Sell through retailers, dealers and agents, or sell to customers yourself.
8 of 12 judged decisions were never undone. The big retreats, Nike's and Macy's, raced toward a target set before the results came in.
The middleman's cut pays for traffic and stock as well as shelf space. Going direct means paying for all three.
The partners Nike cut, and hired back
On June 15, 2017, Nike announced a reorganization built around selling to its customers itself, through its own stores, website and apps, instead of through the shops that carried its shoes.
The plan, Nike said, was designed to let it
“better serve the consumer personally, at scale”Nike, announcing its Consumer Direct Offense, as recorded in the Nike case
Selling direct meant keeping the margin a retailer would have taken, owning the data on every buyer and controlling how the brand was shown. By September 2021 Nike's finance chief said it had "exited about 50%" of its retail partners, from Zappos to Dillard's. In September 2022 Nike's inventory ballooned 44%. In fiscal 2023 its direct business brought in $21.3 billion of $51.2 billion in revenue, about 42%, against a target of 60%.
By March 2024 the same finance chief said the strategy had added "complexity and inefficiency". Nike went back to its wholesale partners, and in 2025 it returned to Amazon after a six-year absence. In the second quarter of fiscal 2025, Nike's North American wholesale sales grew 24% while its direct sales fell 10%.
That is the whole decision in one story. The retailers' cut was easy to see, and the plan was built to keep it. What the retailers did for that cut, bringing in shoppers and holding stock, was harder to see until it was gone. The rest of this guide is about the companies that faced a version of that choice, and what happened next.
What going direct is
Going direct means selling to your customers yourself, through your own stores, website or salespeople, instead of through the retailers, dealers or agents who sit between you and them. The trade calls it direct-to-consumer, or DTC.
The middlemen take a cut, and keeping that cut is usually the point of a direct plan. But the cut pays for things: shelf space, shoppers who walk past it, and stock held in someone else's warehouse when demand drops. Going direct means paying for those yourself, or doing without them. It is rarely all or nothing. Most companies in this guide ended up running more than one channel at once.
Why it is so hard
On paper the case is simple: keep the margin the middleman takes. In a real company, three things make it harder.
The margin shows; the traffic doesn't
The partner's discount is on every invoice. The shoppers that partner brings you are on none of them, so the plan counts the first and forgets the second.
Your partners are watching
Retailers, dealers and agents see the plan before your customers do. A partner who fears being undercut has every reason to push someone else's product before your own channel is ready.
Direct means holding the stock
A retailer that buys your goods takes on the risk of guessing demand. When you sell direct, the forecast and the leftover inventory are yours.
The ways companies did it
The decisions on record take four shapes. Each answers the same question with a different amount of distance from the old channel.
Build the shop the resellers wouldn't
Sell in your own space, on your own terms.
- Apple
- Opened its first two stores on May 19, 2001, because resellers like Sears and CompUSA misrepresented its products. Two days later BusinessWeek ran "Sorry, Steve: Here's Why Apple Stores Won't Work". By 2017 Apple's stores sold about $5,546 per square foot, the highest of any U.S. retailer.
Leave the old channel
Pull out of the stores and sell only through your own channel.
- Dell
- Pulled out of retail stores in July 1994, after inventory writedowns and related costs of $70.3 million in the first half of that year. Its gross margin went from 15.1% in fiscal 1994 to 21.2% a year later, and a net loss of $35.8 million became net income of $149.2 million.
- Nike
- Cut about half its retail partners under a plan announced in 2017, as in the story above, and in 2020 pushed further with Consumer Direct Acceleration. In fiscal 2025 its revenue fell 10%, to $46.3 billion.
Sell direct and through partners, priced apart
Keep the old channel and add your own, with different products or prices in each.
- Progressive
- Sold car insurance through independent agents and direct to drivers. In 2007 it folded its separate agent brand, Drive, back into the Progressive name but kept different pricing in each channel. By 2024 it had about 14.0 million direct auto policies and 9.8 million through agents, on $74.4B of premiums.
Make the middleman an ally
If you are the store, give brands a reason to sell through you rather than around you.
- Nordstrom
- In 2012 led a $16.4 million investment in the clothing brand Bonobos and put its clothes in 20 Nordstrom stores. Brands that had set out to avoid stores followed, among them Reformation, Everlane and Skims, which made Nordstrom its first retail partner in 2020.
What 12 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.
12 going direct decisions, 1994 to 2024
Each mark is one decision. "Never undone" means the company did not reverse it, not that it worked.
Show as a table
| Company | Year | Choice | Outcome |
|---|---|---|---|
| Dell | 1994 | Exit retail, go direct-only | Kept in part |
| Apple | 2001 | Open Apple's own stores | Never undone |
| eBay | 2002 | Acquire PayPal | Never undone |
| Progressive | 2007 | Revert to master brand | Never undone |
| Uber | 2010 | Drop 'Cab', keep going | Never undone |
| Patagonia | 2011 | Run anti-consumption ad | Never undone |
| Nordstrom | 2012 | Strategic investor model | Never undone |
| Atlassian | 2014 | Make the sales hire | Never undone |
| Nike | 2017 | Bypass wholesale, own consumer | Walked back |
| Nike | 2020 | Pivot to DTC | Walked back |
| Macy's | 2023 | Accelerate small-format expansion | Kept in part |
| Macy's | 2024 | Retrench and run marketplace | Never undone |
Of the 12 decisions that can be judged, 8 were never undone, 2 were walked back and 2 were kept only in part.
A few decisions filed here are about something nearby rather than this choice: Uber's 2010 fight with San Francisco regulators, Patagonia's 2011 ad telling customers not to buy its jacket, and Atlassian's 2014 decision to hire salespeople. Read them as neighbors of this decision.
The walk-backs
Only a few of these decisions were undone, and they are worth reading together: 2 reversed outright and 2 kept only in part.
| Company | What it did | How it was undone | Kind of move |
|---|---|---|---|
| Dell1994 | Left retail stores to sell only direct | Began selling through Wal-Mart in June 2007 and was in about 10,000 retail locations by 2008 | Right for its time |
| Nike2017 | Cut about half its retail partners to sell direct, aiming for 60% of revenue | Returned to wholesale and to Amazon; in Q2 fiscal 2025 North American wholesale grew 24% while direct fell 10%Inventory up 44% in September 2022; direct reached about 42% of revenue | Cut the channel at once |
| Nike2020 | Launched Consumer Direct Acceleration, pushing further into its own stores, app and website and out of wholesale | In March 2024 said it must "lean in with our wholesale partners"; replaced its CEO in October 2024Revenue down 10% to $46.3 billion in fiscal 2025 | Doubled down before the results |
| Macy's2023 | Planned up to 30 new small stores away from malls, tripling the format | Had 24 by January 2025, and put four into its first wave of closures | Scaled on early numbers |
The kind of move is Stratrix's reading of each case. Dell's and Macy's were kept only in part; both of Nike's were undone.
Nike and Macy's moved fast toward a number set before the results were in: 60% of revenue from direct sales, up to 30 new stores. Macy's raised its target on comparable sales from the five small stores that had been open more than a year, which grew 8%. Nike had exited about half its retail partners by September 2021, and its inventory problem arrived a year later. It had pressed on in 2020, three years into the first plan, with an acceleration of the same bet, before that plan had met its target.
Dell is the useful contrast. Leaving the stores worked: its gross margin rose from 15.1% to 21.2% in a year. It went back to retail in 2007 only after its customers changed, as falling PC prices made a custom-built machine worth less and buyers moved to ones off the shelf. Its reversal was a response to a new market, not a correction of the old decision.
Going direct can be done in steps. The moves that were undone were made all at once.
How to run this decision
If you are weighing a version of this choice, here is what the record suggests doing, in order.
- Split the middleman's cut into what it pays for
Write down what the partner's margin buys you: shelf space, shoppers, and stock held when demand drops. Put a number on each, and ask how many of the partner's sales would come to you anyway. The rest is demand you are about to start paying for yourself.
Seen in: Nike, 2017 - Go direct where you add something the channel can't
If your own channel sells the same product at the same price with a different logo on the receipt, you have taken on cost and given customers no reason to switch. Apple opened its own stores because resellers misrepresented its products, and the stores did what the resellers would not.
Seen in: Apple, 2001 - Cut partners in stages, and measure in between
Ending relationships one tranche at a time turns one large bet into a series of small tests. Nike exited about half its retail partners and learned what they had been worth only when inventory rose and revenue missed.
Seen in: Nike, 2017 and 2020 - Plan for the stock you will now hold
Selling direct moves the forecast, and the leftovers, onto your books. Dell's short move into retail stores ended after inventory writedowns and related costs of $70.3 million in half a year; the direct model had let it avoid exactly that risk.
Seen in: Dell, 1994 - Give each channel its own price under one name
Partners fear being undercut by your direct channel, so make the difference explicit instead of pretending there is none. Progressive tried a separate brand for its agents, then went back to one name and kept different pricing for each channel.
Seen in: Progressive, 2007 - Set the target after the early results, not before
A small format or new channel that looks strong in its first handful of locations has not yet met its average customer. Macy's raised its small-store target to as many as 30 on comparable sales from five stores, and then closed four of the ones it had.
Seen in: Macy's, 2023 - Keep a way back to the partners you leave
Whatever you wind down, keep the contracts, terms and people that ran it. Nike had to rebuild the relationships it ended, and Dell returned to retail thirteen years after leaving it.
Seen in: Nike, 2017 · Dell, 1994
Warning signs
What it has looked like when this decision was going wrong.
- The plan counts the partner's margin and not the partner's shoppers.Nike, 2017
- Partners are being cut broadly, all at once.Nike, 2017
- A format is promoted from experiment to growth engine on a handful of stores.Macy's, 2023
- Your agents or dealers privately call the new direct channel dead on arrival.Progressive, 2004
- The direct business needs a stock forecast nobody has had to make before.Dell, 1994
All 12 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.
Never undone · 8
- Apple, 2001Open Apple's own stores
- eBay, 2002Acquire PayPal
- Progressive, 2007Revert to master brand
- Uber, 2010Drop 'Cab', keep going
- Patagonia, 2011Run anti-consumption ad
- Nordstrom, 2012Strategic investor model
- Atlassian, 2014Make the sales hire
- Macy's, 2024Retrench and run marketplace
Kept in part · 2
- Dell, 1994Exit retail, go direct-only
- Macy's, 2023Accelerate small-format expansion
Walked back · 2
- Nike, 2017Bypass wholesale, own consumer
- Nike, 2020Pivot to DTC
Questions
- What is direct-to-consumer (DTC)?
- Selling to customers yourself, through your own stores, website or salespeople, instead of through retailers, dealers or agents. The company keeps the margin a middleman would take, and takes on the work and risk that margin used to pay for.
- What are examples of companies that went direct?
- Apple opened its own stores in 2001, Dell sold only direct from 1994, and Nike cut about half its retail partners from 2017. Apple's stores held; Dell went back to retail in 2007, and Nike returned to wholesale and to Amazon.
- Why did Nike go back to wholesale?
- Its direct business reached about 42% of revenue against a 60% target, its inventory ballooned 44% in September 2022, and its finance chief said the strategy had added "complexity and inefficiency". In March 2024 its chief executive said Nike needed to "lean in with our wholesale partners". Nike rebuilt those partnerships, replaced its chief executive in October 2024, and returned to Amazon in 2025.
- Is selling direct more profitable?
- It can be, when the direct channel adds something the partner could not and the company can carry the stock. Dell's margin rose after it left retail stores. When the partner's shoppers are not replaced, the extra margin can be eaten by the cost of finding customers and holding inventory.
More from Stratrix
“How much of the channel's margin do you actually keep?”
See the pack — $499 →All Access
Every Casebook, three Strategy Packs a seat each year and the Analyst, for teams that face these decisions all year.
$1,999 a year · $499 a seatWhat is included →Test a decision, free
Set a live decision against the companies that made it before you.
AcquisitionsThe Distance TestDivestituresThe Divestiture TestDiversificationThe Diversification TestMake or buyThe Make-or-Buy TestNew analyses, by email
The defining moves in business, checked against the record, as they publish.