Building Factor's Weight-Loss Programme
Customers were manually building a weight-loss programme out of a meal delivery service. We gave them the structure they were already trying to create — and a premium tier worth paying for.
- Factor had millions of customers already using the service for weight loss — manually, with no product support. The company had never entered the weight-loss market.
- Built a 0-to-1 programme from strategy through launch: personalised caloric planning, curated meals, premium pricing. No competitor combined all three.
- Programme AOV of $202 vs $120 Classic — driven by additional meal occasions and structured guidance, not just price. Conversion gap closed from -13% at launch to neutral.
- Key decision: merged two planned phases into one MVP after research showed Phase 1 without meal swap would fail the experience bar — a two-month delay that avoided a likely failed launch.
Context
Factor, HelloFresh's ready-to-eat brand, served millions of active customers. Internal research — behavioural data from meal filter usage, menu category selection patterns, and purchase behaviour — showed 52% of them were already using the service for weight loss. They were manually sifting through 100+ meals and add-ons each week to find calorie-appropriate options across four daily meal occasions. They had built themselves a weight-loss programme out of a product that was never designed to be one.
The company decided to enter the weight-loss market with a structured, programme-based offering — initially conceived as a proof-of-concept for a standalone weight-loss brand, later repositioned as the primary market entry vehicle. I took ownership in December 2024. There was a high-level vision prototype and a strategy doc, but no detailed requirements, no workback plan, and no data brief. My job was to define the MVP strategy, align six-plus cross-functional teams, and take the product from zero to market.
The Problem
Factor had a customer base with a clear weight-loss motivation and no product to support it. Customers wanting to lose weight had to evaluate 100+ meals weekly against their caloric targets, manually planning across meals and add-ons for each day. No structured guidance. No calorie tracking. No curated plans. Just a meal delivery service that happened to serve health-conscious food. And the company had never competed in weight loss. Every established player had years of positioning.
| Missing capability | Impact |
|---|---|
| No personalised funnels | Could not tailor sign-up to weight-loss goals |
| No programme-specific pricing | No premium tier to capture additional value |
| No goal-oriented experience | No meal plans, no dashboard, no calorie tracking |
| No legal-approved health claims | Could not market as a weight-loss product |
The initial timeline called for two phased launches in H1 2025.
Insights
52% were already doing it themselves
52% of active customers were already using Factor for weight loss — visible in how they filtered meals, which categories they gravitated toward, which SKUs they reordered. The demand was not theoretical. What was missing was any product that supported what these customers were already trying to do. They had the motivation. Factor had no structure around it. The question: could we build something worth paying a premium for?
Nobody combined the food with the plan
A competitive analysis of 14 weight-loss services revealed two camps:
| Category | Examples | Strength | Gap |
|---|---|---|---|
| Personalised programmes, no food | Noom, WeightWatchers | Structured guidance, tracking | No meal delivery |
| Meal delivery, no personalisation | Trifecta, BistroMD, Sakara | Prepared food, convenience | No caloric planning or goals |
No product combined structured meal delivery with personalised caloric planning. Factor could sit at the intersection. But providing caloric recommendations meant navigating legal constraints around health claims, nutritional guidance, and menu variety that a standard meal service never had to face.
The original two-phase plan would not survive contact with reality
The initial plan split the work into Phase 1 (April: basic programme with pre-assigned meals, no meal swap) and Phase 2 (June: meal selection, dashboard, logging). When Phase 1 designs went through stakeholder and UXR feedback in December 2024 and January 2025, multiple inputs flagged the same problem: launching without meal swap — the ability for customers to choose their own meals within the programme — would produce an experience that fell below the floor customers expected. Research described it as removing the one thing that made Factor appealing (choice) in exchange for a benefit (structure) that the first version could not yet deliver convincingly. The physical product team's evolving assortment strategy also no longer matched the original scope's assumptions about meal availability. I proposed merging both phases into a single June MVP with meal swap included. A two-month delay versus the original Phase 1 date — but it avoided launching a version that had a high probability of failing the experience bar.
Legal went from blocking us to enabling us
Weight-loss claims required legal review, nutritional sign-off, and carefully worded disclaimers. The starting position was restrictive: legal required disclaimers including 'not a weight loss program' — directly contradicting the product positioning. Through iterative negotiation with legal and nutrition teams, we moved from that starting position to an approved claim: 'Lose 7 pounds in 8 weeks' — along with a calorie calculator in the funnel and personalised plan recommendations via a quiz. Each approval unlocked product capability. Legal became a design partner, not a gate.
Why we skipped lock-in
Early scoping explored requiring customers to commit to a multi-week programme duration — the standard model for weight-loss products. Three signals pushed us away from it: no competitor with food delivery had succeeded with long-term lock-in; qualitative feedback showed prospects were unwilling to commit without tasting the food first; and Factor's menu variety did not yet support the allergen/exclusion coverage needed for confident multi-week commitment. A previous experiment with pay-in-advance pricing had shown improved cancellation rates but no meaningful AOR change and significantly negative net revenue per customer. We bet on flexibility. Lock-in might have protected retention numbers, but the product was unproven — asking customers to commit before they had tasted the food felt like the wrong sequence.
Hypotheses
- A structured programme with personalised caloric planning and curated meal selection would convert existing self-service weight-loss customers into a higher-value subscription tier — because the unmet need was structure, not food quality.
- Allowing meal swap within the programme (rather than pre-assigning all meals) would be critical to conversion and retention — because customers chose Factor for choice, and removing it would undermine the core value proposition.
- The unique positioning of personalisation plus convenience plus food delivery would justify a premium price point — because no competitor offered all three, and the competitive analysis showed the gap was real.
- A flexible subscription (no lock-in) would outperform a commitment-based model at the MVP stage — because trust in the product needed to be earned before duration could be demanded.
What We Built
The modified MVP launched in mid-July 2025. It consolidated what had originally been planned as two separate phases into a single release, including capabilities that the original Phase 1 had explicitly scoped out.
Sign-up funnel with personalisation
Built a programme-specific sign-up flow with two calorie-tier options (covering approximately 80% of the weight-loss customer base per nutritionist guidance), dietary preference capture, and a calorie calculator. The funnel was designed to progressively increase commitment. Early data showed that customers who reached the summary step were converting at only 46%. We iterated on reassurance and price communication at each step, improving that micro-conversion to 55%.
Meal planning with customer choice
Customers saw a daily plan matched to their caloric target across four meal occasions, with the ability to swap individual meals. This was the feature that almost got cut. The implementation was harder than scoped. Most eligible add-ons were multi-serve SKUs not designed for single-meal-occasion use. We worked through multiple options with assortment and fulfilment teams — bundles, smaller serving sizes, limiting swap to lunch and dinner only. The final trade-off: we constrained swap eligibility to meals where the fulfilment pipeline could handle single-serve allocation, which covered the majority of selections but left some add-on categories unavailable for swap.
Customer dashboard
Introduced a pared-down dashboard that was originally planned for Phase 2. During design review, we identified that adding both a week-view navigation and a separate dashboard would make the app unnecessarily complex. Simplified to critical information only and integrated into the existing navigation flow. Post-launch UXUM score: 8.1.
Programme-specific pricing and fulfilment
Developed a premium pricing tier reflecting the additional value of structured guidance, integrated with the existing Factor fulfilment pipeline. Coordinated with SCM teams on tray processing logic and delivery scheduling specific to programme orders.
Results
Conversion recovery
The programme launched at 10% rollout in mid-July 2025. The first week was rough. A -13% conversion rate drop versus Classic Factor, made worse by a misconfigured feature gate that blocked the first users from meal selection and delivery problems — 25% of boxes experienced late delivery, with meals and add-ons arriving on separate days and items not packaged properly.
The technical bugs were fixed immediately. The conversion gap took longer. We iterated on the funnel to ensure users entering the programme flow were more intentional, refined reassurance and price communication at key steps, and adjusted the active meal selection experience based on friends-and-family tester feedback.
| Lever | Before | After |
|---|---|---|
| Classic option selection rate | 25% | 34% |
| Summary-to-Registration micro-conversion | 46% | 55% |
| Conversion gap vs Classic | -13% | -8.6% after first round, then neutral |
Post-launch data deep dives identified further subsets of customers who were best fits for the programme. Entry points were redesigned to target these segments.
Revenue and retention
Programme subscribers showed an AOV of $202 versus $120 for Classic Factor — a +68% increase. The higher price reflected the additional meal occasions and structured guidance in the programme tier. Blended AOV improved +1.8%. Customer loyalty increased +4.3%. The retention picture was more nuanced. Predicted AOR was 4.6 — below Classic's 5.3. But the higher AOV meant net customer value was roughly comparable: the programme broke approximately even on CCV despite the shorter predicted lifetime. This is based on early data — weeks, not months — so the AOR prediction carries real uncertainty. Whether retention converges as the active experience improves is the question the next phase needed to answer.
Strategic pivot
The original H2 plan was to launch a second programme (GLP-1 support) in November 2025. Post-launch data told a different story: the active experience — meal selection friction, meal plan usability, logging, educational content — was where the retention gap lived. Launching a second programme on a foundation that still needed strengthening would have spread the team thin across two undercooked experiences. I presented retention and engagement data to senior leadership. The strategy shifted from launching programme #2 to improving programme #1. Stakeholders who had already committed to the GLP-1 timeline in their own planning needed convincing. The data made the case: depth first, breadth later.
What Didn't Work
The conversion rate drop at launch was steeper than anyone had modelled. The -13% impact was outside the range of scenarios we had prepared for. The first week's technical and delivery issues made it worse. The team iterated rapidly, but the early period exposed a gap in the launch plan: conversion monitoring should have been instrumented and rehearsed before rollout, not built reactively once the drop was already visible.
The meal swap implementation was harder than scoped. The add-on assortment was designed for multi-serve consumption, not single-meal-occasion programme use. The hardest trade-off: whether to limit swap to lunch and dinner only (simpler fulfilment, worse experience) or to solve single-serve allocation across all meal occasions (better experience, more downstream complexity). We shipped a working solution, but the underlying product-catalogue mismatch remained a source of friction that would need structural resolution for the programme to scale.
The lock-in decision left retention exposed. Choosing flexibility over commitment meant we had no lever to protect retention beyond the quality of the experience itself. AOR of 4.6 was sufficient to break even but sat below Classic's 5.3. Whether a commitment model would have changed this is an open question — but every week was a retention challenge that the product had to win on its own merits.
What's Next
The AOR gap (4.6 vs 5.3) was the biggest open question. The hypothesis: improving the daily active experience — meal logging, educational content, reduced selection friction — would close it by making the programme a daily habit rather than a weekly delivery. The programme had also established the technical foundation for cycle-based offerings (e.g., 30-day reset, 8-week plans) and integration with external health tools (Apple Health, MyFitnessPal) — capabilities scoped for 2026 that would extend Factor's positioning from meal subscription to goal-driven lifestyle platform.
My Role
Owned the programme from strategy through launch and post-launch iteration. Translated a high-level vision into detailed requirements, a workback plan, and a data brief — none of which existed when I took over. Managed the product through three SLT reviews and ongoing steering committee alignment.
- Merged Phase 1 and Phase 2 into a single MVP — accepting a two-month delay to avoid launching a version that research indicated would fail the experience bar. This was controversial: it meant the team would have no in-market signal until June instead of April. Before taking it to the commercial team, I aligned the day-to-day stakeholders in fulfilment and operations — they had their own concerns about the Phase 1 scope creating downstream problems, so they were easy to get on side. By the time the commercial meeting happened, the operational case was already built. I argued that a failed Phase 1 would cost more than a delayed MVP, and that the Phase 1 scope without meal swap would produce data on a product nobody would actually use.
- Led the strategic pivot from GLP-1 to improving the existing programme. The COO wanted GLP-1; the MD did not want to rush into it. I prepared a doc with the retention and engagement data showing depth would produce more value than breadth, and first met with the MD and COO together to align on the evidence before bringing it to the CEO. The data made the case — programme #1 needed strengthening before launching programme #2 — but the sequencing of the conversations mattered as much as the analysis.
- Turned legal from a blocker into a product lever. Iterated from restrictive disclaimers ('not a weight loss program') to an approved claim ('Lose 7 pounds in 8 weeks') that the funnel, CRM, and landing page could all use.
- Ran the conversion recovery. When the -13% drop hit, I identified the highest-leverage fixes — Classic option visibility, summary-step reassurance copy, price communication — and tracked each through to impact. The first round of changes brought the gap to -8.6%; further iteration brought it to neutral.
- Coordinated across eight teams (physical product, supply chain, CRM, marketing, legal, nutrition, data science, conversions) through an initiative where most of them had competing priorities.
- Mentored a junior PM (Alison) through the initiative. We worked together on breaking ambiguous asks into things a squad could build against. She was promoted to Senior PM by the end of the programme.