The DTC playbook: retention economics and channel mix

# The DTC playbook: retention economics and channel mix

A direct-to-consumer apparel brand doing $40m has $6m of incremental spend to place next year. It can pour it into the Meta and Google auctions, sign three store leases, hire a wholesale sales team and go to market appointments, or put it into email, SMS, product and service so the customers it already has come back a third and fourth time. One budget, four doors, and the door it picks defines what the company is five years from now.

That arbitration is the job of this lesson. The figures underneath it (what a customer costs to acquire, what she is worth across a wardrobe lifetime, how often she repurchases and how much she sends back) are set up by the unit-economics and benchmark lessons in this module. Here they are assumed, and spent.

Why channel purity stopped paying

Warby Parker started selling glasses online in 2010 at $95, in a market where a pair routinely cost $300 or more. It went public in 2021 with hundreds of stores open. Two forces pushed it there, and both still apply.

The first: the acquisition auction turned against everyone standing in it. Early DTC growth ran on cheap Facebook and Instagram inventory. As more advertisers bid for the same attention, prices climbed, and Apple's App Tracking Transparency change (iOS 14.5, April 2021) removed much of the targeting and measurement signal that made those auctions efficient. The second-order effect matters more than the price increase itself. What you pay per customer on Meta, which sells the ad inventory in question, is set inside an auction, which means it is set by how much capital your competitors have just raised. A channel plan anchored to last quarter's paid-social cost per customer is a plan anchored to somebody else's funding round.

The second: online-only caps reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition →. A large share of apparel and eyewear buying still happens where people can touch, try and walk out with the thing. No app fixes the fit question for a frame sitting on a specific nose.

So the disruptors did the disrupted thing. They signed leases. Warby Parker went further and announced shop-in-shops inside Target, taking distribution from a mass retailer of the kind it once defined itself against.

What each channel is actually for

Give every channel one job and one number it is accountable for, and stop asking it to do the others.

• Owned e-commerce and CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition →: the highest margin per order and the only place the customer relationship is yours. Its job is repeat purchase, not first purchase.
• Paid acquisitionPaid acquisitionVisitors arriving via paid ads or sponsored placements, where you pay a platform to display your message rather than earning visits organically.View full definition →: buys strangers at a price you do not control. Its job is volume at a ceiling price you set in advance.
• Physical retail: fit, service and local awareness. It acquires as well as converts, which is why store openings often lift online sales in the same postcode.
• Wholesale and marketplaces: reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → into shoppers your ads will never touch, paid for in margin, data and control.

The arbitration that decides wholesale is narrower than it looks:

Retail price                        $95
Product + inbound cost              $43   (55% DTC gross margin)

DTC, paid acquisition
  Gross profit, first order         $52
  Acquisition cost                  $45
  Contribution, first order         $7    (real profit arrives on order two)

Wholesale at keystone
  Price to the retailer             $48
  Gross profit per unit             $5
  Acquisition cost                  $0
  Contribution, first order         $5

On order one the two channels are nearly indistinguishable. Everything that separates them happens afterwards: the DTC customer leaves an email address and a size, the wholesale customer leaves nothing. So wholesale wins only where the retailer moves volume you could never buy at any auction price, or where your own acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → has drifted above roughly the whole wholesale discount.

The edge case is the one that catches brands out. A company sourcing at a 55% DTC gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → has no wholesale channel available at all: keystone pricing leaves it about $5 a unit before a single markdown allowance. Before wholesale is a decision, it is a repricing and re-sourcing project, usually eighteen months long. Brands that discover this after signing their first big account end up shipping at negative contribution to protect a buyer relationship.

If the arithmetic behind those per-customer figures needs a refresher, see the Corporate Finance Institute's CAC explainer.

The owned channel's real payoff

Sell through a department store and the store knows who bought your coat. You do not.

Owned channelsOwned channelsMedia channels a company owns and controls directly, such as its website, blog, newsletter, social accounts and mobile app. No per-use payment to a publisher is required.View full definition → generate first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition →: what a customer bought, in which size, how often, what she returned, collected directly and with consent. That asset lowers future acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition →, because reactivating a known buyer through email or SMS costs a fraction of winning a stranger in an auction. Warby Parker's home try-on (five frames, shipped free) works as a data instrument as much as a conversion tool, since it reveals which shapes get kept and which come straight back.

Two failure modes here, both common. The first is a list nobody works: a brand celebrates two million email addresses while sending one campaign a month, at which point the data is a storage cost. The second is consent scope. Data collected for order fulfilment and then used for lookalike targeting across borders creates a legal exposure that surfaces years later, usually during due diligence.

And no retention program repairs a product that comes back. Where fit failure drives the return profile the benchmarks lesson describes, personalisation just re-sells the same mistake to the same person faster.

Match the loop to the product

The quickest route to more orders per customer is a reason to come back on a calendar.

Warby Parker built one with contact lenses. Contacts run out, so customers reorder without being persuaded. That replenishment loop is far stickier than frames, which people replace every year or two at most.

Sézane, the French brand that has stayed out of wholesale entirely, gets a similar effect without a subscription: a new collection lands roughly monthly, pieces go out of stock, and the audience learns to check on a rhythm. Scarcity supplies the calendar that consumability supplies elsewhere.

What does not work is bolting a subscription onto a considered purchase. Nobody wants a statement coat every quarter, and forcing the loop produces cancellations plus a support queue. A durable single product with no natural repeat occasion is the hardest retention problem in the sector: the better the product lasts, the longer the gap before the next order.

The wholesale re-entry decision

"Going wholesale" was treated for a decade as an admission of defeat. Wrong frame. Wholesale is a channel with a job, and the question is whether the job is worth its price.

What you hand over goes beyond margin:

• Price control. A retailer that marks down 40% in week six resets what your own customers think the product is worth, and your full-price site is now competing with your own goods.
• Chargebacks and markdown allowances, which arrive after the season and turn a modelled margin into a smaller real one.
• Assortment drift. Buyers order what their customer wants, and a line built to please three big accounts slowly stops looking like your brand.
• Reversibility. A wholesale account takes a season to open and years to unwind without burning the buyer relationship across every door they later work for.

> Does the margin I give up to a wholesaler cost less than the acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → I would otherwise pay for that same customer, and can I hold price once the product is on someone else's shelf?

If the answer to both is yes, wholesale is accretive. If the channel mostly sells to people who would have bought from your site anyway, it is a discount on your existing revenue dressed as growth. Test three things before signing: whether the account reaches buyers your paid channelspaid channelsVisitors arriving via paid ads or sponsored placements, where you pay a platform to display your message rather than earning visits organically.View full definition → cannot, whether you can defend price, and whether you can fill the order without starving your own site of the sizes that actually sell.

Building the channel mix

Channel is a portfolio, not an identity, and the leadership work is mostly organisational.

Sort decisions by how reversible they are. A paid budget can be cut in a week. A wholesale commitment runs a season or two. A store lease runs five to ten years. Commit in that order, and never fund the least reversible bet with the most volatile channel: brands that opened stores on the strength of cheap paid social discovered in 2021 that the auction had moved while the rent had not.

Then settle who owns what before the targets are set. If e-commerce, retail and wholesale each carry an independent revenue number, they will fight over the same customer, the same inventory and the same attributionattributionA framework for assigning credit to the touchpoints that contributed to a conversion, so you can measure which channels and interactions actually drive results.View full definition → credit, and the store team will resent the online team for booking a sale that a fitting appointment produced. One arbitrator ends most of it: contribution margin after acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition →, per channel, with inventory allocated centrally.

Purity is still a defensible choice, with a stated price. Sézane's refusal to wholesale keeps margin, data and price discipline intact, and pays for it in slower geographic reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → and total dependence on the heat of its own audience. That trade is fine as long as it is chosen rather than inherited from a founding story.

Key Takeaways

• The channel decision is an allocation decision. Rank options by reversibility, and do not put lease-length commitments behind auction-length revenue.
• Wholesale is arithmetic plus control. If the margin you surrender costs less than acquiring the same customer yourself, and you can hold price, it is an acquisition channel rather than a retreat.
• A 55% DTC gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → leaves roughly $5 a unit at keystone wholesale. Repricing and re-sourcing come before the channel exists.
• Owned channelsOwned channelsMedia channels a company owns and controls directly, such as its website, blog, newsletter, social accounts and mobile app. No per-use payment to a publisher is required.View full definition → earn their keep through repeat purchase and first-party datafirst-party dataData collected directly from your own customers and prospects through your own channels: your most reliable and privacy-compliant source.View full definition →, not margin alone. An unworked email list is a cost, and personalisation cannot rescue a product that gets returned.
• Match the repeat loop to the product: consumables carry subscriptions, considered pieces need a collection calendar instead.
• Give each channel one job and one number, allocate inventory centrally, and judge everything on contribution after acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition →.