Adjacency expansion: moving into the business next door, and what you carry across

The mirror on the wall

In June 2020, with gyms shut, Lululemon agreed to pay $500 million for Mirror, a startup whose whole product was a sheet of glass that hung on your wall and showed you a fitness instructor. Lululemon sold athletic apparel to a community that met in studios. Selling that community a home workout device looked like one step to the side.

Its chief executive was clear that this was not about clothes.

“It's not an acquisition simply to sell more apparel.”Calvin McDonald, Lululemon's chief executive, as recorded in the Lululemon case

The thesis was that the community in a yoga studio would flow through a screen. It didn't. In December 2021 Lululemon cut its Mirror sales forecast for the year to $125 to $130 million, from $250 to $275 million. Impairments and write-downs followed, about $443 million after tax, and in September 2023 it signed a partnership with Peloton and said it would stop selling the hardware.

Compare a move that looked stranger at the time. At a 2003 executive retreat, Amazon's leaders realized that running reliable, scalable computer infrastructure had quietly become one of the things the company was best at. It was plumbing built to run a shop. In March 2006 it started selling that plumbing to others. In 2024 Amazon Web Services earned $39.8 billion in operating income, over half of Amazon's $68.6 billion total.

Lululemon carried its customers next door and had to buy the skill. Amazon carried a skill it used every day and found new customers for it. That difference, what you actually carry across, is what this guide is about.

What adjacency expansion is

That something can be a skill, a customer base, a brand or an asset like stores or a network. The word that matters is next door. A move into an unrelated business, with nothing carried across, is diversification, which is a different bet.

Why it is so hard

On paper the neighbor's business looks like a short walk. In a real company, three things make it longer.

The ways companies did it

The moves on record take four shapes. Each carries something different across the fence.

What 98 companies did

Here is every decision on record, placed by the year it was made and marked by how it turned out. Hover a mark, or open the table, for the company and the outcome.

Of the 91 moves that can be judged, 76 were never undone, 7 were walked back and 8 were kept only in part.

Many of the moves that held are acquisitions that closed and stayed, and a deal that closes is not yet a deal that works. Crocs kept HeyDude after paying about $2.3 billion for it, and in 2025 took $738.1 million in impairments on it.

The walk-backs

Put the walk-backs side by side and a pattern shows. Most of them bought access to customers the company already had: shoppers, subscribers, store visitors, luxury buyers. Each then needed a skill it did not have, from running a brokerage to running a doctor's office.

A move that stood is not always a move that worked, so a few hard cases are not in this table. JPMorgan never undid its purchase of Frank, though it later sued the founder over a customer count that proved false.

The shared-customer cases fail in a similar way. Sears had shoppers; it needed brokers. Lululemon had a community; it needed a hardware business. Best Buy had installers in customers' homes; it needed hospitals and insurers to pay. Walgreens had store traffic; it needed doctors, and planned to close 160 clinics. Richemont had luxury buyers; it bought a marketplace that competed on selection and price.

The moves that held tended to carry the skill itself. Amazon sold the infrastructure it ran its own shop on. Nvidia opened up chips it was already making. Honda took its engines into the air, slowly: seventeen years passed between the first four engineers and the first flight, and a further twelve before a customer took delivery.

Customers can be bought in a day. The skill to serve them in a new business cannot.

What the share price did

For many of the decisions there is a clean market record: the company's total return over the three years after the decision, against the S&P 500.

The chart is a warning against reading a single move through a share price. Apple's 2003 iTunes Store sits at the top, 868.6 points ahead of the index. But JPMorgan's purchase of Frank ended in a lawsuit against the founder and its shares still beat the index by 21.6 points, while Pfizer's Seagen deal was never undone and its shares trailed by 97.3. Over three years the price mostly reflects the whole company, not the one move.

How to run this decision

If you are weighing a version of this choice, here is what the record suggests doing, in order.

1. Name what you are carrying across, in one word

   A skill, a customer base, a brand or an asset. If the honest answer is customers, the skill to serve them in the new business has to come from somewhere, and that is where most walk-backs on record went wrong.

   Seen in: Lululemon, 2020 · Amazon, 2003
2. Use the skill yourself before you sell it

   Amazon ran its own shop on the infrastructure it later rented out. Nvidia was already making the chips CUDA opened up. A capability you rely on every day has been tested by the hardest customer you have, which is you.

   Seen in: Amazon, 2003 · Nvidia, 2006
3. Check the number you are paying for

   JPMorgan bought Frank on a claimed customer count of 4.25 million; the real figure was fewer than 300,000. Before you pay for customers, a skill or a market share, test it yourself rather than taking the seller's word.

   Seen in: JPMorgan Chase, 2021
4. Stage the bet, and let each stage pay before the next

   Walgreens began with five clinics in a trial, then committed $1 billion for 30% of VillageMD, then $5.2 billion for control before the first money had fully landed. Decide in advance what each stage has to show before the next one is paid for.

   Seen in: Walgreens, 2021
5. Watch the core's numbers while you build next door

   Sears's stores were already lagging when it bought into finance, and lagged further while it owned the firms. Lego nearly ran out of cash after spreading into parks, games and jewelry. Put the core's margin next to the new business's progress in every review.

   Seen in: Sears, 1981 · Lego, 2004
6. Decide when you would sell before you buy

   McDonald's bought Dynamic Yield, learned what it needed and sold it 36 months later at a $271 million pre-tax gain. Kroger kept the fuel centers that fed its stores and sold the ones that did not. Write the exit test down at the start.

   Seen in: McDonald's, 2019 · Kroger, 2018

To work the numbers for your own case, read Adjacent market expansion: how to test what actually transfers.

Warning signs

What it has looked like when this decision was going wrong.

• The case for the move is the customers you share, not the skill you bring.Lululemon, 2020 · Walgreens, 2021
• You are buying because the target is for sale, not because you had a plan.Kodak, 1988
• The number the price rests on comes only from the seller.JPMorgan Chase, 2021
• The core's margins are slipping while attention goes next door.Sears, 1981
• You are raising your stake before the first one has paid.Walgreens, 2021

All 98 cases

Each case is one company at one decision: what was on the table, what it chose, why, and what followed, with every fact sourced.

Questions

What is adjacency expansion?

Moving from a company's core business into a neighboring one, such as a new product, a new kind of customer or a step up or down the supply chain, on the bet that something from the core, a skill, customers, a brand or an asset, gives it an edge there.

What are examples of adjacency expansion?

Amazon selling its computing infrastructure as Amazon Web Services, Nvidia opening its graphics chips to data centers, Honda building a business jet, and Philip Morris buying Swedish Match. Lululemon's purchase of Mirror was walked back, and Walgreens's bet on VillageMD clinics was kept only in part.

How is adjacency expansion different from diversification?

An adjacency carries something from the core business into the new one, a skill, customers, a brand or an asset. Diversification moves into a business with nothing in common. Kodak's purchase of a drug company is closer to diversification than to adjacency.

Why do adjacency moves fail?

On record, most walk-backs bought access to customers the company already had and then needed a skill it lacked, such as Sears in brokerage, Lululemon in fitness hardware and Walgreens in primary care. Others rested on a deal that was available rather than planned, as Kodak's purchase of a drug company did.