Partner Success Has No Benchmarks Yet. That's the Opportunity.

Ask a Customer Success leader what good looks like and they'll bury you in numbers. SaaS Capital's 2025 retention benchmark puts median NRR at 102% for companies with ACVs between $25,000 and $50,000, with a top quartile of 111%. High Alpha's 2025 benchmarks say retaining roughly nine out of ten customers is now the norm. Gainsight can tell you that 72% of CS teams use health score as their primary non-revenue metric and that CS qualified leads jumped from 27% to 52% in two years. Whatever your segment, someone has published a bar to measure against.

Now ask a Partner Success leader the same question. How many partners should one partner success manager cover? What belongs in a partner health score? What's a good partner-attached retention rate? You'll get opinions. You won't get benchmarks, because they don't exist.

I went looking. Across the 2025 and 2026 public research from Omdia, Forrester, PartnerStack, Clazar and Partner Insight, and Crossbeam, there's no benchmark for partner-success staffing ratios, no standard for partner health scoring, and no Partner Success KPI with the shared meaning that NRR and GRR carry in Customer Success. The evidence base is strong on ecosystem revenue motions, marketplace maturity, and what companies hire for. On how to actually run Partner Success day to day, it's close to empty.

Most people would read that as a weakness of the field. I read it as the most interesting opening in post-sale operations right now.

What the evidence actually covers

The money is well documented. Omdia sizes business SaaS at about $420 billion, growing 22.3% a year, and argues partner ecosystems are becoming critical to efficient growth, with Microsoft's 95% partner model and Salesforce's 75% as the mature examples. Forrester's 2025 summary says 67% of surveyed organizations expect indirect revenue to grow above the prior year. PartnerStack's 2026 research says 69% of companies plan to increase partnership investment.

The execution gap is documented too. Clazar and Partner Insight report that 89% of companies are listed on a cloud marketplace but only 22% get more than 20% of their revenue through it. PartnerStack finds the top blocker to partner-driven revenue is alignment across teams, at 37%, and that only 42% of companies use multi-touch attribution.

Even the hiring market is legible. Current partnerships job descriptions ask for co-sell pipeline through AWS ACE, Microsoft Partner Center, and Google Cloud co-sell, plus deal registration, marketplace optimization, and partner-sourced revenue targets. Companies know what they want these leaders to produce.

So we can describe the investment flowing in and the pre-sale motions it funds. What we can't describe, with any shared standard, is the part that happens after the customer signs.

The missing middle

Here's what public data can't answer today. Whether a partner success manager should cover ten partners or a hundred. Which inputs predict that a partner will help customers succeed rather than just resell to them. Whether partner-attached accounts should renew better than direct ones, and by how much. What enablement investment per partner tier is normal.

Omdia says service partners fill implementation gaps and improve retention and customer ROI. That's directional support for the whole discipline, but it comes with no operating standards attached. The closest thing to an operating pattern in public view is Crossbeam's LeanData case: when a customer health score drops to red and the account overlaps with a consulting partner, CSMs get tasks in Slack and Salesforce to engage that partner before the risk hardens. It's a real, wired-in workflow. It's also one company's example, not a benchmark.

The gap makes sense when you consider where Partner Success sits. It's young. It spans organizational boundaries that analysts usually study separately, partnerships on one side and post-sale on the other. And the measurement foundation underneath it is still weak: when fewer than half of companies use multi-touch attribution, standardizing anything downstream of attribution was never going to happen first.

Why this is the opportunity

When a discipline has no benchmarks, the operators who write practicable rules get to define what good means. Customer Success went through this exact phase. The NRR conventions, health score practices, and coverage models everyone now quotes were written by practitioners years before analysts blessed them. Partner Success is at the same point on the curve, and three areas are wide open.

Partner health. Not a satisfaction survey. A weighted model built from operational inputs: certifications kept current, implementation quality, time-to-value on partner-led deployments versus direct ones, escalation frequency, participation in expansion. Then backtest it against what actually happened to the customers those partners touched. If the score doesn't predict outcomes, change the inputs, not the conclusion.

Enablement tied to outcomes. Most partner enablement is measured by consumption: people trained, content downloaded, certifications issued. Consumption says nothing. The rule worth writing links enablement to customer results: do partners who complete a given program onboard customers faster, drive deeper adoption, renew cleaner? If you can't draw that line, you're running a training catalog, not Partner Success.

Partner-attached retention plays. The LeanData pattern generalizes. Wire partner overlap into your risk workflows so that when an account goes red and a capable partner is attached, engaging that partner is a defined play with an owner, an entry condition, and an exit condition. Not an improvisation someone remembers to try in the second escalation call.

None of this requires new software. It requires writing the rules down, running them, and keeping what survives contact with real accounts.

One honest caveat

Much of the evidence in this series comes from vendor research: Gainsight, ChurnZero, PartnerStack, Clazar, Crossbeam, and their peers. That work is current, specific, and closer to workflow reality than most analyst reports, but it's vendor-led and survey-based, so treat it as directional, not as law. The same honesty applies to the operating rules you write yourself. Version one will be partly wrong. Backtest it, revise it, and say so when you do. That discipline is what turned Customer Success from a philosophy into a profession, and it's what will do the same for Partner Success.

This post closes the series on post-sale and partner operations in the AI shift. The earlier pieces, on partnerships as an operating layer, AI's admin-first ROI, the ways retention metrics lie, what companies screen for in senior post-sale leaders, and running programs across markets, are all at https://www.santiagomarin.net/blog.