ovulation, is the day ligament laxity peaks and ACL risk is highest. Force clean inputs with a chest strap, set your true maximum heart rate manually, and contextualise every readiness score against cycle phase, training load and life stage. The athlete has to become the final algorithm. Chapters: 0:00 The 25-Year-Old Man Inside Your Watch 2:53 The 7-Day Week vs the 28-Day Cycle 4:29 Follicular Phase: Why the Data Looks Perfect 7:48 Luteal Phase: The Thermogenic Tax 10:56 How the Algorithm Misreads Hormones as Sickness 13:08 VO2 Max: The Male-Biased Math 16:18 Directional Error: Garmin vs Polar 18:53 Why Elite Women Get Penalised Hardest 21:17 Optical Sensor Errors: ±34 BPM 24:36 Cycle-Synced Training in Practice 27:49 The Ovulation Trap: Peak Score, Peak ACL Risk 30:02 Skin Temperature Sensors & Natural Cycles 31:45 Melanin, Skin Tone and PPG Physics 35:34 Wrist Size: A Hostile Environment for Optics 39:35 RED-S: Where Bad Data Turns Clinical 43:17 The Early Warning Signs in Your HRV 47:08 Why a Monthly Bleed Proves Nothing 49:45 Eight Sleep Pod 5: Pregnancy Mode Done Right 55:17 The Baseline Principle: Become the Algorithm Research Sources: Female VO2 Max on Garmin: Why the Number Lies Hormones, Endurance Training and the Menstrual Cycle Wrist Heart Rate Accuracy and Skin Tone RED-S, Female Athletes and Wearable Data Cycle-Synced Running on Garmin More From the5krunner: the5krunner.com Newsletter Sign-Up Subscribe for Premium Content
Whoop Killer War: 8 Brands, 1 Survivor? (ft. AI Insights)Eight brands could launch a screenless band. The hardware is cheap. The algorithms are not. Who actually survives?