The membership, and where it stopped working.

The squad

A small team, shipping daily

Our squad was myself, the PM and six developers, all working in Claude directly in the repo, with no Figma handoff. It worked autonomously, outside the larger organization’s roadmap and goals, and it moved like a startup.

Most HelloFresh teams design every variation in Figma and ship at the end of a four- or six-week sprint, but we challenged ourselves to ship something every day. I could get four or five flows done in half an hour and validate them quickly, where prototyping them in Figma would take me well over a week.

Even the PM can generate a design, which is where I fall back on my 15 years of visual design experience and never take what Claude gives me as the bar. The squad was an incubator that sent ideas back up to the mothership for other teams to take, so everything I designed had to work for every brand.

Brief

A subscription on top of a subscription

The question we were given was vague: what would an Amazon Prime look like for Factor? I wasn’t sure what that meant, because Factor is already a weekly subscription you pay for, so why would anyone want another one? Customers paid for shipping and extra for premium upgrades like a larger steak, so the idea was a membership that got you free shipping and some of those add-ons.

A phone showing the Factor cart after joining Factor Plus, with premium meal upgrades and shipping struck through

Decision

I argued for Rewards. The team chose standalone.

We already had Factor Rewards, which was free to join and gave you free shipping, discounts on upgrades and even a few extras the paid version didn’t. Showing customers two competing programs that look the same, where one is free and one costs money, is a UX nightmare, and obviously I would pick the free one.

I pitched merging them, but the team wanted a membership we could price, forecast and measure on its own. So it was on me to make Factor+ as easy to understand as possible.

Experiment

Not everyone, not yet

The plan was to release Factor+ to everyone, until I modeled a year across every loyalty level. From 11 boxes up it turned negative. Our best customers already order weekly and pay for shipping and premium meals, so each of them would cost the company $641 a year. It was a great deal for the customer and a horrible deal for the business.

With the team moving fast, there's potential for something like this to be overlooked. Luckily I caught it before initial rollout and argued we should be more strategic about who we offered it to. We started with a smaller group against a control group, released in tiers by loyalty level and compared how each level took it, then moved up gradually while keeping the highest levels out.²

Design

Show the saving in this order

In the cart, we A/B tested two ways of introducing Factor+, and option B, which led with today’s saving, won resoundingly. Option A explained what Factor+ is, what it costs and how often we’d charge you. Option B was an impulse buy, like the gum and chocolate bars beside you in the grocery store line, because who doesn’t want to save money?³

Option A, the enrolment module expanded: You could save $41.46 with Factor Plus. Premium meal upgrades $28.47 free, shipping $12.99 free, new total $67.96, then Join Factor Plus — Fig. 01Explain the membership, or show today’s saving?

Option B, the order total with a single checkbox: Save $42.99 on this order with Factor Plus, then Continue — Fig. 01Explain the membership, or show today’s saving?

We A/B tested the join step that opens next as well. Option B led with the dollar amount every four weeks or billed annually. It performed better than option A, a table of what you get with and without Factor+.

Option A, the comparison modal: 4 weeks or Yearly, $24.99, a Without Factor Plus and With Factor Plus comparison, the terms checkbox and Start Factor Plus — Fig. 02Same step, drawn two ways.

Option B, the join sheet: Get more with Factor Plus, save $41.46 on this order, two plan cards at $24.99 every 4 weeks and $99 billed annually, the card on file, and Join Factor Plus — Fig. 02Same step, drawn two ways.

Grubhub and the other food apps all have something similar, so I researched them to make this feel familiar. The offer sits below the total and above the call to action, the most important point in the flow, so I kept it quick enough that nobody gets sidetracked into second-guessing.

01
The offer, in the cart
What joining would save on this order, right above the button that pays for it.
02
One sheet to join
Two plans, the card on file, and the terms in a sentence.
03
Savings, applied
The new total, before the order is placed.

Fig. 03Joining from the cart.

The cart on desktop: five meals on the left; on the right the order total of $109.42, the Save $42.99 on this order with Factor Plus checkbox, and Continue — Fig. 04The same flow on desktop.

The cart on desktop after joining: a Factor Plus confirmation that the membership is now active, with premium upgrades and shipping struck through and a new total of $67.96 — Fig. 04The same flow on desktop.

A big part of the work was making Factor+ show up correctly everywhere in the app, so people knew where to find their benefits and how to cancel, because what it started with is not what it will end with.⁴

Manage Subscriptions in account settings: Factor meals and the Factor Plus membership listed as active subscriptions, each with its own manage button, and Factor Form under cancelled subscriptions — Fig. 05Factor+ in account settings, next to the meal subscription.

The Factor Plus membership page: annual plan, member since March 15, 2026, $99.00 a year, next billing date, $156.00 saved, and the benefits included — Fig. 05Factor+ in account settings, next to the meal subscription.

Factor+ membership hub showing active member status, free shipping usage, premium meal credit, and benefit expiry — Fig. 06The membership hub as shipped, the one place for benefits that apply automatically: what’s included, used and saved, and when benefits expire.

When someone tries to cancel, these screens show the savings they still have.

The first cancellation screen: Hold up, you still have benefits to use, showing $131.96 saved since joining, a Keep my bonuses button and a Cancel Anyway button — Fig. 07Leaving, in three screens: what you’ve saved, one question, then a confirmation that billing stops and benefits run to the end of the term.

A short survey sheet asking why you are canceling, with Not using the benefits selected, and buttons to go back or submit and cancel — Fig. 07Leaving, in three screens: what you’ve saved, one question, then a confirmation that billing stops and benefits run to the end of the term.

Results

The results supported a wider audience

Customers offered Factor+ placed 30.5% more orders and brought 12.6% more predicted value per customer, which surprised me, because I thought a subscription on top of a subscription would be too much. Free shipping is always a hurdle, so once you have it, ordering feels easier, especially with those premium upgrades.⁵

Cancellations fell 7.4%, because it’s harder to cancel when you’re getting really good benefits, and one in 10 customers joined. The number that really shocked me was 45,609 members by week 34, roughly 9% of all active customers.⁶

We targeted people with two to 10 previous boxes, because the higher you go, the more you get into cannibalization, so we paused the highest loyalty level and took the rest group by group.

Predicted value vs control

+12.6% −1.3%¹

Predicted net value over 52 weeks per customer offered Factor+, by previous order history. LL indicates previous box count.

LL2–10+12.6%Target cohort
LL10–15+5.0%Expand
LL15–20+3.7%Expand
LL20–30+4.5%Expand
LL30–50+1.7%Monitor
LL50+−1.3%Paused