The Somatic Team 5 min read

A coach that knows your history. Nothing else does.

Most fitness apps store your data but never use it. Here's what a truly personalized training recommendation actually requires.

You’ve Been Logging Everything. Nothing’s Been Reading It.

Two years of sessions. Every weight, every rep, every exercise you’ve touched since you started taking this seriously. Your training log is more complete than most people’s. And your app still recommends the same split it gave you on week one.

A 2010 study in the Journal of Strength and Conditioning Research found that athletes who received programming adjusted to their individual performance response showed significantly better outcomes than those following fixed schedules.1 The variable that mattered wasn’t the program structure. It was whether the program actually knew who was running it.

Most apps store everything and read nothing.

The Gap Between a Plan and a Coach

The fitness app market is full of products that are essentially digital logbooks with a recommendation layer bolted on top. The recommendation layer doesn’t read your logbook. It reads a template.

This matters because training isn’t static. Your readiness changes week to week. Your muscle group frequency balance shifts. Your body responds differently at week three of an accumulation block than it does at week one. A rigid program can’t see any of that. It tells you what day it is and what’s scheduled.

The result is a recommendation that’s theoretically sound and practically tone-deaf. It doesn’t know you’ve been hammering push patterns all week and your pull volume has lagged. It doesn’t know you had two hard sessions back to back and today probably isn’t the day for a heavy deadlift attempt. It doesn’t know you’ve been adding reps to your bench consistently for six weeks and a weight bump is overdue.

A plan knows your goals. A coach knows your history. Most apps give you the former and call it the latter.

What Real Personalization Actually Requires

Track more than sets and reps

The minimum useful dataset for a personalized recommendation includes: sessions completed, muscle groups trained, weights and reps per exercise, session duration, and heart rate data where available. Volume by muscle group per week — not just total volume — matters in particular, since imbalances develop quietly and compound over time. If your logging system doesn’t capture this, it can’t generate a useful output.

Understand your training phase

Programming should look different at different points in a training block. Early accumulation (weeks 1–3) calls for higher volume, moderate intensity. Later, intensity rises and volume drops. Deload week reduces both. A static program treats all weeks the same. A responsive system knows where you are and adjusts accordingly.2

Surface imbalances before they become problems

The most actionable insight a coach provides isn’t “what to train today” — it’s “what you’ve been neglecting.” Volume imbalances between muscle groups, lopsided push/pull ratios, a cardio gap in an otherwise strength-focused month. These patterns are visible in the data. The lifter who’s deep in the routine can’t see them.

Use acceptance patterns, not just performance data

What sessions you complete versus skip is as informative as how you perform in the sessions you do. Consistently avoiding certain session types reveals fatigue, boredom, or a mismatch between the recommendation and your actual state. A smart system reads this signal and adjusts.3

Soma does all of this. Before every recommendation, it reads your full session history, volume by muscle group, performance trends, training split classification, and acceptance patterns. The suggestion it generates is specific to where you are today — not where a theoretical version of you is supposed to be.

What It Looks Like in Practice

Take a lifter two months into consistent training. On paper, they’ve been hitting four sessions a week. But when you look at the distribution, it’s heavy on upper body push, light on back work, and cardio has been skipped three weeks running.

A generic app recommends the next scheduled session. A coach looks at the log and says: your posterior chain is falling behind, your pull-to-push ratio is off, and your cardiovascular base is eroding. Let’s address that before it becomes a structural imbalance.

The plan doesn’t care what happened last month. The recommendation is built entirely from it.

The data you’re collecting in your training log is more valuable than most apps make it. A good recommendation isn’t built from your goals — it’s built from your history. Where you’ve been, how your body has responded, what you’ve been neglecting, where you’re ready to push.

No subscription. Your data never leaves your iPhone. Join the beta via TestFlight.

  1. Mann JB et al. (2010). “The Effect of Autoregulatory Progressive Resistance Exercise vs. Linear Periodization on Strength Improvement in College Athletes.” Journal of Strength and Conditioning Research, 24(7), 1718–1723. ↩

  2. Kiely J (2012). “Periodization paradigms in the 21st century: evidence-led or tradition-driven?” International Journal of Sports Physiology and Performance, 7(3), 242–250. ↩

  3. Israetel M, Hoffmann J, Smith C (2019). Scientific Principles of Strength Training. Renaissance Periodization. ↩