Chapter 10: Channels, Retail and the Customer Journey

On a Saturday morning at a bike shop in Portland, a couple takes turns riding a Lumen Cargo around the parking lot with their four-year-old strapped into the child seat. The shop's mechanic adjusts the handlebars, explains how the battery charges and promises a free tune-up after the first month. They buy the bike that afternoon. The same weekend, a family near Columbus, Ohio, reads about the Cargo online, likes what they see and gets stuck on a simple question: where could they try one, and who would fix it if something went wrong? The nearest Lumen dealer is more than a thousand miles away.

Both families want the same bike at the same price. What separates them is the channel: the set of businesses and touchpoints that gets the product from Lumen to the customer and supports it afterward. Lumen's plan to take the Cargo national is, in large part, a channel problem. The product and the price were the subjects of the last two chapters. This chapter covers the third P, place: how companies choose and manage the intermediaries that reach customers, how products physically move through supply chains, how retailing works in stores and online, and how customers move among all of these on a single journey. AI is changing each of these, from forecasting how many bikes to build to shopping agents that may one day choose a bike for a customer.

10.1 Marketing channels

Few producers sell directly to every final user. Between the factory and the customer there are usually other businesses (wholesalers, distributors, retailers, dealers, online marketplaces) that each do part of the work.

A company's channel decisions affect every other marketing decision. Prices depend on the margins intermediaries need, as you saw in Chapter 9. Pricing. The sales force and advertising depend on how much selling the intermediaries do. And channel decisions are long-term commitments. A company can change its prices or its ads next month. Contracts and relationships with dealers, and the warehouses and systems that serve them, take years to build and are hard to unwind.

10.1.1 How channel members add value

Why would a producer give part of its revenue to someone else? Because intermediaries usually do some jobs more efficiently than the producer could.

The simplest reason is contact efficiency. Suppose three producers each sell to three customers directly. That takes 3 × 3 = 9 separate contacts, each with its own shipment, invoice and conversation. If all three producers sell through one distributor, the distributor makes 3 contacts with producers and 3 with customers, a total of 6. The savings grow quickly as the numbers grow: 20 producers and 1,000 customers need 20,000 direct contacts, or 1,020 through a single intermediary. That's why a grocery store stocks products from thousands of producers instead of each food company running its own shops.

Intermediaries also solve a mismatch. Producers make large quantities of a narrow range of products. Customers want small quantities of a wide range. Intermediaries buy in bulk, break the bulk into smaller lots and assemble assortments that match what customers want. A bike shop that sells several brands of bikes, helmets, locks, tires and repairs gives a customer one place to get everything needed to ride.

The table below lists the functions channel members perform. Some help complete transactions and others help fulfill them.

Channel functions

Channel function	What it involves	Example in an e-bike channel
Information	Gathering and sharing information about customers, competitors and the market	A dealer tells Lumen which models customers ask about and what they complain about
Promotion	Developing and spreading persuasive communications	A shop features Lumen in its window, social posts and group rides
Contact	Finding and communicating with prospective buyers	Customers walk into the shop; a marketplace lists the bike
Matching	Shaping the offer to the buyer's needs, including assembly and fitting	The mechanic fits the bike and recommends a child seat
Negotiation	Agreeing on price and terms so ownership can transfer	Trade-in value, financing, accessory packages
Physical distribution	Transporting and storing goods	Holding bikes in stock so a customer can ride one home
Financing	Acquiring and using funds to cover the costs of the channel	The dealer pays for its inventory before customers buy
Risk taking	Assuming the risks of carrying out channel work	The dealer is stuck with unsold bikes if demand falls

One principle follows from this list: you can eliminate intermediaries, but you can't eliminate their functions. If Lumen sells directly instead of through a dealer, someone still has to provide test rides, assemble the bike, hold inventory, fix flat tires and absorb the risk of unsold stock. Either Lumen does those jobs itself, or a different partner does them, or customers do without them. The question in channel design is always who can perform each function best for the value it creates.

10.1.2 Channel structures

The number of intermediary levels between producer and final buyer describes a channel's length.

Channel levels in consumer markets

Channel	Structure	Example
Direct	Producer → consumer	Lumen's website; Warby Parker's own stores and site
One level	Producer → retailer → consumer	Lumen → independent bike shop → rider
Two levels	Producer → wholesaler → retailer → consumer	A snack maker → food distributor → convenience store → shopper
Three levels	Producer → wholesaler → jobber → retailer → consumer	Some packaged foods and meat products

Business markets have similar structures. Many B2B sellers use direct sales forces for large customers and distributors for small ones. Lumen's fleet program, which sells and leases bikes to universities, corporate campuses and delivery businesses, is a direct B2B channel. The way those buyers make decisions is covered in Chapter 5. Business Markets and Buyer Behavior.

10.1.3 Vertical, horizontal and multichannel systems

In a traditional channel, each member is an independent business trying to maximize its own profit, sometimes at the expense of the channel as a whole. Many channels are now organized more tightly.

There are three types of VMS:

• Corporate VMS. One company owns successive stages of production and distribution. Inditex, the owner of Zara, designs, makes much of its clothing and sells through its own stores, which lets it move new designs into stores quickly.
• Contractual VMS. Independent firms join through contracts. Franchising is the most common form: McDonald's restaurants are mostly owned by franchisees who operate under a contract with the company. Retailer cooperatives such as Ace Hardware are another form.
• Administered VMS. Leadership comes not from ownership or contracts but from the size and power of one member. Large retailers such as Walmart can shape the prices, packaging and delivery practices of their suppliers, and large brands can win shelf space and cooperation from resellers.

In a horizontal marketing system, companies at the same level of the channel join forces. Starbucks cafés operating inside Target stores, under a licensing arrangement, are a familiar example: each company gets access to the other's customers.

Lumen is already a multichannel company: its website (about 55% of sales), its independent dealers (about 35%) and its fleet program (about 10%). Multichannel distribution lets a company reach more segments and match each with the channel that suits it. Its cost is complexity and, often, conflict.

10.1.4 Channel conflict

Horizontal conflict occurs among firms at the same level, such as two Lumen dealers in the same city competing for the same customers, or one dealer complaining that another discounts accessories. Vertical conflict occurs between levels, and it's usually more serious. When a producer starts selling directly to customers its dealers already serve, the dealers can feel they're competing with their own supplier.

Real examples are easy to find. Nike moved sharply toward direct sales through its own stores, website and apps around 2020 and cut back on many wholesale partners. In the years that followed, it worked to rebuild some of those retail relationships, a sign of how hard it is to replace what retailers do. Tesla sells cars directly to consumers, and in several U.S. states it has fought laws, backed by franchised auto dealers, that require new cars to be sold through dealers. In each case the conflict is about who performs channel functions and who earns the margin.

Some conflict is healthy. It can push channel members to improve. Too much can destroy the channel. Companies manage conflict by:

• Defining roles clearly. Who handles which customers, territories and services?
• Sharing the rewards of new channels. For example, giving a dealer credit for an online sale in its area.
• Keeping prices consistent. A minimum advertised price (MAP) policy, which sets the lowest price at which resellers may advertise a product, keeps online and store prices from undercutting each other.
• Communicating. Channel partners who learn about a change in advance, and have a say in it, react very differently from those who learn about it from customers.

Online selling has caused disintermediation in many industries: travel agencies lost business to airline websites, and bookstores to online retailers. New intermediaries often arise as old ones fade. Online marketplaces, comparison sites and, as you'll see later in this chapter, AI shopping agents are all intermediaries.

10.1.5 Channel design decisions

Designing a channel starts with customers, not with the company's existing relationships.

1. Analyze customer needs. What service do target customers want from the channel? How close a location, how quick a delivery, how wide an assortment, how much help before and after the sale? A family spending $3,000 on its first cargo bike wants a test ride and nearby service. A student buying a $1,399 Fold might be happy to order online and assemble it.
2. Set channel objectives. State the level of service the company will provide to each segment and at what cost. Delivering every service every customer wants is usually too expensive.
3. Identify the major alternatives. These include the types of intermediaries (company stores, independent retailers, marketplaces, a direct sales force), their number and the responsibilities of each.
4. Evaluate the alternatives. Compare them on economic criteria (sales and costs), control (how much say the company keeps over how the product is sold) and adaptability (how easily the company can change course later).

The number of intermediaries at each level is a strategic choice in its own right.

Intensive, selective and exclusive distribution

Strategy	Typical products	Benefits	Costs
Intensive	Soft drinks, snacks, toothpaste	Maximum availability for convenience goods	Little control over how the product is sold or displayed
Selective	Appliances, electronics, mid-priced bikes	Good coverage with more control and stronger relationships	Some customers must travel or search
Exclusive	Luxury cars, high-end fashion	Strong dealer commitment and service; supports a premium image	Limited reach; dependent on a few partners

Lumen uses selective distribution: about 60 dealers chosen in the Mountain West and Pacific Northwest. Moving to intensive distribution, such as selling through big-box stores, would add reach but would put a $2,999 bike in places without trained mechanics, and would strain Lumen's relationships with the dealers it already has.

10.1.6 Managing channel members

Once a channel is designed, the company has to select, motivate and evaluate the members. Producers look for intermediaries with a good reputation, the right customers, financial strength and the ability to provide the services the product needs. They motivate them with margins, training, co-operative advertising, sales leads and, increasingly, shared data. Many manufacturers now treat key resellers as partners, practicing partner relationship management: planning jointly, sharing forecasts and solving problems together rather than simply selling to them. Finally, the company evaluates members against standards such as sales, inventory levels, customer satisfaction and service quality, and rewards those who perform well.

10.2 Supply chains and logistics

A marketing channel is the downstream part of a larger system. Before Lumen can sell a bike, it needs frames, motors, batteries, displays and tires from suppliers, often in several countries.

Logistics used to be treated as a cost center: get the goods out the door cheaply. Now most companies see it as part of the customer experience. Fast, reliable delivery and easy returns are reasons people choose one seller over another, and a stockout of a popular product is lost revenue that may never come back.

The goal of logistics is not maximum service at any price or minimum cost at any service level. It's the right balance. Delivering every order the next day is possible, but expensive. Holding enough inventory to never run out ties up cash in products that may sit for months. Each company has to decide what level of service its customers will pay for.

10.2.1 Major logistics functions

Warehousing. Companies store goods between production and sale. Traditional warehouses store goods for a long time. Distribution centers are designed to move goods quickly: receive, sort and ship out, often within a day. Large e-commerce companies run highly automated distribution centers near major cities so they can deliver within a day or two.

Inventory management. Too little inventory means stockouts, lost sales and disappointed customers. Too much means high carrying costs, products that go out of date and eventual markdowns. Many companies use just-in-time systems, which keep small inventories and rely on frequent, reliable deliveries. Just-in-time is efficient when supply is predictable and fragile when it isn't, as the shortages of the pandemic years showed. Companies also hold safety stock, extra inventory that protects against forecast errors and supply delays.

Transportation. Each mode has strengths and weaknesses.

Transportation modes

Mode	Strengths	Weaknesses	Typical use
Truck	Flexible, door to door	Higher cost per mile than rail or water over long distances	Most domestic freight, final delivery
Rail	Low cost for heavy, bulky loads over land	Slower, less flexible	Coal, grain, cars, containers
Water	Lowest cost for very large loads	Slowest; weather and port delays	Imported goods in containers
Air	Fastest	Most expensive; weight limits	High-value or urgent items, perishables
Pipeline	Low cost, continuous	Only for liquids and gases	Oil, natural gas
Digital	Instant, nearly free per unit	Only for digital products	Software, music, video, e-books

Companies often combine modes, which is called intermodal transportation: a shipping container might travel by ship, then rail, then truck. Products with special risks bring their own rules. Lithium-ion batteries, like those in e-bikes, are classed as dangerous goods for transport and must be packed, labeled and shipped under specific regulations, which limits how they can travel by air.

Logistics information management. Channel partners share information electronically so that goods can move without delays: orders, shipping notices, inventory levels and point-of-sale data. Barcodes, RFID tags and GPS tracking let companies see where products are throughout the chain.

Many companies hand some or all of these functions to third-party logistics (3PL) providers, companies that run warehouses, ship goods, handle returns and manage inventory for their clients. A 3PL can give a smaller company like Lumen access to warehouses in new regions without building them.

10.2.2 Demand forecasting

Every logistics decision rests on a forecast. How many Cargos will customers buy next spring, in which cities, through which channels? The answer determines how many bikes Lumen orders, where it stores them and how much cash it ties up.

Common forecasting methods include:

• Time-series methods, which project past sales patterns forward, adjusting for trend and seasonality. Bike sales, for example, rise in spring and fall in winter in most of the United States.
• Causal (driver-based) methods, which link sales to factors that drive them, such as price, advertising, weather, fuel prices or new dealers opening.
• Judgmental methods, which gather estimates from salespeople, dealers, executives or outside experts. These are essential when there's no history, as with a new product or a new market.
• Market tests, which launch in a limited area and use the results to forecast for a larger one.

Forecasts are always wrong to some degree. The question is how wrong and in which direction. Long lead times make errors more expensive. Suppose, for illustration, that Lumen's contract manufacturer needs about four months from order to delivery at Lumen's Denver warehouse. If Lumen expects to sell 300 Cargos a month during the selling season, it must commit to roughly 1,200 bikes before it sees how the first month of sales goes. At a manufacturing cost of $1,480 each (the illustrative figure from Chapter 9. Pricing), that's 1,200 × $1,480 = $1,776,000 tied up in inventory based on a forecast.

Forecast errors also grow as they travel up a supply chain, a pattern called the bullwhip effect. Small changes in what consumers buy turn into larger swings in retailer orders, larger still in distributor orders and largest at the factory, because each member adds its own safety margin and orders in batches. Procter & Gamble famously found large swings in orders for Pampers diapers even though babies' use of diapers is quite steady. The bicycle industry lived through a bullwhip during and after the pandemic: demand surged, shops and brands ordered heavily to catch up, and when demand cooled, much of the industry was left with too much inventory and turned to deep discounts.

10.3 Retailing and omnichannel

The last step in most consumer channels is retailing.

Retailing is no longer just a store. Producers sell directly online, marketplaces connect millions of sellers with buyers, and social platforms let people buy without leaving an app. But physical stores remain where a large share of retail spending happens, and many purchases that end online start in a store, or the other way around.

10.3.1 Store retailers

Store retailers differ in the amount of service they offer, the breadth and depth of their product lines, their prices and how they're organized.

Major types of store retailers

Type	Description	Examples
Specialty store	Narrow product line with a deep assortment	Independent bike shops, running stores, REI
Department store	Several product lines, each run as a separate department	Macy's, Nordstrom
Supermarket	Large, low-cost, self-service store for food and household products	Kroger, Safeway
Convenience store	Small store near homes, open long hours, limited line of high-turnover items	7-Eleven
Superstore and category killer	Very large store; a category killer offers a deep assortment in one line	Walmart Supercenters; Best Buy, Home Depot
Off-price retailer	Buys brand-name goods at less than regular wholesale prices and sells for less than retail	TJ Maxx, Ross
Warehouse club	Membership store selling a limited selection in bulk at low prices	Costco, Sam's Club

Retailers also differ in service level. Self-service retailers serve customers who want to "locate, compare and select" on their own. Full-service retailers, such as specialty stores and high-end department stores, help customers at every stage of the purchase. An independent bike shop is a full-service specialty retailer: its value lies in fitting, test rides, assembly, expert advice and repairs. That makes it a natural channel for products that need explanation and service, and an expensive one to replace.

Finally, retailers differ in organization. Corporate chains own and control many outlets. Franchises operate under contracts with a franchisor. Independent stores sometimes join retailer cooperatives to buy and advertise together.

10.3.2 Online and marketplace retailing

Online retailing includes pure online retailers, direct-to-consumer (DTC) brands that sell their own products through their own sites, and store retailers' websites and apps. Many DTC brands, such as Warby Parker, began online and later opened physical stores, because stores build awareness, let customers try products and often make online sales in the same area rise.

Amazon's marketplace, eBay and Etsy are well-known examples. For a producer, a marketplace offers huge reach and fulfillment services, but at a cost: fees, competition on the same page with rivals and copycats, limited control over how the brand is presented, and less direct access to customer data. Social commerce, buying directly inside social media platforms, is a growing form of the same idea. You'll see the digital side of these channels in Chapter 12. Digital, Social and Search Marketing.

10.3.3 Retailer marketing decisions

Retailers make the same kinds of marketing decisions producers do:

• Segmentation, targeting and positioning. Who does the store serve, and why should those customers choose it?
• Product assortment and services. What range of products, and which services: fitting, delivery, installation, repairs, returns?
• Store atmosphere and experience. The layout, design and activities that make visiting worthwhile. Experiential retailing gives customers reasons to visit beyond buying. REI's large flagship stores, with spaces to try gear and classes on outdoor skills, are an example.
• Price. EDLP or high-low, as described in Chapter 9. Pricing.
• Promotion. Advertising, events, loyalty programs and personal selling.
• Location. For store retailers, perhaps the most important decision of all, because a poor location is very hard to fix.

For a bike shop, the experience might be group rides, repair clinics and a test-ride loop. Those are reasons for a family to visit that no website can match.

10.3.4 Omnichannel retailing

Customers don't think in channels. They read reviews on their phones while standing in a store, order online and pick up at the curb, or try a product in a shop and order it online that night.

Multichannel and omnichannel aren't the same. A multichannel company sells through several channels that may operate separately, with different prices, inventory and customer records. An omnichannel company connects them: the same customer record, inventory visible across channels, and the ability to start a purchase in one channel and finish it in another.

Common omnichannel practices include:

• Buy online, pick up in store (BOPIS) and curbside pickup. Target's order-pickup and drive-up services are a well-known example.
• Ship from store, using store inventory to fill online orders from the nearest location.
• Endless aisle, letting store staff order products that aren't on the shelf for delivery to the customer's home.
• Easy cross-channel returns, such as returning an online purchase to a store.
• Unified loyalty and customer service, so a customer's history follows them across channels.

Two shopping behaviors make omnichannel essential. Showrooming is examining a product in a store and then buying it online, often from a different seller. Webrooming is the reverse: researching online and then buying in a store. Retailers that connect their channels can win both kinds of shoppers instead of losing them at the switch. Best Buy, for example, responded to showrooming by matching online prices and using its stores as pickup and shipping points.

10.3.5 The integrated customer journey

In Chapter 4. Consumer Behavior you followed the buyer decision process from need recognition to post-purchase behavior. From a channel perspective, the question is which channel serves the customer at each stage, and whether the handoffs between them work.

Here is a simplified journey map for a family buying a Lumen Cargo in a city where Lumen has a dealer.

Simplified customer journey for a Lumen Cargo

Stage	Customer goal	Touchpoints and channels	Possible pain points
Awareness	Find a way to replace some car trips	Social posts, a friend's bike, search results, AI assistants	Never hearing of Lumen
Research	Compare cargo bikes and learn whether one is practical	Lumen website, reviews, website chat assistant, dealer website	Conflicting information; no local reviews
Try	Ride one with a child on board	Dealer test ride, demo event	No dealer nearby; dealer out of stock
Buy	Get the right bike and accessories at a fair price	Dealer, Lumen website, phone	Different prices in different channels; financing questions
Delivery and setup	Get a safe, adjusted bike	Dealer assembly and fitting, or home delivery	Self-assembly of a heavy bike; delivery damage
Use and service	Ride safely; fix problems fast	Lumen app, dealer service, Lumen Care, chat assistant	Long waits for repairs; unclear warranty process
Loyalty and advocacy	Share the experience, buy accessories, recommend	Reviews, social media, app, dealer events	No reason to stay in touch

Notice how many stages depend on the dealer. In a city without one, the "Try," "Delivery and setup" and "Use and service" stages have no good channel at all, which is exactly the Columbus family's problem.

10.4 AI in channels

AI is reshaping channels in three main ways: predicting demand, running the physical movement of goods, and, newest and least settled, acting as an intermediary between customers and sellers.

10.4.1 AI forecasting and inventory

Machine-learning forecasting models can draw on far more signals than traditional time-series methods: sales by store and day, prices and promotions, weather, local events, website traffic, search trends and competitors' prices. Large retailers use them to forecast demand for each product at each location and to decide automatically how much to reorder and where to send it. Better forecasts at that level mean fewer stockouts and fewer markdowns.

AI forecasting has clear limits, though:

• Regime changes. Models learn from the past. When conditions change suddenly, as in a pandemic, a new competitor or a policy change, the past may be a poor guide, and a model can be confidently wrong. The bike industry's boom and bust is a reminder that the biggest forecasting problems are often the ones no history covers.
• New products and new markets. With no sales history, a model has little to learn from. Lumen's national Cargo launch is exactly this situation, and it calls for judgmental methods, market tests and close watching of early results.
• Data quality. A model fed with incomplete dealer sales data, or with stockout periods recorded as low demand, will learn the wrong lesson. If a dealer had no Cargos in stock for a month, zero sales doesn't mean zero demand.
• Opacity. It can be hard to tell why a model forecast what it did, which makes it hard to know when to override it.

10.4.2 Automated logistics and fulfillment

AI and automation now run large parts of the physical supply chain. Amazon bought the warehouse robotics company Kiva Systems in 2012, and robots that bring shelves of products to human pickers are now common in large fulfillment centers. Routing software decides the order in which delivery drivers make their stops. UPS's ORION system, which plans driver routes, is a well-known example. AI also helps decide which warehouse should fill each order, how to pack it and when to reorder. Companies are testing autonomous trucks, sidewalk robots and drones for delivery, with results that vary by place and regulation.

For marketers, the point is that these systems change what customers expect. When some retailers deliver in a day, a week starts to feel slow. A smaller company usually can't build its own robotic warehouses, but it can buy access to the same capabilities through 3PLs and marketplaces, at the cost of some control.

10.4.3 AI shopping agents as a new intermediary

The biggest potential change is in who, or what, stands between the customer and the seller. You met AI shopping assistants as part of the buying journey in Chapter 4. Consumer Behavior. From a channel perspective, they look like a new kind of intermediary.

Go back to the list of channel functions. An AI shopping agent can perform several of them: information (summarizing products and reviews), contact (finding sellers the shopper never heard of), matching (recommending the option that fits the shopper's stated needs) and even negotiation (finding the best available price or applying a discount). It doesn't perform physical distribution, financing or risk taking, and it can't give a test ride or fix a brake. Those still need someone else.

If agents become common, several things follow for marketers:

• The agent becomes a gatekeeper. A shopper who asks an agent for "the best cargo e-bike for two kids under $3,500" may see only a few options. Being on that short list matters as much as shelf space once did.
• Product information has to be machine-readable and accurate. Agents rely on product data, reviews, specifications and policies they can find and parse. Vague marketing claims carry less weight than clear facts: weight capacity, battery certification, range, warranty terms, service locations.
• Brands may lose the direct relationship. If an agent handles the purchase, the brand may never see the customer's questions or build a connection with them, much as producers selling through marketplaces lose some customer data.
• Agents have their own incentives and errors. An agent may be run by a platform that earns fees from some sellers, may favor sellers with the cleanest data rather than the best products, or may simply get facts wrong. Who is responsible when an agent buys the wrong item, and who handles the return, are open questions.

How fast agents will spread, and how much buying they will actually do, is uncertain. Many shoppers will still want to see, touch and try products, especially expensive ones. The safest course for a marketer is to treat agents as one more channel to understand and prepare for, not as a replacement for the others. You'll look at how AI-generated answers are changing search, a closely related shift, in Chapter 12. Digital, Social and Search Marketing.

10.4.4 What stays a human judgment

AI can forecast, route, pick and even shop. It can't decide which partners a company should trust, how to share the rewards of a new channel fairly with dealers who built the business, or what level of service the brand should promise. Channels are relationships as much as systems, and the hardest channel decisions, like whether to add a channel that competes with existing partners, are judgments about strategy and fairness that a person has to own.

10.5 Lumen Cycles: a channel strategy for going national

Dana Okafor's team lays out the channel problem plainly. Today Lumen reaches customers through its website, about 60 dealers in the Mountain West and Pacific Northwest, and its fleet program. The national Cargo launch will bring Lumen to families in metro areas where it has no dealers. Those families are exactly the customers who most want a test ride and local service.

The economics. In Chapter 9. Pricing you saw that Lumen keeps about $1,350 of contribution on a Cargo sold through its website but only about $380 on a Cargo sold through a dealer, because the dealer keeps a 35% margin. If the national launch's $1,350,000 of added fixed costs were covered entirely by online sales, Lumen would need to sell 1,000 Cargos. If half the sales went through dealers, the weighted average contribution would be (0.5 × $1,350) + (0.5 × $380) = $675 + $190 = $865 per bike, and the break-even volume would rise to $1,350,000 ÷ $865 = 1,560.7, so 1,561 Cargos.

That comparison makes direct sales look far better, but it leaves out the channel functions. Through a dealer, the dealer pays for inventory, provides the test ride, assembles and fits the bike, handles warranty repairs and promotes Lumen locally. Selling direct, Lumen must pay for those jobs itself or do without them, and much of its online advertising, which sits in fixed costs, exists to do the "contact" work a dealer's storefront does for free.

The options. The team sketches three channel strategies for new metros.

Channel options for the national Cargo launch

Option	How it works	Strengths	Risks
Direct only	Website sales with home delivery; a mobile-repair partner for service	Highest contribution per bike; full control of price and data	No test rides; families hesitate; service quality depends on a new partner
Recruit new dealers	Selective distribution: sign a few strong shops in each new metro	Test rides and service from day one; dealers share inventory risk	Slow to build; lower contribution per bike; dealers may hesitate to stock an unknown brand
Hybrid ("buy anywhere, ride with a dealer")	Customers buy online or in store; online orders in a dealer's area are assembled, delivered and serviced by that dealer for a fee	Combines online convenience with local service; reduces channel conflict	Requires shared inventory and order systems; fee levels must be fair to both sides

Marcus models the hybrid option with an illustrative fee of $250 paid to the dealer for each online Cargo order it assembles and delivers. Lumen would then keep about $1,350 − $250 = $1,100 on those bikes (ignoring the small difference between shipping to a dealer and shipping to a home), far more than the $380 on a dealer sale, while the dealer gets paid for work it does and a new customer for service and accessories. Whether $250 is enough to motivate dealers, and too much for Lumen, is something the team would test with a few dealers first.

Managing conflict. Lumen's policy that its website and dealers advertise the same price for each bike, discussed in Chapter 9. Pricing, already removes the most common source of conflict. Under a hybrid model, Lumen would also need clear territories for crediting online orders and a commitment to share customer data with the dealer that services the bike, with the customer's consent.

Logistics. Lumen's bikes are built by a contract manufacturer and shipped by ocean freight to its Denver warehouse, a lead time of about four months. Serving families on the East Coast from Denver means slow and costly shipping of large, heavy bikes with batteries that must be shipped as dangerous goods. The team is considering a 3PL warehouse in the eastern United States rather than building its own.

AI. Marcus wants to use AI forecasting for the existing markets, where Lumen has years of history, and judgmental forecasts plus close tracking of early sales in the new metros, where it has none. He's careful to flag dealer stockouts so the model doesn't mistake missing inventory for missing demand. Jess Patel adds an item no one expected: making sure Lumen's product pages state the Cargo's specifications, weight capacity, battery certification, warranty and service locations clearly and consistently, so AI assistants and shopping agents describe the bike accurately.

Your marketing plan. Draft the Channel strategy section of your plan, in about one page. If you're using Lumen, focus on the national Cargo launch.

1. Describe your target customers' channel needs at each stage of their journey, using a short journey map.
2. Choose a channel structure: direct, indirect or multichannel, and the intensity of distribution (intensive, selective or exclusive). Explain which channel functions each member will perform.
3. Show the channel economics: contribution per unit in each channel and how the channel mix affects your break-even volume from the pricing section.
4. Explain how you will manage channel conflict, including pricing consistency and how partners are rewarded.
5. Summarize key logistics decisions: where inventory will sit, how products will reach customers and how you'll forecast demand.
6. Note how AI fits: forecasting, fulfillment, and how you'll prepare for AI shopping assistants and agents.

Disclose any AI help, and check every figure, as the rules from Chapter 1. Marketing in an AI Era require.

10.6 Try it with AI

AI assistants can produce a tidy customer journey map in seconds. The map will look plausible. Whether it matches what a real customer experiences is another matter: AI may describe services a retailer doesn't offer, miss ones it does, or describe an older version of the website. The only way to know is to walk the journey yourself.

Choose a retailer you can actually visit, online and, ideally, in person: a sporting goods chain, an electronics retailer, a home improvement store, a grocery chain with pickup, or a local shop with a website. Choose a specific product a customer might buy there, such as a pair of running shoes, a laptop or a cordless drill. Then start a new conversation and type the prompt below, filling in your choices.

Now test the map by walking through the journey yourself. Don't buy anything and don't create an account you don't want.

1. Verify each touchpoint. Visit the retailer's website and app, and the store if you can. For each touchpoint the AI listed, mark it confirmed, not found or different from described. Add touchpoints the AI missed.
2. Test the handoffs. Try at least two cross-channel moves: check whether the website shows in-store stock, look up the pickup and return policy, see whether an item in your app cart appears on the website, or ask the store's chat assistant a question about pickup. Record what happened.
3. Check the AI's uncertainty list. Were the details it flagged as uncertain actually wrong more often than the others? Did it flag anything that turned out to be wrong?
4. Apply the chapter. Is this retailer omnichannel or only multichannel? Which practices from the chapter does it use (BOPIS, ship from store, endless aisle, cross-channel returns)? Where does the journey break?

Then ask the AI to apply the same thinking to Lumen:

Check the reply against the channel functions table in this chapter. Did the AI use each function correctly, or did it, for example, treat "promotion" and "information" as the same thing? Did it notice that removing the dealer shifts work to Lumen or to the customer rather than eliminating it? Compare its two suggestions with the three options in the Lumen section, and decide which you'd recommend. Write a short paragraph with your recommendation and a one-line note on how you used AI.

10.7 Summary

Marketing channels are sets of interdependent organizations that make products available to customers. Intermediaries add value through contact efficiency and by matching producers' large, narrow output to customers' small, varied needs, and they perform functions such as information, promotion, contact, matching, negotiation, physical distribution, financing and risk taking. Intermediaries can be removed, but their functions can't. Channels may be direct or indirect, organized as vertical, horizontal or multichannel systems, and designed for intensive, selective or exclusive distribution. Conflict between channel members is common, especially when a producer adds a direct channel, and is managed with clear roles, shared rewards, consistent pricing and communication.

Behind the channel sits the supply chain. Logistics balances customer service against cost through warehousing, inventory management, transportation and information management, all guided by demand forecasts whose errors grow with long lead times and the bullwhip effect. Retailers range from specialty stores to warehouse clubs, online retailers and marketplaces, and the most effective now practice omnichannel retailing, connecting channels so customers can move among them on a single journey. AI is improving forecasting and automating fulfillment, though it struggles with new markets and sudden change, and AI shopping agents may become a powerful new intermediary that rewards accurate, machine-readable product information. Choosing and managing channel partners remains a human judgment.

10.8 Review Questions

1. Explain contact efficiency. How many contacts are needed for 5 producers to reach 40 customers directly, and how many through one intermediary?
2. List the eight channel functions in this chapter. For an online-only mattress brand, who performs each function?
3. Distinguish corporate, contractual and administered vertical marketing systems, and give an example of each that isn't in the chapter.
4. A premium headphone brand is deciding between selective and exclusive distribution. What factors should it weigh?
5. What is the bullwhip effect, and how can sharing sales data across a channel reduce it?
6. Explain the difference between multichannel and omnichannel retailing. Describe a retailer you've used that does one well and one that does it poorly.
7. Which channel functions can an AI shopping agent perform, and which can't it? What should a brand do to prepare for shoppers who use agents?
8. Lumen earns far more per bike on direct sales than on dealer sales. Make the case for keeping and expanding the dealer network anyway.