You work as a product manager at an e-commerce company. Your team prioritizes a performance tweak you hope will improve page load time. You release an improvement and find that time spent on site increases, correlating with a higher average order value (AOV) and increased ad revenue.
Hypothesis-->Experiment-->Win!
You change jobs and now work at a B2B SaaS company. Some key details:
Your team strongly believes that the product needs to have top 90th percentile, consumer-grade UX—as good as, or better than, the other apps your customers use daily.
You also hypothesize that your product has a virtuous loop/flywheel/network effect. Encouraging customers (and making it easy for them) to automate vendor transactions positively influences their relationships with vendors. This ultimately helps them lower costs and enables you to build a competitive moat using all that data. When you activate this flywheel, it generates outsized benefits for both your customers and your company—or so you think, based on faint signals.
One of the key benefits of your product occurs outside the product itself and is influenced by many environmental factors. Dysfunctional companies won't magically become functional by using your product, but it can help healthier companies improve.
Your team has an "irrational" belief in product quality. They don't see quality and speed as trade-offs. Instead, they believe quality has many positive second and third-order effects—increasing Revenue, protecting Revenue, avoiding costs, and reducing costs.
So: e-commerce vs. a typical B2B SaaS product.
The difference here highlights one of the biggest challenges in product management. In cases where there's a short hop from improvements to dollar signs, you don't necessarily need strong conviction. But in cases involving short- and long-term effects, where causality is hard to determine, and dealing with abstract strategic bets...you need conviction. You must make leaps of faith and attempt to shape a future that may not yet exist and can't be extrapolated from existing data.
Even in the e-commerce example, factors like brand loyalty, customer lifetime value (CLV), and organic word-of-mouth are harder to quantify. These require stubborn diligence and trust in early but inconclusive indicators.
The high-performing product teams I've met worldwide—especially those operating in contexts like the second example—almost always have conviction AND a willingness to be proven wrong and change course. Convictions loosely held. Their key strategic insights frequently come from qualitative research and keen market observation. Ultimately, they're willing to make a leap of faith.
I remember conducting a workshop with a team that included their finance partner. The team was doing amazing work then and continues to excel. They mix a lot of controlled experiments and quant, with deep qualitative research.
During the workshop, the finance partner asked, "But how will we REALLY know if that's the right metric?"What transpired was enlightening:
X: "If it goes up, and by next year, revenue reaches X, and our margins hit Y, will we worry too much about whether it's the perfect metric?"
F: "Hmmmm…"
X: "Would you prefer we just do the bare minimum rather than building confidence in these ideas and experimenting with our strategic assumptions?"
F: "No…"
X: "If this were the perfect metric, and we knew it unlocked long-term growth, wouldn't that be suspect? Wouldn't our competitors likely have figured it out, too—if it were that straightforward? Maybe uncertainty is a signal we're on to something."
F: "Huh, right…"
X: "Revenue is concrete. But there are ways to boost Revenue in the short term that can ruin a company, right? So we shouldn't solely rely on that."
F: "Right…"
X: "If this metric helps us stay focused and improves decision-making, isn't that valuable?"
F: "I get it…"
In summary, no magic measurement trick will give your team ultimate conviction for the things that matter in the long term. You can build confidence, get closer, and "measure what matters." You can remain open to feedback. But what truly matters will often be harder to quantify and requires a leap of faith and conviction.
This is excellent John. I’ve often felt that product velocity is essentially a function of thrust (conviction, team/talent, clarity) minus drag (risk aversion, coordination costs, WIP).
The chase for metric purity is often rooted in risk aversion & mitigation more than anything else. It’s also why [measure what matters] might be true, but not everything that matters can/should be measured.
OTOH conviction can act as “jet fuel” to achieve “escape velocity” in the face of substantial ambiguity and risks.
How do you think this more nuanced view of metrics-based strategy for big bets / experimentation fits together with "North Star"-style metrics? I feel like this post is way more relatable and representative of the early-stage startup experience, but I'm guessing the two aren't mutually exclusive concepts from your perspective?