I have spent time analyzing the partner programs of the Big Three. The verdict: they are scrambling, and they have not figured it out yet.
They are still designed for the SaaS era. Sell a license, implement the software, renew the license.
Salesforce is the closest to getting it right. They are moving to consumption pricing, charging per action. Smart for them. It aligns their revenue with usage. But they have not solved the partner problem. If a partner optimizes a bot to complete a task in two steps instead of five, the partner just reduced Salesforce's consumption revenue. There is a structural conflict of interest baked into the model.
HubSpot is sending mixed signals. They tightened retention metrics, which forces partners to care about long-term success. But in January 2026, they removed the Managed Services requirement for tiering. They are saying keep the client happy while also saying you do not have to sell managed services. Pick one.
Microsoft is still throwing money at license activations. Incentives are heavily weighted toward Copilot deployments. Classic volume play.
The missing piece across all three programs is optimization. Who pays the partner to make the agent smarter after it is deployed?
The customer thinks the software should just work because AI is supposed to be magic. The vendor thinks the partner should just fix it because that is their job. The partner is stuck in the middle, doing unpaid R&D and prompt tuning to keep churn from spiking.
This is unsustainable. Partners cannot be the unpaid support tier for Big Tech's beta software. The programs need to evolve to reward outcomes, not transactions. Until they do, the smartest partners will build their own IP and service layers on top. Decoupling their revenue from the vendor's broken incentive structure entirely.