What Europe's GLP-1 users actually need — and the daily product that meets them across the journey

A bottom-up analysis of 13,886 real social conversations. Prepared for the founding team. Read top to bottom: the recommendation is earned in §8–9, not assumed.

The data supports a two-product range sold as one journey: a daily on-drug companion as the hero — broad market, daily consumption, high lifetime value — anchored by a maintenance / transition product for the emotionally-charged moment users come off. They are the same customer at two stages, they share a formula spine (muscle preservation + nutrition adequacy + key micronutrients), and they tell one story: "support your results while you're on it; protect them when you come off."

This report builds that conclusion from the ground up. We first establish what the corpus is and how it was tagged (§1), map every concern by volume, emotion and journey-stage (§2–4), test where private search demand diverges from public conversation (§3), and resolve the concerns into product territories (§5). We then weigh the European competitive reality (§6) and the daily-ritual market signal — including GRUNS' explicit use of a GLP-1 hook (§7) — before reasoning through four product architectures (§8) and arriving at the blended recommendation (§9), its formula logic (§10), the honest risks (§11), and a validation roadmap (§12). Full verbatims and the positioning vocabulary are in the annexes.

1 · Method & corpus

The analysis draws on a purpose-built listening engine that continuously captures public GLP-1 conversation across four platforms, pseudonymises authors, and classifies every item with a two-stage model pipeline: a cheap pass tags pain, journey phase, emotion and geography across the whole corpus (the census), and a stronger pass scores reply-worthy opportunities. A hard safety gate excludes medical-emergency, eating-disorder and crisis content from all downstream use.

Geographic reality, stated up front: 10,145 items are geography-unknown; of the identifiable remainder, the signal skews US (1,024) and is thinner for the intended launch markets — FR (576), UK (545), with Spain folded into "other (890)". This is the single most important caveat in the report: the thesis is derived largely from English-language conversation and must be validated directly in ES/FR before it sets the launch market (see §11–12).

2 · The concern landscape

Every concern, quantified. Raw volume is only one axis — a high-volume concern that is unsolvable (cost) or low-emotion (injection logistics) is a worse product target than a slightly smaller one that runs hot and concentrates at a buyable moment. The table below shows mentions, share, emotional intensity (1–4), and how each concern splits between the on-drug phase and the late (stopping + post-drug) phases.

Two readings matter. First, the top of the table is dominated by daily, on-drug concerns — nutrition, food noise, digestive, muscle, energy — which is the raw material for a daily product. Second, the hottest concerns by emotion (regain actual 2.51, mood 2.60, regain fear 2.49, cravings 2.44) cluster around the late phases — the raw material for the emotional, retention-driving second product. Both signals are real; they describe two stages of one customer.

3 · Cross-signal: where private search diverges from public talk

Conversation volume alone can mislead — people post about some things and silently Google others. Fusing conversation with Google search volume and emotional intensity exposes the gaps.

4 · The journey: where each concern lives

The same concern means different things at different stages, and the product decision lives in where demand concentrates. The corpus splits across five phases:

On-drug-stable (4,212) is by far the largest phase — and this is the crux of the strategic re-think. It is tempting to dismiss it as "just because most of the corpus is mid-treatment," but that is precisely the point: most GLP-1 users, most of the time, are mid-treatment. A daily product addressed to that state has the largest addressable audience and the longest consumption window. Meanwhile the regain cluster spikes at stopping + post-drug — smaller in population but acute in emotion and equipped with a clean commercial trigger (the decision to come off). One range can serve both: the daily product owns the long middle; the maintenance product owns the exit.

5 · Clustering: one product or many?

Which concerns co-occur in the same post tells us whether to build one multi-benefit product or separate SKUs. The pairs resolve into four territories:

Critically, the maintenance and discontinuation territories share a spine — muscle, nutrition, the same micronutrients. That shared spine is what makes a two-SKU range coherent rather than scattered: one formula philosophy, two expressions.

6 · The European context changes the competitive maths

A reflexive objection to an on-drug muscle/nutrition product is "the market is saturated — every protein brand already shouts muscle preservation on GLP-1." That is true in the United States. It is substantially not yet true in Europe, and especially not in Spain, France and the broader EU launch footprint.

This reframes the entire opportunity. In the US the daily on-drug product is a red ocean and the discontinuation niche is the only clean lane. In Europe, the daily on-drug product is itself an open lane — there is no dominant, GLP-1-native daily companion brand with the clean-label, anti-pinkwash positioning this team is built for. The first-mover argument that, in the US, applies only to the discontinuation niche applies across the whole range in Europe. The competitive case against the broad daily product largely dissolves once geography is stated honestly.

7 · The daily-ritual market signal

The fastest recent success in consumer supplements, GRUNS, won not by solving a sharp clinical pain but by being a delightful, simple, universal daily ritual — and it now explicitly markets a GLP-1 support angle as an acquisition hook. Two lessons follow directly.

• The GLP-1 daily-companion framing is commercially validated. The category's breakout brand is leaning into exactly this association.
• The winning shape is a daily ritual, not an intervention. Broad, easy, habitual — consumed every day for the duration of treatment. That is the LTV engine a sharp, churny discontinuation product can never be on its own.

The corpus corroborates the appetite for ritual and guidance: starters describe Mounjaro as "a full personality shift… between protein, hydration, side-effects and constant googling," and creators monetise maintenance demand directly — the recurring "comment PLAN" posts in the data are users selling the very maintenance protocol our second SKU would productise. The demand for a structured daily-and-after system is visible and already being captured by amateurs.

The caution GRUNS also teaches: a daily product needs an emotional reason-to-believe or it is just another greens powder. Ours is supplied by the corpus — the muscle → metabolism → "protect your results" narrative. That story is the difference between a commodity and a brand.

8 · Four product architectures, weighed

Reasoning to the recommendation rather than asserting it. Four candidate heroes emerged from §2–7:

The error to avoid — and one this analysis initially made — is treating these as mutually exclusive and picking the most defensible (B). Optimising for defensibility yields a sharp but shallow business. Optimising for a large daily product everyone on the drug takes (the stated objective) points to A, with B as the natural retention companion and D as a fast-follow. A and B share a formula spine and a single narrative, so combining them is coherent, not scattered.

9 · Recommendation

The blend resolves the central tension. The daily product carries the volume and lifetime value; the maintenance product carries the emotion and defensibility and converts the daily product's main weakness (churn when people feel fine) into a reason to keep buying ("you'll want this when you stop"). One brand, one belief, two stages of one customer. The hair SKU then de-risks further by adding a third, appearance-driven purchase occasion with demonstrated willingness-to-pay.

10 · Formula logic — drawn from what users already reach for

The ingredient rationale is not imposed; it is lifted from the language of the corpus. Users coming off the drug are already stacking these by guesswork — the product simply does it credibly.

11 · Risks & open questions

12 · Roadmap — to close the thesis before committing capital

Annex A · Representative verbatims

Unedited customer voice from the corpus, organised by theme. Doubles as a positioning and ad-copy bank.

On-drug · muscle & body composition

"GLP-1 drugs can lead to reduced bone mineral density and fracture risk, primarily from rapid weight loss — loss of fat and muscle decreases stress on bones, weakening them."

unknown · on-drug · emotion 2/4

"I have lost 18 kg since March… also eating less protein than ideal. I had long THICK hair and have lost 2/3 of it, it's a nightmare. Did you manage to recover the hair and stop the loss?"

unknown · on-drug · emotion 3/4

"Naked, there was no life left in me. I was smaller in clothes but the muscle was gone. Weight training, walking, more fibre, more protein seem to be my future now."

US · post-drug · emotion 4/4

Food noise (the on-drug relief, and its return)

"I lost 70lbs over a year… at 59 it's a game changer. It takes away the food noise and makes me feel better in menopause. My A1c went from 5.9 to 5.0. My chronic pain is gone."

unknown · on-drug · emotion 5/4

"This phase scared me more than starting. Not because of hunger — because of fear. Fear of slipping back into old cycles. The difference this time? I had a plan."

US · post-drug · emotion 4/4

"When hunger returned after stopping, I panicked. My first thought was to go straight back on the jab. Instead I made a maintenance plan and stuck to it — that's when things settled."

UK · post-drug · emotion 3/4

Stopping & regain (the emotional core)

"Three months down the line, yikes. The comfort eating. Even at a fraction of what I ate before — instant pounds. You lose muscle mass, which drives your metabolism."

UK · post-drug · emotion 4/4

"Stopping doesn't mean going back to how things were. I was worse off after stopping than if I'd never used it. It messed with my metabolism so I gain weight even faster now."

unknown · post-drug · emotion 4/4

"Former users regained weight four times faster than those using just diet and exercise. 'What was particularly striking was just how fast it was regained.' (Oxford study, cited by a user)"

US · post-drug · emotion 2/4

"If you're scared of what happens after Mounjaro… you're not alone. Maintenance isn't perfect but it IS possible with the right plan."

UK · post-drug · emotion 3/4

Daily nutrition & the 'full personality shift'

"The Mounjaro beginner experience is truly something else. Between protein, hydration, side-effects and constant googling — it's a full personality shift. Save this so you're not alone."

unknown · starting · emotion 2/4

"Google Telogen Effluvium — hair loss after rapid weight loss, surgery, illness, stress. It'll eventually stop. Add biotin if you're not already, but high doses can cause cystic acne."

unknown · on-drug · emotion 1/4

"Over the past 3 weeks my life has completely changed! Here are some go-to meals I've found helpful — and very yummy."

UK · starting · emotion 2/4

Hair loss (the under-served SKU-2 signal)

"I lost 50 pounds and feel amazing — the downside is I've lost ~50% of my hair and it's dry and brittle. I'm on iron and Lustriva. What else can I do?? 😭"

— · on-drug · emotion 3/4

"My eyebrows started falling out, then my hair. All labs normal. A year later, zero eyebrow regrowth. I've tried minoxidil, zinc, iron for 9 months — only worse."

— · post-drug · emotion 4/4

"Hair loss? Check iron (100–150), vitamin D (>60) and B12 (>500). (Creator self-help post, score 9,460)"

BR · on-drug · emotion 2/4

Annex B · Positioning vocabulary

The exact recurring phrases people use, by concern — the raw material for headlines and ad copy that mirror the customer's own words.

Annex C · Methodology notes & caveats

• Corpus: 13,886 items; 5,653 carry ≥1 pain tag. Sources: Reddit 5,862, YouTube 5,480, TikTok 1,771, Instagram 68.
• Tagging: two-stage model pipeline; census tags applied to all non-risk items, improved to capture lay-language symptoms. ~one pass — recall imperfect; ~45–57% of items legitimately carry no pain (debate, off-topic, pleasantries, ads).
• Emotion scored 1–4 (some 5s in raw data); journey phase one of considering / starting / on-drug / stopping / post-drug / unknown.
• Search volume from Google Keyword Planner (launch-geo basis); treat banding as directional given API tier.
• Geography: 10,145 items geo-unknown — the dominant limitation. Identifiable signal skews US. ES/FR demand is not yet validated by this corpus.
• Safety: medical-emergency, eating-disorder and crisis content is hard-excluded from all outputs by design and does not inform product claims.
• Not medical or regulatory advice. All product/claim decisions require qualified clinical and EU regulatory review.