Context
An established luxury furniture retailer trading as VIG Furniture: three showrooms in Los Angeles, one in Miami, and an online store. Costs were climbing, the departments were working past each other, and the campaigns weren’t paying for themselves. The brief described a marketing problem, which is usually where the conversation starts and usually too late in the story.
Diagnosis
The audit covered sales, marketing, logistics, and IT at the same time. What came back was operational: overlapping systems, duplicated work, KPIs nobody agreed on, and decisions stuck between teams who each assumed the other owned them.
It also produced the number the whole case turns on. Over sixteen months the site drew 926,812 users and produced 809 orders. That’s a conversion rate of 0.087%, one order per 1,146 visitors, on $1.19 million of revenue.
Nearly a million people came and almost nobody bought.
The engagement numbers rule out the obvious explanation. 2.81 pages a session, 2:17 on site, 66% bounce. That’s a visitor who’s looking, not one who landed on a broken page and left. The traffic was interested and the business couldn’t convert it, which is a completely different diagnosis from "we need more traffic", and it points at the offer, the pricing, and everything that happens after the click.
One more thing the audit settled quietly: ecommerce and ERP reported the same revenue to the cent. In most audits those two disagree and the first week goes on reconciling them. Here the data was trustworthy, which meant the problem was real rather than a reporting artefact.
Retention was 4%. In luxury furniture the buyer is expensive to acquire and the second purchase is where the margin lives, so a repeat rate that low means the relationship ends on the delivery date. Average order value said the same thing from the other side: at $253 they were selling one item to people furnishing a room.
The device split was the last piece. 35% Safari against 32% Chrome, 35% iOS against 14% Android or Windows. A majority-Apple, majority-mobile audience buying luxury furniture, and the creative had been built for neither.
Fixing the ads first would have brought more customers into a business already losing money on the ones it had.
Decisions
- Rebuilt the operating model first and the campaign mix second, on clean data
- Replaced four separate departmental processes with one framework
- Rebuilt the workflows around named ownership of each decision
- Rebuilt Facebook, Google, SEO, and email together rather than channel by channel, mapping which proposition ran down which channel rather than tuning each one separately
- Planned promotion against a calendar of retail events, with the same set of questions asked of each one, instead of running always-on
Results
Average order value rose from roughly $240 to roughly $760. Retention moved from 4% to 18%, and the business added approximately 3,800 customers a year.
The rebuilt funnel is visible in the account. Prospecting bought reach cheaply, at $0.75 a landing page view and $0.004 an engagement, and recorded no purchases at all. One retargeting ad set converted that audience, returning $2,371.21 on $149.44 of spend in the same window. Google Ads came out at a 314% return across the rebuild.
Analytics says the same thing from outside the ad account. Between March and May 2023 social users rose 19.94% against 6.55% site-wide, and seven-day active users rose 61.76% while one-day active users fell 27%. The gain was in people coming back, not in first clicks.
The slow part was the argument. Founders want the campaign fixed, because campaigns are visible and operating models aren’t.