Partner KPIs: The Metrics That Measure a Partnership

A partner program produces a lot of numbers. Signed partners, trained staff, registered deals, closed revenue, portal logins, MDF claims. The hard part isn't collecting them. It's knowing which ones describe the health of a partnership, and which ones just describe activity.

This article is a reference catalogue. It groups partner KPIs by the stage of the partner lifecycle, gives a definition and a formula for each, and separates leading indicators from lagging ones. It then looks at how large vendor programs score their own partners, using their published documentation, and closes with the pitfalls that distort partner reporting. How to arrange these metrics into a one-page report is covered in partner scorecard, and how to credit revenue between partner and vendor is covered in partner attribution.

What a Partner KPI Is

A key performance indicator is a measurable value that shows how effectively an objective is being met. The general idea is covered in KPIs. A partner KPI applies that idea to the channel: it measures what partners do, what they produce, or what the relationship costs.

Partner metrics fall into three groups, and mixing them up is the most common reporting mistake:

• Activity metrics count things partners do: trainings completed, leads submitted, portal logins.
• Output metrics count what those activities create: pipeline, deals, revenue.
• Economic metrics compare output with cost: return on MDF, cost of channel per dollar of revenue.

Activity metrics are cheap to collect and easy to inflate. Output and economic metrics are harder to measure and much harder to fake. A useful set includes all three, weighted toward the last two.

Partner KPIs by Lifecycle Stage

The sections below follow a partner from recruitment through to long-term economics. Every formula is a plain definition. Your own program may define a term slightly differently, and that's fine, as long as you write the definition down and apply it the same way every period.

Recruitment

Recruitment metrics tell you whether you're finding the right partners, not just more of them.

Fit mix matters more than raw volume. Fifty signed partners outside your ideal profile will cost more in onboarding than five inside it.

Onboarding and Activation

Onboarding ends at a partner's first productive deal, so these metrics measure speed and conversion to that point. The process itself is described in partner onboarding.

Fix the activation window (90 or 180 days, say) before you look at the data, and keep it the same across cohorts. Segment by partner type, because a referral partner and a systems integrator will never share a time to first deal.

Enablement and Certification

These metrics ask whether partners have the skills to represent the product. See partner enablement and partner certification for the programs behind them.

Certifications are an activity metric. They become more useful when you check them against output: do certified partners register and close more than uncertified ones in your own data? That's a comparison you can run internally, and it will tell you more than any published average.

Pipeline

Pipeline metrics sit between activity and revenue. They're leading indicators of revenue, with the usual caveat that pipeline is only worth what it converts to.

Registration data is one of the cleanest signals in the channel, because a partner who bothers to register a deal is putting a claim on it. How registration works is covered in deal registration. A low approval rate can mean partners misunderstand the rules, or that your rules are too restrictive, or that channel conflict is rejecting deals your direct team already touched. The number alone doesn't say which.

Revenue

Revenue metrics are the lagging indicators everyone cares about most.

Always state which revenue base you divide by. "Partners drive 30% of revenue" means different things if the numerator is sourced revenue or all partner-touched revenue, and the two should never be quoted interchangeably.

The three comparisons against direct are where partner reporting becomes honest. A partner channel with a lower win rate, smaller deals and a longer cycle than direct may still be worth running, but it costs more per dollar, and you should know that. Compare like with like: same segment, same product, same period. For the broader revenue-metric family, see RevOps metrics.

Retention and Customer Outcomes

Partner-sold customers can churn faster or slower than direct ones. Nobody knows until they measure it.

Retention of the customers a partner brings is a quality check on the partner, not just the deal. A partner that closes quickly and churns heavily is leaking money you've already paid commission on. Some vendor programs build exactly this into tier criteria, as the next section shows.

Engagement

Engagement metrics describe how present a partner is in the relationship.

Define "active" by an output, not a login. A partner who opens the portal weekly and never registers a deal is not active in any sense that matters. And note that response time is a vendor-side metric: partners disengage when the vendor is slow, and that deserves its own line on the report.

Program Economics

This group answers whether the channel pays for itself. See market development funds and partner incentives for what the spend consists of.

Cost of channel is the number executives most often want and teams most often leave out. Include salaries and tooling, not only payouts. A channel that looks cheap because the partner team's time isn't counted is cheap on paper only.

Leading vs Lagging Indicators

Some of these metrics tell you what's coming and some tell you what already happened. You need both, and you need to know which is which.

Leading indicators give you time to act. A falling activation rate this quarter predicts weaker revenue in two or three quarters, depending on your sales cycle. Lagging indicators confirm whether the leading ones were pointing the right way, and they're the ones you can defend in front of a finance team.

A practical rule: review leading indicators monthly and lagging ones quarterly. Check each year whether your leading indicators actually predicted your lagging ones. If registration volume rose for three quarters and revenue didn't move, registration volume isn't predicting anything for your program, and it needs a rethink.

How Vendor Programs Measure Their Partners

Large vendors publish how they score partners. These documents are useful reference points because they show which metrics a mature program chooses to weight, even if you'd never copy the thresholds.

Microsoft: Partner Capability Score

Microsoft's Partner Capability Score documentation describes a composite score across three categories: performance, skilling and customer success. Each of six solution areas carries a maximum of 100 points. Metrics within each category have weights and thresholds, and a partner earns partial points for partial progress.

The categories map neatly onto the lifecycle above:

• Performance uses net customer adds: eligible customers added in the trailing 12 months, with lost customers subtracted from new ones.
• Skilling counts intermediate and advanced certifications held by people linked to the partner's organization.
• Customer success uses usage growth and deployments.

To qualify for a Solutions Partner designation, a partner needs a score of at least 70 points in the solution area, and every metric in that area must be above zero. That second rule is the interesting one: it stops a partner from reaching the bar on a single strength, since all three categories have to show something. The same page says performance and customer success data are typically refreshed by the 20th of each month, and skilling data within 10 days of a certification being completed.

The design lesson is that Microsoft blends output (customer adds), capability (certifications) and outcomes (usage and deployments) into one score, rather than relying on any one of them.

AWS: Services Partner Tiers

AWS's Services Partner tier criteria combine knowledge requirements with experience requirements. On the knowledge side, the Select, Advanced and Premier tiers ask for four, eight and twenty accredited individuals respectively, split evenly between technical and business roles, plus foundational and technical certified staff. On the experience side, the tiers require launched opportunities with a minimum total monthly recurring revenue: three opportunities and at least $1,500 for Select, twenty and at least $10,000 for Advanced, and fifty and at least $50,000 for Premier.

The page also lists additional requirements for Premier, including a formal business plan, an executive business review and sustained attainment of the criteria for more than six months. In effect, AWS pairs a certification metric (coverage) with a pipeline-to-revenue metric (launched opportunities and MRR), then adds a relationship requirement at the top tier.

HubSpot: Points and Retention

HubSpot's partner tier page shows a points-based model. Sourced points are earned by sourcing and closing deals with customers, at 5 points per $100 of MRR in non-growth markets and 10 in growth markets. Total points are the sum of sourced, assisted and managed points, which separates originating a deal from helping on one or managing the account afterwards. The page also states that Elite partners must hold at least 80% average gross revenue retention over the trailing 12 months, and Diamond partners at least 75%.

So HubSpot explicitly separates sourced from assisted contribution and puts a retention floor on its top tiers. That's the same logic as the retention section above, built into the program rules.

Across the three, the pattern is consistent: a revenue or pipeline measure, a skills measure, and (in two of the three) a quality or customer-outcome measure. None of them rely on activity metrics alone.

Pitfalls in Partner Measurement

Key Facts: Partner KPIs

• Microsoft's Partner Capability Score gives each solution area a maximum of 100 points across performance, skilling and customer success, and a partner needs at least 70 points with every metric above zero to qualify (Microsoft Learn).
• AWS's Premier Services Partner tier requires 50 launched opportunities with total MRR of at least $50,000, against 3 and $1,500 for Select (AWS).
• HubSpot requires Elite partners to hold at least 80% average gross revenue retention over the trailing 12 months, and Diamond partners at least 75% (HubSpot).
• HubSpot awards sourced points for sourcing and closing deals, and counts assisted and managed points separately in the total (HubSpot).
• Partner metrics fall into three groups: activity, output and economics. Reliable reporting weights output and economics most heavily.

Vanity Metrics

Signed partners, portal logins, courses completed and leads submitted all rise easily and say little about revenue. They're fine as diagnostics. They're a problem when they become targets, because partner managers will then optimize for them. If a team is rewarded for partners signed, expect a long tail of partners that never sell.

Double Counting Sourced and Influenced

Sourced and influenced pipeline are different claims, and adding them together overstates partner contribution. A deal can be sourced by one partner and influenced by two more, and a single deal can show up three times in a naive total. Keep the categories separate, count each deal once under a single primary role, and report totals only for a defined, non-overlapping set. The crediting rules are the subject of partner attribution.

Mixing Revenue Bases

A partner revenue share built from sourced revenue in one quarter and partner-touched revenue in the next will move for reasons that have nothing to do with partners. Pick the definition, write it into the scorecard and leave it alone.

No Direct Comparison

Reporting partner win rate without direct win rate in the same segment gives you a number with no meaning. The same goes for deal size and sales cycle. Without the comparison, a good number and a bad number look identical.

Averaging Across Partner Types

A blended activation rate across referral partners, resellers and integrators hides what's happening in each. Segment every metric by partner type at least, and by tier or region where volumes allow.

Measuring Activity That Partners Control Cheaply

Registration volume can be gamed by submitting speculative deals. Pair it with approval rate and registration-to-close rate so volume without quality shows up.

Ignoring Vendor-Side Metrics

Partner results depend on what the vendor provides. Slow access, slow deal-registration decisions and slow support all show up as weak partner performance. Track your own response times next to the partners'.

Related Reading

• Partner scorecard
• Partner attribution
• Partner onboarding
• Partner incentives: margins, rebates and SPIFFs
• Market development funds
• Deal registration
• Partner-led growth
• RevOps metrics