A numbered framework: how the market is segmented; what essential work the target segment requires; which portfolio of participants should do that work; and whether to go direct or indirect.

Direct versus indirect comes last. It follows from identifying the work, such as explaining, trialing, installing and recommending, and deciding who can do each part well. Starting with 'should we sell direct?' skips the analysis that answers it.

A flow diagram for the existing AV 609 mixer and AF pump: Aqualisa sells through its own sales force, distributors and trade shops to plumbers and showrooms, which reach consumers.

Plumbers sit between the manufacturer and most consumers. The installer, not the homeowner, picks the product in most replacement jobs, so the product has to win the plumber before it can win the household.

The channel diagram annotated with each participant's work: the sales force gives plumbers information; trade shops handle logistics, assortment and aggregation; plumbers install and recommend; showrooms provide trial and evaluation.

Each link does different work. Plumbers make the recommendation, and showrooms are the only place a consumer can see and try the product before a plumber is involved, which is why showrooms became the turnaround route.

The diagnosis: too expensive; plumbers are risk-averse about electronic showers after past failures; installation is easier, so it is unclear who benefits; the sales force would not compete with its bread-and-butter products and saw sales to existing accounts as zero-sum; and marketing was poorly targeted and did not use showrooms.

Only one reason is about the product. The others concern participants' incentives and risk, which is the lesson's main point. 'Who benefits from easier installation?' matters most: a plumber billing by the day may lose income when a two-day job becomes a one-day job.

A bar diagram. Start from the old product's price plus lifecycle costs. A new product's EVC equals the old total less its reduced lifecycle costs, plus any incremental value. The new price is then set between the supplier's cost and EVC, which splits the gap into the supplier's margin and an inducement left for the customer.

EVC is a ceiling, not a price. The inducement is the share deliberately left with the customer to make switching worthwhile, and when intermediaries are involved, part of the value also has to reach them.

The existing system costs about EUR 2,420 installed: EUR 1,160 of materials (AV 609 mixer EUR 715 plus pump EUR 445, plumber markup under 5%), EUR 960 of labor (2 days at EUR 60/hr) and about EUR 300 of demolition (EUR 200-500). The Quartz installs in one day (EUR 480) without demolition, putting its EVC at at least EUR 2,000, plus better aesthetics, functionality and ease of use.

Most of the value comes from labor and demolition the homeowner avoids, which is exactly the plumber's income. EUR 2,420 - EUR 480 = EUR 1,940 is the price at which a homeowner merely breaks even before the qualitative benefits. The value is real, but it flows away from the participant who chooses the product.

The Quartz price chain: retail EUR 1,080, trade margin EUR 505, manufacturer's selling price EUR 575, manufacturer margin EUR 345, cost EUR 230. Aqualisa nets EUR 345 per Quartz against EUR 330 from an AV 609 plus pump.

Trade shops keep EUR 505 of the EUR 1,080 retail price (47%) but do little to promote the product, while Aqualisa earns only EUR 15 more than on the old product. The innovation barely changes the manufacturer's take, so the sales force has little reason to push it.

Aqualisa pursued none of the case options and waited. The market began turning in January 2002, and by May the Quartz sold 500 units a week. The turnaround began in showrooms, where interior designers and wealthy consumers specified the Quartz to their plumbers, who then discovered its advantages.

Adoption came through the participant who gained most from the product's appearance and ease of use, the specifying consumer, who then pulled plumbers along. The bottleneck was broken by changing who made the choice, not by persuading plumbers directly.

The marketer's mission is a mutually beneficial exchange between company and customers. That requires considering the whole value chain, including intermediaries. If any stage lacks an exchange of value, the route fails, and good R&D alone is not sufficient.

Use this as a test for each link in a route map: what does this participant gain, and what do they risk? Any link without a credible answer is the likely bottleneck.

Aqualisa's existing route did not suit an innovative product: trade shops took a high markup without pushing it, and plumbers had to price jobs differently to capture value. Going direct risked conflict with trade shops that still carried the bread-and-butter products, and the sales organization was unmotivated to sell the Quartz.

A route built for an incumbent product can actively work against a new one. The companion caution (slide 31) applies to a startup that depends on an incumbent's channel as well: the channel owner protects its existing business first.