Building Startups that Last (Part 4 of 5): Creating a Betting Culture

This is Part 4 of a 5-part series on the critical factors that startups need to win.

The hardest decisions in startups aren't about execution—they're about direction. Should we pivot? Double down? Try something adjacent? Most founders agonize over these calls because they're trying to predict the future. But prediction is a losing game.

What separates companies that last from those that flame out isn't better forecasting. It's better betting. And better betting comes from a culture that knows when to leap and how to decide.

At Parse.ly, we made one of these leaps in 2011. We'd built a recommendation engine for publishers—think "related articles" widgets. We had about $200K in ARR, growing 20% quarter-over-quarter. Not bad for a small team. But I could feel us hitting a ceiling. The market was getting crowded, our growth was flattening, and the product felt increasingly commoditized.

We had two choices: optimize what we had or leap to something new. We leaped. We pivoted to analytics—a completely different product serving the same customers. As we moved away from recommendations, revenue dropped to near zero. The team was confused. Some customers churned. It was terrifying.

But within 18 months, we hit $2M ARR and never looked back. That pivot set us on the path to building a market-leading SaaS analytics business that Automattic acquired in 2021.

The question isn't whether to make these leaps—you will have to. The question is: how do you know when to leap, and how do you decide where to leap to?

Strategy 1: The Local Max Trap

There's a concept from optimization theory that every founder should internalize: the local maximum versus the global maximum.

Imagine you're climbing a mountain in thick fog. You can only see a few feet ahead. You keep climbing upward until you reach a peak. Congratulations—you've reached the top! Except you haven't. You've reached a top. There are taller mountains all around you, but you can't see them through the fog.

This is the local maximum problem. And it's endemic in startups.

You build a product. It gets traction. You optimize it. Growth accelerates. Then it plateaus. You optimize harder—better sales process, improved onboarding, feature enhancements. Growth edges up slightly, then plateaus again. You're at the top of your mountain. But it's not the tallest mountain.

We hit this more than once at Parse.ly. After the pivot, we thought winning content analytics for digital media was the peak. Then we realized a much higher peak existed: marketers.

The only way to reach a higher peak is to go down first. That means intentionally making your metrics worse in the short term to reach a better position in the long term. This is psychologically brutal. Everything in your body says: "We're winning! Keep climbing!" But the math says you're trapped.

How do you know you've hit a local max?

Three signals:

Optimization produces diminishing returns: You're working harder for smaller gains.
Your best customers want something adjacent: At Parse.ly, our happiest recommendation customers kept asking, "Can you show us the data behind the recommendations?" That question was the seed of the analytics pivot.
The market is compressing your margins: When competitors flood in and prices drop, you're probably in a commodity mountain range. Time to find a new range.

The key is making everyone in your company fluent in this mental model. When your team understands that going down can be the only way to go up, they'll stop resisting necessary pivots. They'll start spotting the signals themselves.

Strategy 2: Effectual vs. Causal Decision Making

Once you've decided to leap, how do you figure out where to leap to? This is where most founders reach for the MBA playbook: market sizing, competitive analysis, financial projections, strategic planning.

This is called causal reasoning: you set a goal, analyze paths to reach it, and pick the best path. It works great when you have predictable environments and abundant resources.

Startups have neither.

There's a better approach for startups, and it's called effectual reasoning. It was developed by Saras Sarasvathy (now at UVA's Darden School), who studied how expert entrepreneurs make decisions. She found they reason in almost the opposite direction of MBAs.

Causal reasoning (MBA style):

Start with a specific goal
Analyze the market and competitive landscape
Calculate expected returns and risks
Gather necessary resources
Execute the plan
Avoid surprises

Effectual reasoning (startup style):

Start with means at hand (who you are, what you know, who you know)
Calculate affordable loss (what you can afford to lose)
Form partnerships to expand resources
Leverage contingencies and surprises
Shape markets rather than predict them

At Parse.ly, the analytics pivot wasn't the result of a market analysis. It came from effectual reasoning:

What we had: A team that understood publisher workflows, relationships with 30+ media companies, and deep expertise in content data pipelines.

Affordable loss: We had about 12 months of runway. We could afford to spend 3 months building an MVP of analytics before we'd know if it would work.

Partnerships: Our existing customers became design partners. Instead of trying to imagine what "the market" wanted, we built exactly what The Atlantic and Wired asked for.

Contingencies: When Google launched real-time analytics, instead of panicking, we used it as validation and positioned Parse.ly as "Google Analytics for content."

We didn't have a business plan. We didn't do market sizing. We just started with what we had and moved forward.

This is how elvex started too. I didn't analyze the "AI management platform" market (it barely existed). I started with what I knew: enterprises struggle to deploy AI. We built the first version for ten design partners in 30 days. The market emerged around us.

Why effectual reasoning works for startups

Startups don't have the resources to execute MBA-style causal plans. You can't outspend incumbents on customer acquisition. You can't out-analyze them on market research. And your predictions will be wrong anyway because you're creating new markets, not entering existing ones.

But you can move faster. You can leverage partnerships. You can turn surprises into advantages. You can shape markets around your strengths.

Building the Betting Muscle

Creating a betting culture isn't about being reckless. It's about being honest about uncertainty and systematic about decision-making in the face of it.

Two practices we use at elvex:

Max peak reviews: Quarterly, we examine our key metrics and ask: "Are we optimizing, or are we trapped on a local max?" This isn't a strategy review—it's specifically about identifying diminishing returns. If we spot them, we explore what adjacent mountains might be taller.
Effectual post-mortems: When a bet fails (and they will), we don't ask "Why was our prediction wrong?" We ask: "Did we start with the right means? Did we calculate affordable loss correctly? Did we leverage partnerships and contingencies?" This keeps the focus on decision process, not outcome luck.

The hardest part of building startups that last is knowing when the path forward requires going down first. And the only way to get good at that is to make it a core part of how your team thinks and decides.

Because the companies that win aren't the ones that never make wrong turns. They're the ones that recognize when they're on the wrong mountain—and have the courage to climb a new one.

Next up in Part 5: Farming Talent over Finding Talent.

This is Part 4 of a 5-part series on building startups that last. Read Part 1 here. Read Part 2 here. Read Part 3 here.