We’re at the point where we need to get serious about measuring ROI on our referral program, but the math gets weird fast when you’re dealing with US and Russian partners at the same time.
Right now, we’re tracking basic metrics—number of referrals, conversion rate, revenue per referral. But the problem is that US partners close deals differently than Russian partners. US brands want detailed performance reports and attribution. Russian brands want to see volume and speed. We’re also paying partners different rates depending on market, and I’m not even sure if we’re comparing apples to apples anymore.
Plus, there’s the question of what we should actually be optimizing for. Short-term referral volume? Long-term partner value? Revenue per partner? We’re spending time and resources on partners that might not be the most profitable, and I can’t tell if it’s because we’re measuring the wrong thing or because the partnerships actually aren’t working.
How are you guys actually structuring your ROI analysis when you’re managing referrals across different markets and partner types? What gets measured, and what do you actually use that data for?
Okay, this is exactly what I deal with. The key is creating normalized metrics that let you compare across markets without losing context.
Here’s my framework: First, separate conversion metrics (did the referral turn into a deal?) from value metrics (what’s the deal worth adjusted for cost?). Conversion metrics let you see efficiency. Value metrics let you see profitability.
Second, calculate ROI per partner, not globally. Track: referrals sent → deals closed → revenue → minus your cost to manage that partner. Then rank. You’ll immediately see which partners are actually profitable. Some will have high volume but terrible margins. Others will be lower volume but highly profitable.
Third, adjust for market differences explicitly. Create a conversion rate baseline for each market. If Russian partners close at 25% and US partners close at 15%, that’s your baseline—don’t penalize the US partners for that. Instead, look at how each partner performs relative to their market baseline.
I’d also separate operational costs from referral costs. If managing a US partner takes 3x the time, that affects ROI too. When I started tracking that, suddenly some “high performer” partners dropped off because they were eating all my team’s bandwidth.
One more practical tip: create tiered targets. Decide what you need to see from a partner to keep them. Maybe it’s “minimum 3 referrals per month with 20% conversion rate” or “$5K revenue per partner per quarter.” Then track partners against those targets monthly. Partners who consistently miss? Those conversations need to happen. Partners who exceed? Double down on the relationship.
Without tiers, you end up managing partners emotionally instead of strategically, and that kills your ROI fast.
We struggled with this too. Here’s what changed things for us: We realized we were measuring activity instead of impact. Referrals sent is activity. Revenue closed is impact. We started reporting two separate dashboards—one internal for our team (showing activity, process, effort) and one external for partners (showing impact, revenue, results). The external dashboard is what actually matters for ROI.
Also, we found that market differences matter way more than we thought. A US brand might take 90 days to close but pay 3x. A Russian brand might close in 30 days but pay 1x. The ROI timeline is completely different. We started measuring both “close time” and “revenue per day” separately so we could actually compare.
One thing that helped: we color-coded our referral partners by profitability tier and literally stopped sending low-tier partners new referrals. Counterintuitive, right? But the bandwidth we freed up let us focus on high-ROI relationships, and overall program ROI went up.
Step back and ask: what decision does this ROI number actually drive?
If you’re trying to decide whether to keep a partner, that’s a different ROI calculation than if you’re trying to decide how much budget to allocate next quarter. If you’re trying to pitch the program to leadership, that’s different again.
What I’d recommend: Create one master ROI model that shows total program performance (all referrals, all markets, all partners rolled up). Then create specific models for each market, each partner type, each product line—whatever segments matter to your business.
For cross-market work specifically, I’d track these four metrics religiously:
- Referral conversion rate by market (to understand market differences)
- Revenue per referral by market (to see value differences)
- Cost to acquire a partner by market (to understand investment required)
- Partner lifetime value by market (to see long-term profitability)
Then overlay your partner tiers. You’ll see patterns fast—like maybe your US partners deliver high LTV but slow conversions, while Russian partners do the opposite. That’s data you can act on. From there, you optimize your partner mix and allocation strategy.
Real talk: most agencies are terrible at ROI tracking because we don’t want to see that most partnerships are barely breaking even. Once you actually measure it, you realize you’re burning time on low-value relationships.
What’s worked for us is brutal simplicity. We track: revenue per partner per month. That’s our main metric. We also track referral volume and close rate, but everything else is secondary. Every month, we rank partners by revenue. Bottom 25%? We either renegotiate terms or stop actively referring to them.
For cross-market stuff, we also account for deal size differences. US deals tend to be bigger but take longer. Russian deals close faster but smaller. So we measure revenue per quarter, not per month, and we’re patient with partners in their ramp-up phase.
One more thing: we tie partner comp to actual results. We don’t just say “we’ll send you referrals,” we say “here’s our referral process, here’s what we track, and here’s how we’ll compensate you based on what actually closes.” That alignment makes the ROI conversation way clearer.