Key Partnerships: What Goes in This Business Model Canvas Block

What the Key Partnerships block means

The Key Partnerships block is the network of suppliers and partners that make your business model work. It sits on the left side of the canvas, in the infrastructure zone, and it answers a question most founders skip: what should we not build, do, or own ourselves?

Every business model rests on things the company does not control. A software company runs on someone else’s cloud. A restaurant chain depends on food suppliers. A marketplace depends on payment processors and app stores. This block forces you to name those dependencies explicitly, so you can manage them instead of discovering them the day a contract changes.

Osterwalder defines four types of partnership:

1. Strategic alliances between non-competitors: a bank partnering with an accounting software company, for example.
2. Coopetition: strategic partnerships between competitors, such as rival airlines sharing a booking alliance.
3. Joint ventures: two companies creating a new business together that neither would build alone.
4. Buyer-supplier relationships: arrangements that assure a reliable supply of something critical.

Most early-stage canvases are dominated by type four. That is fine. The point is not to have all four types; the point is to know which type each partnership is, because each type carries a different risk profile.

What goes in this block

Partners end up in this block for one of three reasons:

• Optimization and economies of scale. You outsource something because a partner does it cheaper or better than you could. Cloud infrastructure is the classic case: almost no startup should run its own data center.
• Reduction of risk and uncertainty. In an uncertain market, an alliance spreads the bet. Companies form standards alliances precisely so that no single player carries the risk of backing the wrong technology alone.
• Acquisition of particular resources and activities. Licenses, distribution access, customer access, or capabilities you cannot or should not build in-house. When a partner performs one of your key activities or supplies one of your key resources, they belong in this block.

The filter that keeps this block honest: only partners whose loss would materially damage the model belong here. Your electricity provider is a supplier. Your accountant is a vendor. Neither is a key partnership, because both are replaceable in a week. The payment processor that handles 100 percent of your revenue is a key partnership, because replacing it is a quarter-long project and losing it stops the business.

The guiding questions professionals ask

Consultants and experienced facilitators work through this block with a specific set of questions:

• Who are our key partners? Who are our key suppliers?
• Which key resources are we acquiring from partners?
• Which key activities do partners perform for us?
• What can’t we, or shouldn’t we, build in-house?
• Are these optimization partners, strategic partners, or co-creation partners?
• What is the dependency risk? What happens if this partner raises prices 3x or shuts us off tomorrow? (Ask this hard about app stores, ad platforms, and sole suppliers.)
• What does the partner get out of it? Is the partnership durable, or one-sidedly wishful?

That last question is the one amateurs skip. A partnership only holds if both sides win. If you cannot articulate the partner’s motivation in one sentence, you have a hope, not a partnership.

Three concrete examples

McDonald’s: the partnership that is the business

McDonald’s is the cleanest illustration that Key Partnerships can be the center of a model, not a side note. Roughly 95 percent of its 45,000+ restaurants are run by franchisees. The franchisees contribute capital and labor; McDonald’s contributes the brand, the operating system, and usually the land underneath. Growth is largely funded off McDonald’s balance sheet, by partners.

Its canvas also lists dedicated long-term suppliers (Tyson through Keystone, Lamb Weston for potatoes, Coca-Cola, which treats McDonald’s as its largest restaurant customer and provides bespoke support), delivery platforms (long-term global deals with DoorDash and Uber Eats, negotiated from scale), technology partners like Google Cloud for in-restaurant edge computing, and entertainment IP partners such as film studios for Happy Meal toys and music artists for celebrity meals. Notice what each partnership does: franchisees fund growth, suppliers secure quality and cost, delivery platforms add an occasion, tech partners buy speed, and IP partners rent cultural relevance. Every entry maps to a reason. The full breakdown is in the McDonald’s business model canvas example.

A B2B SaaS startup: infrastructure and distribution

A typical 15-person SaaS company lists three or four partnerships. AWS or Google Cloud hosts the product (optimization: no data center). Stripe processes every dollar of revenue (acquisition of a capability: payment infrastructure is not worth building). A systems-integrator or agency channel resells into enterprises the founders cannot reach directly (acquisition of customer access). Each is genuinely key: losing the cloud provider means a migration measured in months, losing the payment processor means revenue stops, losing the channel partner means the enterprise pipeline dries up. The founders should also write down the risk column: what does an AWS price increase or a Stripe account suspension do to the model?

A specialty food brand: co-manufacturing and retail

A packaged-food startup selling hot sauce lists its co-manufacturer (the partner performs the key activity of production, because building a certified facility would consume all available capital), its primary pepper grower under a season-ahead contract (buyer-supplier relationship securing the one input that cannot be swapped without changing the product), and its regional grocery distributor (customer access the brand cannot buy any other way at its size). Three entries, each irreplaceable in under 90 days, each with a clear motivation on both sides: the co-manufacturer fills capacity, the grower gets guaranteed volume, the distributor gets a differentiated product for its category.

Common mistakes (and the fix for each)

1. Listing every vendor. The canvas fills up with the web host, the office landlord, the accountant, and the courier service. None of these are key. The fix: apply the material-damage test to every entry. If you could replace the partner in a month without the model flinching, delete the note.

2. Wishful partnerships. “Partnership with Google” or “Distribution deal with Amazon” written as fact when no conversation has happened. This is the single most credibility-destroying entry a canvas can carry in front of an investor. The fix: unconfirmed partnerships are hypotheses. Mark them as such, visually and verbally, and treat “can we actually get this partner?” as an assumption to test, not a box already filled.

3. Ignoring platform dependency risk. A model that lives or dies by one marketplace’s algorithm, one app store’s policies, or one supplier’s goodwill, with that risk named nowhere on the canvas. The fix: for every partner, write one line answering “what happens if they raise prices 3x or shut us off?” If the answer is “the business ends” and you have no mitigation, that is the most important sentence in your entire canvas.

How Key Partnerships connects to the rest of the canvas

The consistency check reviewers run is simple: every key partner must exist to support something on the right side of the canvas. If you cannot draw an arrow from a partner to a value proposition, a channel, or a revenue stream, the note should come off the board.

The tightest connections run within the infrastructure zone. Partnerships are the alternative to ownership: anything a partner supplies is a key resource you did not have to buy, and anything a partner performs is a key activity you did not have to staff. That trade shows up directly in the cost structure, usually converting fixed costs into variable ones. McDonald’s is the extreme case: by pushing operations onto franchisee partners, corporate exported food costs and labor inflation off its own P&L.

Partnerships also feed the right side. A delivery platform is simultaneously a partner and a channel. A reseller is a partner and a customer-relationship owner. When one company plays two roles on your canvas, note it in both blocks, because the dependency is doubled.

Where it appears in the fill order

Key Partnerships is block 8 of 9 in the recommended fill order: after Customer Segments, Value Propositions, Channels, Customer Relationships, Revenue Streams, Key Resources, and Key Activities, and just before Cost Structure. The logic is deliberate. You can only decide what not to do yourself after you know what must be done at all. Resources and activities define the workload; partnerships decide which parts of that workload someone else should carry; cost structure then prices the whole arrangement.

The full sequence, with the reasoning behind it, is in how to fill in a business model canvas. If you want to work through this block on a real canvas today, grab a free business model canvas template in Word, PowerPoint, Excel, or PDF and start with the material-damage test on your current vendor list. Most teams find they have two or three genuine key partnerships and a dozen vendors they had been mislabeling.