You convinced 15 partners to sign. Six months later, 2 sold something and 13 forgot your product exists. The problem wasn't recruiting — it was accepting anyone who showed interest. This article is the framework for structuring the program before you open the door.
- Post 1 — The myth of the channel that sells itself
- Post 2 — How to structure the program: criteria, tiers, and contracts (you are here)
- Post 3 — Recruiting, qualifying, and compensating partners
- Post 4 — Channels in Brazilian B2B SaaS: what changes in practice
In Post 1 of this series, we established that a channel is a distribution model — not a synonym for sales. Now we go operational: who to accept, how to organize, and what goes on paper.
But before the operational details, a real case. In a recent diagnostic with a B2B SaaS startup, the founder wanted to launch a reseller channel to "increase sales." When we analyzed the operation, the picture was different: poorly defined inbound and outbound playbooks, pricing and solution packaging still maturing, a generic ICP, and extremely high churn. The strategy was to increase sales through channel to compensate for churn — exactly the opposite of what they should have been doing. The channel wasn't going to fix the problem. It was going to amplify it.
To make matters worse, their only prior experience with partners had been traumatic. They partnered with a company, worked closely together — including the partner operating inside the startup's office. The partner copied the business model and the software to compete directly. It ended in litigation. No NDA, no intellectual property clause, no contractual protection. A predictable disaster.
Before structuring: why channel?
Before defining criteria, tiers, and contracts, the most important question is: what problem does the channel solve for your business? If the answer is "increase sales," stop. A channel isn't a generic solution. It's a distribution model that serves specific objectives:
- Territorial extension — reaching regions or markets where you have no direct presence and can't justify building your own operation.
- Segment access — reaching a customer profile you can't access with your current team. The partner already has the client base and credibility in that segment.
- Offer complement — creating a more competitive solution by combining your product with the partner's service or product. Together, you solve a problem neither can solve alone.
If the channel doesn't clearly fit at least one of these three objectives, the program starts without direction — and without direction, every partner looks like a good idea.
In the beginning, less is more
The natural temptation is to recruit as many partners as possible. More partners = more coverage = more sales. This logic is false — and dangerous.
Start with 2-3 partners. No more. This approach has a double benefit. First: it's easier to sell the opportunity to the partner. With fewer partners, each one has a broader market and their only competition is third-party solutions — not other partners in the same program. Second: you test the model and mature the process before scaling. You discover what works in enablement, adjust the contract, calibrate margins — all at a controlled cost. Scaling a model that doesn't work is the most expensive way to learn.
The right partner is rarer than you think
Wrong partners aren't neutral. They consume pre-sales resources, generate unproductive support tickets, and — worst of all — burn your brand in the market with poorly qualified approaches.
IPP: the mirror of your ICP
Just as you have an ICP (Ideal Customer Profile), you need an IPP — Ideal Partner Profile. The principle is simple: the partner's customer base must overlap with the vendor's ICP. If there's no overlap, the partner will try to sell to whoever they already know, not to whoever you need.
In practice, this means mapping the partner's industry focus, the size of their clients, the verticals they operate in, and their average deal size. If your ICP is a tech company with 200-500 employees and the partner serves 10-person accounting firms, it doesn't matter how enthusiastic they are — the fit doesn't exist.
Minimum technical capability
Before accepting any partner, define what they need to have — not what they promise to have in 90 days. When I created the first certified partner program at Novell Brazil, each category had specific requirements: certified professionals, demo infrastructure, level-1 support capability. I later brought that model to build channel programs from scratch at Oracle and Informatica — both companies that were 100% direct sales where I created the channel from zero. The principle was the same: partners who couldn't prove capability before entering didn't get in.
For startups, the requirements don't need to be as extensive. But the minimum is non-negotiable: at least 1 person dedicated to selling your product, the ability to run a demo without your team, and basic CRM tools for pipeline tracking.
Portfolio complementarity
Your product needs to fit naturally into the partner's existing offering. If they sell network infrastructure and you offer HR software, their sales rep won't know how to position your product in a client conversation. An "alien" product in the portfolio = zero priority.
The ideal fit is when the partner already solves an adjacent problem. Their rep is already at the right table, having the right conversation. Your product enters as a natural extension of the solution.
Decline the partner if you spot 2 or more of these signals:
- More than 15 products in their portfolio — your product will compete for attention with all of them
- No dedicated sales rep — "everyone sells everything" means no one sells anything
- History of conflict with other vendors — a partner who fights with suppliers is a pattern, not an exception
- Signs without negotiating — a partner who accepts any contract without questioning terms isn't committed
- Demands territorial exclusivity without committing to minimum pipeline
Verifiable commitment
Words are cheap. Commitment is measured in actions. Before formalizing the partnership, verify whether the partner is willing to invest in training (not just send the intern), dedicate at least 1 headcount, and participate in joint marketing activities.
In my experience, the best indicator of future commitment is the effort the partner puts in before signing the contract. A partner who invests time in the qualification process, studies the product on their own, and brings specific questions is the partner who will perform. A partner who wants to sign in the first meeting will forget you by the second week.
Tiers: governance, not gamification
A tier model isn't a loyalty program. It's not a badge on a website. It's the governance mechanism that defines what you invest in each partner — and what you expect in return.
The most common mistake in startups is copying the model from large vendors — 5 or 6 levels with elaborate names, complex criteria, and granular benefits. Those models were built for ecosystems with thousands of partners. If you have 10-30 partners, you need 2-3 tiers at most.
| Dimension | Authorized | Preferred | Strategic |
|---|---|---|---|
| Requirements | Signed agreement, 1 certified person | 3+ sales per quarter, 2 certified people, NPS ≥ 8 | Dedicated team, shared pipeline, regular QBRs |
| Margin | 15-20% | 20-30% | 30-40% + SPIFs |
| Pre-sales support | Self-service + documentation | SE (Sales Engineer) access via ticket | Dedicated SE + direct support |
| Leads | No lead sharing | Qualified leads by region | Shared pipeline + active co-selling |
| Co-marketing | Logo in directory | Co-branded materials, joint events | MDF (Market Development Funds), dedicated campaigns |
| Governance | Automated monthly report | Monthly pipeline review | Formal QBR + joint annual business plan |
Progression criteria — and demotion
Moving up a tier requires objective criteria: minimum revenue in the period, number of active certifications, NPS from clients served by the partner. No ambiguity.
But what few people discuss: demotion criteria are equally important. A partner who was Preferred in Q1 but sold nothing in Q2 and Q3 needs to go back to Authorized. No exceptions. If they don't move down, the tier becomes an honorary title — and the partners who actually perform lose motivation when they see inactive peers enjoying the same benefits.
Programs without demotion criteria create two problems: inactive partners consuming resources (support, materials, SE access) with no return; and active partners losing motivation because effort and inertia get the same treatment. Review tiers every quarter.
The contract that protects both sides
The channel agreement isn't a legal formality. It's the document that defines the rules of the game. Every missing clause is a future conflict — and conflict in a channel destroys trust faster than any competitor.
Territories and accounts
Define who sells to whom. Channel conflict — partner vs. direct sales, partner vs. partner — is the silent killer of programs. In every channel program I built, territory definition was one of the first decisions. Without that clarity, every deal becomes an internal dispute over who's entitled to the commission.
For startups, the practical rule is: start with geographic territories or by vertical. Never overlap two partners in the same territory without clear priority rules.
Deal registration
Deal registration mechanics protect the partner who found and developed the opportunity. Without it, the partner who spent weeks qualifying a prospect loses the deal to a more aggressive partner — or to the vendor's own direct sales team.
Clear rules: the partner registers the opportunity in the system; the vendor confirms within 48 hours; protection lasts 90-120 days (renewable with proven progress); if the partner doesn't move the deal for 30 days, protection expires. Full transparency — both sides see the status in real time.
Vendor SLAs to the partner
The channel agreement isn't one-sided. The vendor has obligations too. Pre-sales response time (24-48h for technical support), turnaround for special discount approvals, SE availability for joint demos. If the vendor doesn't meet its SLAs, the partner loses deals — and trust evaporates.
Minimum performance and data ownership
Minimum performance clauses define the sales volume a partner must deliver per period to maintain their tier. Without them, the program becomes a directory of logos with no results.
Customer data ownership is the most neglected clause — and the one that creates the most conflict on day 365. Who owns the relationship? Who has CRM access? What happens to the data if the partnership ends? Define it on day 1. In my experience, what works is: customer data belongs to the vendor, but the partner retains access while the partnership is active and for 90 days after termination for transition.
Exit clause
Partnerships end. The contract needs to plan for it. What happens to the clients the partner brought in? Who takes over support? What's the transition period? Clients can't be orphaned because the vendor and partner didn't plan for the divorce.
Intellectual property, NDA, and non-compete
Remember the diagnostic I mentioned at the beginning? The partner copied the business model and the software. Litigation. No NDA, no intellectual property clause, no non-compete. When you open your product, your playbook, and your business model to a partner, you're sharing what took years to build. The contract must protect that.
Include in the contract from day 1:
- NDA (Non-Disclosure Agreement) — protects commercial, technical, and strategic information shared during the partnership. Minimum duration: 2 years after termination.
- Intellectual property — defines that code, methodology, materials, and product data belong exclusively to the vendor. Access is not transfer of rights.
- Non-compete — prevents the partner from creating a competing product using knowledge acquired during the partnership. Typical duration: 12-24 months after termination, in the same market/vertical.
What every program forgets to include
Contracts and tiers are the structure. But the program needs operational components that nobody puts in the first draft — and that make the difference between a channel that works and one that exists only on paper.
Sales playbook for the partner
If the partner's sales rep can't qualify and sell without calling your team, you don't have a channel — you have assisted selling through a third-party entity. The playbook must cover: detailed ICP, qualification questions, common objections with responses, demo script, and proposal template.
In every channel program I built, each certified partner received a playbook that enabled their reps to run the entire cycle — from prospecting to close — without needing anyone from the vendor until the proof-of-concept stage. That's a channel. Everything else is dependency in disguise.
RACI matrix
Who prospects? Who qualifies? Who runs the demo? Who delivers the proposal? Who negotiates pricing? Who onboards the client? Who provides post-sale support? If the answer to any of these is "it depends on the deal," you don't have a process — you have improvisation. And improvisation doesn't scale.
The RACI matrix (Responsible, Accountable, Consulted, Informed) clearly defines who does what at each stage. Without it, every joint deal becomes an internal negotiation before it becomes a sale. Understanding how this applies in practice when you need to diagnose and fix revenue bottlenecks →
Enablement calendar
Enablement isn't "whenever we get around to it." It's a fixed schedule. Training dates, product refreshers, certification, feature launches — everything published with 90 days' notice. The partner needs to plan their team's calendar, and "let's schedule something next month" doesn't work.
The minimum format: 2-day initial training at onboarding, quarterly half-day refresher, monthly 1-hour product and market update. If you're managing multiple partners and want to keep your GTM synchronized between direct sales and channel →, this calendar is non-negotiable.
Channel conflict policy
When two partners compete for the same client, who wins? When the vendor's direct sales team competes against a partner on the same prospect, who stands down? Without written, published, and enforced rules, trust dies at the first dispute.
The policy needs to cover at least three scenarios: partner vs. partner (priority to whoever registered first), direct sales vs. partner (absolute protection of the registered deal), and conflict over existing accounts (ownership rule based on who brought the client). You don't need more than 2 pages. You need clarity and consistency in enforcement — because the first time you make an exception, the policy ceases to exist.
🌳 Is the partner ready for the program?
Does the partner's customer base overlap with your ICP by at least 40%?
→ Decline. Without ICP overlap, the partner will sell to the wrong audience and damage your brand.
→ Continue to Filter 2.
Does the partner have at least 1 person who can dedicate 50%+ of their time to your product?
→ Decline or condition. Without dedicated headcount, your product goes to the bottom of the priority list.
→ Continue to Filter 3.
Does the product fit naturally into the partner's existing portfolio?
→ Decline. An alien product in the portfolio = zero priority for the partner's sales rep.
→ Partner is qualified. Start onboarding at the Authorized tier and review in 90 days.
What this series covers
This is the second of four posts on building a B2B channel program that actually works. In Post 1, we covered why channel isn't a synonym for sales and what needs to be in place before any partnership.
In the upcoming posts:
- Post 3: How to recruit, qualify, and compensate partners — commissions, SPIFs, MDF, and why the biggest mistake is rewarding contract signatures.
- Post 4: What changes in the Brazilian B2B SaaS context — adapting the value proposition, local market specifics, and lessons from Oracle and Informatica in Brazil.
If you're evaluating how to make structural decisions about your sales team before scaling through channels, the article on when (and how) to make the decision to let a sales rep go → complements this discussion.
Conclusion
Structuring a channel program isn't a PowerPoint exercise. It's business engineering: every decision about who to accept, how to organize, and what to put in the contract determines whether the program generates revenue or generates cost.
I created the channel program at Novell Brazil and brought that model to build programs from scratch at Oracle and Informatica across Latin America. The ones that worked had one thing in common: the structure was ready before the first partner signed. The ones I've seen fail — including the recent diagnostic case — had another thing in common: the rush to recruit without having the rules defined.
A channel program without selection criteria, tiers with consequences, and a contract that protects both sides isn't a program. It's a list of partners that will cost you 12 months of learning you could have avoided.