Case study. Cloud Cruz is a fictional company and every number is synthetic. The program design, the systems and the decisions are my work.Fictional company, synthetic data, my real work.

Tony GilharryTalk to Tony

A partner program, designed from zero.

Cloud Cruz had customers and no partner program. I designed one from zero and built the systems to run it, the way I've built them at WP Engine, AWS and Webflow. In the modeled Year 1, 445 customers went live through partners against a target of 400, on a partner team that stayed at 9.

Read the Program BlueprintSee the program on one map

The modeled Year 1, against its targets.

Synthetic results from the program's data generator, on a fictional calendar.

  • Customers live through partners, the north star

    400Target445Year 1

  • Partners activated with a live joint customer

    250Target264Year 1

  • Partners activated within 90 days of signing

    20%First cohorts40%Latest cohorts

  • People on the partner team, for about 1,000 partners

    9Apr 20279Mar 2028

The problem.

Cloud Cruz sells the data and AI platform that apps and agents run on. Its buyers are mid-market software teams with a mandate to ship AI this year, and most of them don't have the data engineers to do it alone.

Agencies and AI consultancies could build with those customers. Software vendors could embed the platform. Cloud marketplaces could reach enterprise data teams. None of it was organized: no tiers, no incentives, no system of record, no way to know which partner moved a deal.

The goal I set: 400 customers live through partners in Year 1, on a partner team that stays at 9 people.

Agencies, SIsISVs, tech partners0 DevelopAgentic PDR1 AcquireFit score, terms2 Onboardintrow, Tipalti KYC, Slack3 ActivateHealth check in Salesforce4 EnableCourses, Pearson VUE, Credly5 Co-sellSalesforce, Crossbeam6 MaintainZendesk, the Support Hub7 OptimizeOne scorecardCloud marketplacesAWS, Google Cloud, AzureHeld 60 daysSalesforce ledgerPayoutTipalti

The whole program on one map.

Lines are the paths a partner, or its money, travels. Stations are the eight stages and the systems that do the work at each one.

Choose any station to see who owns it and how it is measured.

Drawn as a transit diagram, after Harry Beck (London, 1933) and Massimo Vignelli (New York, 1972): a partner needs what a rider needs, which stage they're at and what gets them to the next one.

  • Services pathAgencies, consultancies and SIs, from launch on Apr 1, 2027
  • Technology pathISVs and technology partners, from Oct 1, 2027
  • The lifecycleEvery partner, whichever path it joined on
  • Marketplace co-sellDeals through AWS, Google Cloud and Azure, from Oct 1, 2027
  • The moneyCommissions, earned when the customer pays and held 60 days

Cloud Cruz is fictional. Every partner and number here is synthetic.

How the program works.

The map shows the eight stages. Three more things to know: who the partners are, the tiers they climb, and the play they sell. The Lifecycle covers each stage, and any term you don't know is in the glossary.

Who the partners are

  • Services path

    Agencies, consultancies and SIs. Build and implement Cloud Cruz for customers.

    Opens Apr 1, 2027, at launch

  • Technology path

    ISVs and technology partners, then resellers. Embed Cloud Cruz, integrate with it or resell it.

    Opens Oct 1, 2027

Anchor
Holding-company data and digital units. Invite only, white glove, the route to large brands.
Scale
The digital arms of large consultancies. The enterprise route.
Specialist
Independent digital and data agencies. Select and Premier targets.
Builder
Boutique AI studios and consultancies. The beta cohort and the volume.

The tiers they climb

Partners earn tiers by proving capability and customer results, counts first and revenue later. A partner is promoted within 30 days of qualifying.

  1. Member

    10% commission

    Takes An agreement and a named contact. Free.

    Unlocks Deal registration, courses, the community

  2. Select

    15% commission

    Takes 4 certified people, 3 live joint customers, 1 public reference

    Unlocks Co-selling with Cloud Cruz account executives

  3. Premier

    20% commission

    Takes 10 certified people (3 at Level 2), 10 live customers, 3 references, $175K sourced revenue

    Unlocks A co-marketing fund and a dedicated Slack channel

  4. Elite

    25% commission

    Takes By invitation: 25 certified people (8 at Level 3), 25 live customers, 6 references, $500K

    Unlocks Specialized campaigns and a joint business plan

The solution play partners sell

The 30-day Launch plan. A partner takes a customer from kickoff to production on Cloud Cruz in 30 days, led by a Level 2 technical person. On enterprise deals Cloud Cruz pays for the partner's first consulting and hands-on hours, so the customer gets a fast, low-risk start and the partner wins the business. It's what partners sell, and why customers buy through them (decisions 45, 52, 53, 61).

Six builds on one line.

Building nowPropensity and tier agent

  1. Shipped Sep 2026

    Foundation

    What does the program's data look like, and what can AI safely query?

    All 8 stages

    Read the code

  2. Shipped Oct 2026

    Program Blueprint

    How do you design a partner program from zero, stage by stage?

    All 8 stages

    Read the code

  3. Shipped Oct 2026

    Measurement Layer

    What did partners source and influence, and what did it cost?

    Stages 5, 7

    Read the code

  4. Building now

    Propensity and tier agent

    Which partners deserve investment, and which tiers are wrong?

    Stages 0, 1, 6

  5. Planned

    Answer Engine and deal registration triage

    What do the rules say, and which registrations need a person?

    Stages 4, 5

  6. Planned

    Year-1 review

    What worked in Year 1, where did it break, and what changes in Year 2?

    Stage 7

Cloud Cruz is fictional. Every partner and number in its program is synthetic.

How it was decided.

I make every call. AI researches the market, lays out the options and builds what I decide. 110 decisions so far, each logged with the options considered and the reason for the choice: read the decision log on GitHub.

The hardest calls, like the Year-1 targets and the commission rates, went through a simulated leadership room: sales, finance, partner and product voices that disagree on purpose, with me approving the result. It's how I'd run a real program: decisions in the open, reasons written down.

Where to start.

Pick what you hire for. Each one starts with finished work.

  1. Partner programsHow do you design a partner program from zero?The Program Blueprint
  2. Running the programHow does the program run day to day?The Salesforce setup
  3. Implementation and enablementHow do partners get customers live, fast and right?Credentials and the 30-day Launch plan
  4. Where AI does the workWhere does AI do the routine work, and where does a person decide?Systems and the AI agents