For a while, direct-to-consumer became more than a route to market. It became a belief system.
Owning the customer relationship was presented as inherently superior to wholesale. Marketplaces were treated as necessary evils. Stores were sometimes described as legacy infrastructure waiting for digital economics to make them irrelevant. If enough customers could be persuaded to buy directly, the theory went, brands would gain better data, stronger margins and much more control over the relationship.
Some of that was true. The problem was the word inherently.
A direct channel can be strategically valuable and economically unattractive at the same time. A marketplace can surrender part of the economics and still create customers the brand would never have reached itself. A store can appear less efficient in a channel P&L while quietly influencing a digital transaction that gets credited somewhere else.
The customer, meanwhile, has been largely indifferent to the argument.
They move between stores, websites, marketplaces, social platforms and increasingly AI-assisted discovery according to whatever is easiest or most useful at that moment. They do not wake up thinking about channel ownership. They think about the product, the price, whether they trust the seller and how quickly they can get what they want.
That is why I do not think DTC is dead.
I think the idea that one channel should defeat the others is.
DTC exposed a problem worth solving
The original DTC argument was compelling because it identified something genuinely broken.
Brands had become too distant from their customers.
Wholesale could provide distribution at enormous scale, but it often left the brand with limited visibility into who bought, why they bought and what happened afterwards. Retail partners owned much of the transaction data. Feedback travelled slowly. The brand could spend heavily creating demand while knowing surprisingly little about the people responding to it.
Direct commerce changed that.
A customer buying through an owned channel could be recognised, served again, understood over time and connected to a broader relationship. Product teams could see behavioural signals. Marketing could measure more of the journey. Service could have richer context. Loyalty could become more than a generic rewards mechanic.
Those capabilities still matter enormously.
Where the argument went too far was assuming that because customer closeness was valuable, the direct transaction itself must always be the most valuable transaction.
The economics became harder as acquisition costs rose, digital competition intensified and fulfilment expectations increased. Building traffic, conversion, payments, service, fraud capability, logistics and returns is expensive. For some categories, the theoretical margin gained by removing an intermediary was quietly consumed by the cost of becoming one.
The lesson was never that direct was wrong.
It was that ownership of the relationship and ownership of every transaction are not the same strategic objective.
That distinction matters because it allows a brand to pursue customer intimacy without forcing every purchase through one channel.
Channel P&Ls encourage strange behaviour
Many of the worst omnichannel decisions I have seen were perfectly rational inside the way the organisation measured itself.
The eCommerce team wants the transaction online because online revenue sits in its P&L. The store team wants the sale through the store because its targets depend on store performance. Wholesale wants to protect partner economics. Marketplace teams pursue GMV. Marketing may be measured against attributed revenue regardless of where the customer eventually buys.
Each function behaves sensibly within its own incentives.
The customer experiences the combined result.
This creates familiar absurdities. A store associate hesitates to help with an online order because the revenue will not count toward the store. Digital marketing gets credit for a sale even though the customer visited a store twice before purchasing. Inventory is technically available somewhere in the network but commercially inaccessible because moving it would hurt another channel’s economics.
Technology can connect those systems. It cannot resolve the incentives underneath them.
That is why omnichannel programmes often reach a frustrating plateau. The organisation integrates inventory, customer data and fulfilment, yet the behaviour remains stubbornly channel-led because the commercial architecture has not moved with the technology.
The more useful unit of strategy is the customer relationship.
How much did it cost to acquire the customer? How often do they buy? What margin do they generate over time? Which interactions increase their likelihood of returning? Which channels make the total relationship more valuable?
Once the economics are viewed that way, the question changes from “Which channel won the transaction?” to “Did the enterprise create more customer value?”
That sounds like a subtle shift.
It is not.
It changes how budgets, incentives and decisions need to work.
Stores and digital were never opposites
The store-versus-eCommerce debate has always bothered me because it confuses an interface with a business model.
A good physical store can be a discovery environment, fulfilment point, service centre, acquisition channel, brand theatre and source of customer data. A good digital business can extend all of those capabilities before, during and after the visit.
The strongest retailers increasingly make the distinction difficult for the customer to see.
A person might discover a product on social media, check availability online, visit a store to try it, ask an associate for advice, order a different size from the associate’s device and have it delivered home. Which channel made the sale?
Operationally, somebody will need an answer because accounting still exists.
Strategically, obsessing over that answer can be counterproductive.
The store influenced the transaction. Digital provided inventory visibility. The associate added confidence. The fulfilment network completed the experience. The customer should not care which department receives the internal applause.
This is particularly important as stores become more digitally capable.
Clienteling, mobile point of sale, endless aisle, appointment management, personalised recommendations and access to cross-network inventory turn the store into part of the broader commerce operating system. An associate who can see the customer relationship and sell from the entire inventory network is no longer constrained by what happens to be hanging three metres away.
At that point, describing the business as “store plus eCommerce” starts to feel increasingly artificial.
It is one commerce business with multiple surfaces.
The operating model should eventually reflect that.
Marketplaces are neither salvation nor surrender
Marketplaces created another version of the same ideological debate.
For some brands, they provide extraordinary reach and customer acquisition. They concentrate demand, simplify discovery and offer logistics or payment capabilities that would be expensive to reproduce independently. In categories where customers naturally shop comparatively, refusing to participate can amount to refusing to appear where the category is being bought.
The trade-offs are real.
Margin may be lower. Customer data can be constrained. Brand presentation may be weaker. The marketplace can become powerful enough to influence pricing, visibility and eventually the economics of the category itself.
That does not make the strategic decision binary.
The better question is what role the marketplace should play inside the portfolio.
It may be an acquisition channel for customers who later deepen their relationship directly. It may make sense for certain products, markets or customer segments rather than the entire assortment. It may provide reach in a geography where building a full owned operation would not yet be justified.
The mistake is allowing convenience to become dependency without noticing.
A brand that cannot create demand outside a marketplace is vulnerable. A brand that refuses marketplaces despite customers clearly preferring them may be equally misguided.
The strategic objective is optionality.
Strong brands should be capable of using distribution partners without becoming invisible behind them.
That requires maintaining an owned centre of gravity: first-party customer relationships, differentiated product, distinctive brand experience, useful loyalty and enough direct capability to understand demand rather than merely receive orders from somebody else’s platform.
Direct still matters enormously in that model.
It simply stops needing to win every transaction.
AI will make channel ownership even less meaningful
The next disruption to this debate may come from AI.
Product discovery is already moving into interfaces that sit above retailer websites, search engines and marketplaces. A customer may describe a need to an assistant, receive a shortlist and eventually allow that assistant to compare prices, check availability or transact.
When that becomes common, the traditional idea of a customer “entering a channel” becomes even weaker.
The customer may never consciously visit the brand’s website. They may not browse a marketplace in the conventional sense either. An agent could assemble information from multiple sources and decide which merchant best satisfies the customer’s constraints.
That makes the underlying commerce capabilities more important than the surface.
Accurate product information matters. Inventory visibility matters. Competitive pricing matters. Trust matters. Fulfilment reliability matters. Identity and loyalty matter if the agent can legitimately use them.
The retailer can no longer assume that controlling the digital storefront means controlling discovery.
This is another reason the old channel war becomes strategically unhelpful.
If an AI interface begins mediating the journey, arguing internally about whether digital or marketplace “owns” the customer looks even more detached from reality. The enterprise needs to make itself easy to discover, easy to trust and easy to transact with regardless of which interface initiated the interaction.
The competitive advantage moves further into the operating system underneath commerce.
The future is a portfolio, not a winner
There is still a strong case for owned commerce.
A brand needs somewhere it can express itself without compromise. It needs direct customer relationships, first-party data and the ability to learn. It needs enough control over the experience to experiment and enough economic visibility to understand which customers are actually valuable.
That owned centre becomes more important as the ecosystem becomes more fragmented, not less.
But an owned centre is different from an owned-only strategy.
The customer relationship can begin in a store, a marketplace, social media, an AI assistant or a wholesale partner. The commercial skill lies in deciding how those channels reinforce one another, what economics each should carry and how the enterprise gradually builds a stronger relationship with the customer wherever that relationship began.
This requires more discipline than declaring one channel strategically superior.
It means accepting that the highest-margin transaction may not create the highest-value customer. A marketplace sale can be commercially sensible if it acquires a relationship profitably. A store visit can create digital revenue weeks later. Direct fulfilment can improve a store experience. Wholesale can build awareness that owned channels subsequently monetise.
The organisation needs to understand those interactions well enough to invest accordingly.
That is ultimately what the DTC era taught us.
Brands were right to want greater proximity to customers. They were right to build their own capabilities, collect better data and reduce their dependence on intermediaries.
The error was turning proximity into exclusivity.
Customers do not reward companies for winning internal channel arguments. They reward businesses that are present where they want to shop and still manage to recognise them when they move somewhere else.
The strongest commerce organisations will therefore own what matters most: the customer understanding, proposition, data, experience and operating capabilities that make the relationship valuable.
They can rent distribution when it makes sense.
They can partner when somebody else provides reach or capability more efficiently.
They can use stores, marketplaces and owned digital without pretending one must eventually eliminate the others.
DTC is not dead.
It has simply grown up.