Distribution strategy: choose and connect your channels.
Evidence status
What was reproduced, and what it does not prove.
The public CSV and the Python run verify the weights, the direct and indirect scores, and the positive-contribution gate of the illustrative grid. They verify neither the coverage, nor the control, nor the real economics of the channels.
- Technical specification of the channel choice grid, specification in French (opens in a new tab)
v1, 31 August 2026 · CC BY 4.0 · specification in French · Method document written by marketing-mix.net. The figures are illustrative: they are not market measurements.
Key distinctions
Three conditions before calculating.
Coverage and control are traded against each other
Intensive distribution maximises availability but often reduces command of assortment, advice and experience. Selective or exclusive distribution increases control at the price of narrower coverage. The right level depends on the role of the category, the capacity to serve and the cost of one additional point of presence.
Margin and volume must share one account
D2C, marketplace, retail and partners do not carry the same commissions, acquisition costs, returns, inventory, lead times and services. A channel with an apparently high margin can destroy contribution if it displaces more profitable sales or adds under-used infrastructure.
Data and intermediation create dependency
Access to customer data, to listing rules and to an audience can accelerate growth while increasing dependency. That value must be counted once only, and tied to usage rights, activation capability and a realistic exit option.
Method
Build a channel architecture before selecting the players
- 01Choose the degree of intermediation
Define what must stay direct, what can be entrusted to an intermediary and what justifies a hybrid model; separating acquisition, transaction, delivery, advice, service and loyalty avoids reducing the choice to direct versus indirect.
- 02Set the coverage intensity
Arbitrate between intensive, selective and exclusive distribution according to expected accessibility, product complexity, need for advice, positioning, capacity to control and deployment cost; the chosen level must be observable by territory and segment.
- 03Assign a role to each channel
State the contribution expected from D2C, marketplaces, retail, distributors and partners; a single-channel setup favours concentration, multichannel juxtaposes several routes, and omnichannel organises their transitions, their data and their responsibilities.
- 04Measure the economics and the transfers
Compare net price, commission, acquisition, logistics, inventory, returns, service, fixed costs and working capital; measure genuinely new sales, cannibalisation, transfers between channels and partner reaction over the same horizon.
- 05Install coexistence and exit
Document price, assortment, promotion, attribution, data-sharing, compensation and partner treatment; every channel receives a conflict signal, a contribution threshold, an owner, a review date and a condition for reduction or closure.
The distribution architecture matrix
The matrix compares economics, coverage, speed, control, data, reversibility and conflict. The scores never authorise a channel that stays below the contribution threshold.
| Criterion | Indirect | Direct | Decision question |
|---|---|---|---|
| Annual contribution | €152k | €80.6k | Economic threshold |
| Coverage | 9/10 | 5/10 | Market access |
| Speed of rollout | 8/10 | 4/10 | Real capacity |
| Control of the experience | 5/10 | 9/10 | Command of the journey |
| Customer data | 3/10 | 10/10 | Access and consent |
| Reversibility | 6/10 | 7/10 | Exit cost |
| Conflict control | 8/10 | 4/10 | Risk to address |
| Weighted score | 6.30/10 | 6.85/10 | The economic threshold remains the priority |
Direct controls the relationship better but does not yet clear the economic threshold
An indirect channel produces €152k of annual contribution. A direct channel produces €80.6k after acquisition, logistics, service and fixed costs, while creating data access estimated at €90k.
After deducting €45k of conflict and transfer risk, the adjusted direct value reaches €125.6k. It remains €26.4k below the indirect contribution, despite better control of the experience and of the data.
The partner stays primary. Direct becomes a pilot bounded to a segment where contribution, data and service quality are measurable before any extension.
Acceptance conditions
What must be true to act.
- 01
The direct-or-indirect choice compares the same role, the same territory and the same horizon.
- 02
The intensive, selective or exclusive level is tied to a coverage target and to a capacity to control.
- 03
D2C, marketplace, retail and partners each receive an explicit function in the journey.
- 04
Single, multi or omnichannel is chosen according to the transitions that can really be operated between channels.
- 05
Margin and volume are assessed after full costs, transfers and cannibalisation.
- 06
The value of customer data is bounded by its accessibility, its consent and its activation.
- 07
A channel conflict triggers a documented action rather than an implicit compensation.
- 08
Every channel has an exit condition, an owner and a review date.
Limits
What this analysis does not prove.
- Control and speed scores depend on observed internal capabilities and do not constitute a universal scale.
- A hybrid model can create coordination costs that two separate profit-and-loss accounts do not capture.
- A sale moved from one channel to another is not incremental growth.
- Theoretical access to data is worth nothing without usage rights, quality, identity and activation capability.
References
Works cited.
- Anderson, E. & Coughlan, A. T. (1987), International Market Entry and Expansion via Independent or Integrated Channels, Journal of Marketing.
- Williamson, O. E. (1985), The Economic Institutions of Capitalism.
- Coughlan, A. T. et al. (2006), Marketing Channels.