Nagesh Koritala, in full color, wearing a forest-green knit quarter-zip over a white collared shirt and holding a navy member card, beside a cut-paper collage: a tall sign-up bar crossed out in ink, a sofa in a delivery box, a calendar page with one day circled, and a slowly rising line circled in red.

Writing

Vanity metrics vs. actionable metrics

A made-up home goods membership shows why sign-ups and member sales look great on a launch slide, and what to check instead: what’s left after discounts, and who renews.

EssayNovember 2026About a 9 minute readBy Nagesh Koritala

Watercolor: a launch review. A team leader points proudly at one very tall blue bar on a wall screen while two teammates smile up at it; at the end of the table a colleague quietly holds up a small card with a gently rising red line.
Everyone is looking at the tall bar. One person brought a different chart.

You’ve probably sat in this meeting. A new membership goes live, and the first slide is all good news.

To show what I mean, I made up a home goods store called Larkfield and gave it a paid membership, Larkfield Plus. In its first month, 40,000 people join. Members place $4 million in orders. Everyone in the room is happy.

Larkfield Plus · launch month

40,000people joined

$4Min member orders

Illustrative

Then someone asks three questions. How much of that $4 million does Larkfield keep? How many will still be members next year? And how many of those orders would have happened anyway?

Sign-ups tell you people noticed the offer. To know it’s worth paying for, look at what’s left after the discount, and who renews and orders again. When I look at a membership, I pair every total with a number that can fall.

Most of those numbers are rates, the same idea as conversion rate: out of a group, the share who did the next thing, like new members who ordered again.

Try it first

Which number would tell you Larkfield Plus is working?

The first two count activity, and running totals only go up. Member orders are also counted at full price. The last two are rates that can fall, so they can warn you.

Where the $4 million goes

Start with the money. The $4 million on the slide adds up every member order at full price, a total often reported as gross merchandise value, or GMV. But members don’t pay full price. A lower price is a big reason they joined.

Here’s one made-up Larkfield order: $180 at full price, with 10% off and free shipping for members.

One member order · $180 at full price

Illustrative
  1. What the launch slide countsthe full price$180
  2. What Larkfield collectsafter 10% off$162
  3. What Larkfield keepsafter the products and free shipping$44
  4. Kept from a non-memberthe same order, full price$72
Made-up numbers: a 40% product margin, so the products cost Larkfield $108, and $10 of shipping that Larkfield pays for members. A non-member pays full price and their own shipping. Both are before returns and other order costs.

The slide counts $180. Larkfield collects $162 and keeps $44, after the cost of the products and the shipping it now pays for. If a non-member placed the same order, Larkfield would keep $72. The member price cuts what Larkfield collects by 10%. Add free shipping, and what it keeps on that order drops by almost 40%.

If every launch order looked like this one, the $4 million on the slide would be about $3.6 million collected and a little under $1 million kept. The $79 fees come on top. They’re there to help pay for these discounts, and the lifetime value table below counts them.

GMV also means different things at different companies. Some count it before discounts, others after discounts but before returns and cancellations. So when someone shows me GMV, I ask one question first: before or after discounts, returns, and cancellations?

Money collected after discounts is closer to the truth, but it’s still a total. It can’t tell you whether those orders needed a discount at all.

Two kinds of numbers

Eric Ries calls the first kind vanity metrics. In a 2009 post he named total hits his personal favorite example, because a total like that can go up for reasons nobody can trace. He calls the second kind actionable metrics: numbers broken down far enough that you know what to do differently.1

Totals that count activity

People who joined, member GMV (members’ orders, often counted at full price), savings given away, app downloads. They tell you how many.

Rates that show behavior

Who kept it, who ordered again, who renewed, and what each member is worth against what they cost. They tell you what people do.

Totals still matter. A membership nobody tries can’t grow. But a running total like sign-ups only goes up, and it can’t tell you why it moved. If you’re newer to this, my lesson on picking the right metrics covers the basics: vanity and actionable metrics, cohorts, and guardrails.

How I’d design Larkfield Plus

Watercolor: a woman in a sky blue sweater kneels beside a large delivery box in a sunny living room, folding back the wrapping from a new mustard yellow sofa. A navy member card and a coffee cup sit on the side table.
A member furnishing a home, one big piece at a time.

Home goods aren’t groceries. According to EMARKETER, 64% of shoppers buy home furnishings less than once a month.2 So a home membership has to earn its fee across a few orders a year, not a few a week.

Two real programs show the range. RH charges $200 a year for 30% off full-price items.3 In its fiscal 2023 report, members accounted for about 97% of sales in its core business.4 Wayfair Rewards costs $29 a year for 5% back and free shipping. Wayfair says it set that price to be “effectively breakeven” for a customer spending about $600 a year.5

Larkfield Plus sits between them, at $79 a year

Illustrative
  1. Member prices

    10% off full-price items, on every order.

  2. Free shipping

    On every order, plus $50 off white-glove delivery for furniture.

  3. Easier returns

    60 days instead of 30, with free pickup for large items.

  4. Design help

    One free 30-minute room styling call a year.

  5. First look

    48 hours of early access to new collections and sales.

I’d price it so it pays for itself for a household that orders about three times a year. At a made-up $180 average order, 10% off saves about $54, and free shipping about $30. That’s $84 back on a $79 fee, before returns and design help.

One member’s year · three orders

10% off three $180 orders$54

Free shipping on three orders$30

Saved$84

Membership fee$79

Pays for itself, before returns and design help. Illustrative.

I’d leave two things out. No points, because two kinds of savings make people do math at checkout. And no two-day promise, because a sofa can’t keep it.

Paid from day one, with a 30-day refund. I’d skip the free trial. I think Amazon Prime can offer one because people use it so often. CIRP reported in 2022 that more than two-thirds of Prime’s free trial members became paying members.6 Someone furnishing a living room is different. Their trial often ends before a second order, so it tells you nothing and makes the sign-up total look bigger than the business. The refund takes back any member savings already used, so nobody can join, buy a sofa at the member price, and cancel.

Either way, the renewal has to be clear. Since July 2025, California’s automatic renewal law requires the shopper’s express consent to the renewal terms, whether the plan starts as a free trial or paid.7 At checkout, I’d show it like this:

Larkfield checkout Illustrative

Larkfield Plus · $79 a yearYou save $28 on this order ($18 off, plus free shipping)

Renews every year on your join date until you cancel. Cancel online anytime in two clicks. Full refund within 30 days, minus any member savings you’ve used.

The box starts unchecked. Joining waits until the shopper agrees.

The price and renewal date show up before anyone joins, and consent gets its own unchecked box. Canceling takes two clicks, and a reminder before renewal shows what each member saved.

The numbers that decide

Now follow the same made-up launch group for a year, as each number arrives:

  • Day 3085%kept it after the 30-day refund window. That’s 34,000 members, not 40,000.
  • Month 12+1extra order a year per member, compared with similar shoppers who didn’t join.
  • Month 1250%renewed a year later, in the group that joined in the launch month.
  • Day 30$45to win each member who kept it, called the customer acquisition cost (CAC). That’s $1.53 million in launch marketing, divided by 34,000.

Watch what you divide by. Divide the same marketing by 40,000 sign-ups and each member seems to cost about $38. Only the people who stayed count.

Watercolor: a product manager with a long braid points a red marker at a whiteboard with two columns of small house cards; one column has an extra card circled in red. A data analyst in a wheelchair beside her shows a chart on his tablet.
Two groups of similar shoppers. The circled card is what membership changed.

Watch what membership changed. Members place three orders a year. Similar shoppers who didn’t join place two. So membership changed one order, not three. Here’s what worries me. Regulars join first, because they save the most on orders they were going to place anyway. Wayfair said as much on its own earnings call, where early sign-ups were “weighted towards recurring Wayfair shoppers.”5

Count lifetime value by what changed. Lifetime value, or LTV, is what one member is worth over the whole relationship. With made-up numbers (a 40% product margin and $10 of shipping per order), switch between the two ways of counting:

$134lifetime value over two years

About 3 to 1against $45 to win each member

Lifetime value of one Larkfield Plus member, counted two ways (made-up numbers)
Each yearCounting every orderCounting what membership changed
Membership fee$79$79
Order profit, after the member price$162 (three orders)$54 (the one extra order)
Discount on orders they’d have placed anywaynot counted−$36
Free shipping−$30−$30
Value a year$211$67
Over two years (LTV)$422$134
Why two years: with half of members renewing each year, the average stay is one year, plus half a year, plus a quarter, and so on, which adds up to two. It leaves out the cost of returns, the white-glove credit, and design calls, so the real number is a little lower. Illustrative numbers.

Counting every order, LTV is more than nine times what each member cost to win. Counting what membership changed, it’s about 3 to 1, or $134 against $45. David Skok, whose guide to subscription software (SaaS) metrics is widely cited, says the best subscription software businesses have a ratio higher than 3, and calls his numbers “only guidelines.” He also watches how long it takes to earn back what a customer cost, and calls profitability anemic past 12 months.8 Spread over the year, Larkfield earns its $45 back in about eight months. In cash it’s sooner, because the $79 fee is paid up front.

Try it: change the renewal rate

2.0 yearsaverage stay

$134lifetime value

3.0 to 1LTV to CAC

Right at Skok’s guideline of 3.

Each member is worth $67 a year, counted by what membership changed. The average stay is 1 ÷ (1 − the renewal rate). Most teams also cap lifetime value at three to five years and discount later years; this keeps it simple. Illustrative numbers.

Now drop renewals. If 30% renew instead of 50%, the average stay falls to about 1.4 years, LTV to about $96, and the ratio to about 2 to 1. The sign-ups and member orders haven’t changed, but the business has.

What real programs report

The companies that run memberships at scale report the rates, not only the totals.

  • 92.3%Costcorenewal rate in the US and Canada, end of fiscal 20259
  • 15%+Wayfair Rewardsof US revenue, from more than 1 million members5
  • ~97%RHof core sales came from members in fiscal 20234

Costco publishes its renewal rate every year. Wayfair also says the program costs it some gross margin, the profit left on each order after product costs, and that lower advertising spend more than makes up for it, because members come back on their own.5 So it judges the program on profit after those costs.

RH shows the other edge. When members account for about 97% of core sales, member GMV is almost the same as core sales. At that point the membership is the price, and the useful questions are renewal and margin.

How I’d put them to work

  1. Report what you keep

    Show member sales after member prices, free shipping, and returns, not only at full price. And say how GMV is counted.

  2. Pair every total with a rate

    Sign-ups with the share kept after 30 days. Member orders with the share of members who order again within 90 days.

  3. Read groups, not grand totals

    Follow everyone who joined in the same month, called a cohort. A grand total blends this month’s eager regulars with last year’s quiet leavers.

  4. Ask what membership changed

    Compare members with similar shoppers who didn’t join, or hold back a small group from the offer if you can.

  5. Keep LTV to CAC and payback in view

    CAC is what it takes to win one member who stays. Compare it with what that member is worth, and count the months it takes to earn that cost back.

  6. Set guardrails

    Watch return rates, full-price margin, cancellation complaints, and how often non-members still buy. A membership that wins members by losing everyone else isn’t working.

What I take from it

Watercolor: a year later, the same woman sits comfortably on her mustard sofa, now with cushions, a throw, a lamp, and plants, smiling at her phone with a coffee in hand. A calendar on the wall has one day circled in red.
A year later, the renewal email arrives. This is the moment the launch slide couldn’t show.

The launch slide wasn’t wrong. 40,000 people did join. The slide answered a different question: did the offer get noticed?

The questions that decide are the three from that meeting: what Larkfield keeps, who’s still a member next year, and which orders would have happened anyway.

If you run a membership, try one thing this week. Take everyone who joined in a single month, and count how many ordered again within 90 days. That one number will tell you more than the launch slide.

Which number on your launch slide needs a second number beside it?

Written November 2026. Larkfield, Larkfield Plus, and every Larkfield number are made up to show how the metrics work. The research and company figures come from the sources below. The program design and its reasoning are my own, and none of this is legal or financial advice.

Sources

  1. Eric Ries, Why vanity metrics are dangerous (opens in a new tab), Startup Lessons Learned, December 23, 2009. He develops actionable metrics further in The Lean Startup (2011). ↑
  2. Radial, Key Consumer Trends in the Home Furnishings Industry (opens in a new tab), citing EMARKETER. Undated page, read November 2026. ↑
  3. RH, Frequently asked questions (opens in a new tab). Member fee and savings as listed in November 2026; RH may change them. ↑
  4. RH, Form 10-K for fiscal 2023 (opens in a new tab). Members accounted for about 97% of sales in the core RH business, with about 281,000 members at year end. ↑
  5. Wayfair, fourth quarter 2025 earnings call transcript (opens in a new tab), February 19, 2026. Company statements about Wayfair Rewards. ↑ ↑ ↑
  6. CIRP, Amazon Prime Retention Rates are Steady and Very High (opens in a new tab), August 2, 2022. ↑
  7. Barnes & Thornburg, California Expands Automatic Renewal Law (opens in a new tab), 2025. A summary of AB 2863, in effect from July 1, 2025. ↑
  8. David Skok, SaaS Metrics 2.0 (opens in a new tab), For Entrepreneurs. The LTV to CAC guideline and the 12-month payback. ↑
  9. Costco, Form 10-K for fiscal 2025 (opens in a new tab). Renewal rate in the US and Canada at the end of fiscal 2025. ↑

Say hello

Disagree with
something here?

Tell me where and why. Pushback makes the next piece sharper.

Nagesh Koritala in a grey sweater over a white collar, holding up a printed page covered in red-pen notes, smiling, beside cut-paper essay pages, two speech bubbles with a question mark, and a pencil.