Mishneh Torah, Sanctification of the New Month 18-19

Hook

The founder’s greatest delusion is the belief that "the data speaks for itself." You stare at your churn rates, your CAC, and your MRR growth, convinced that if the numbers are black-and-white, the strategy is bulletproof. But here is the truth that keeps founders awake at 3:00 AM: data is not reality; data is a sampling of reality, filtered through the geography of where you are standing.

In Mishneh Torah, Sanctification of the New Month, Maimonides dissects the mechanics of observation. He notes, "It is well-known and obvious that although the calculations indicate that the moon should be sighted... its sighting is only probable." He goes on to describe how a witness in a valley, blocked by a mountain, or obscured by seasonal dust, might miss the very thing that is objectively there.

This is the quintessential founder dilemma: Contextual Bias. You are building a product in a "valley"—a niche, a specific segment of the market, or a particular technological silo. You are making critical decisions based on what you see from your limited vantage point, while competitors—or your customers—are standing on a "lofty mountain," seeing a completely different horizon.

When you scale, you assume your internal metrics represent the universal truth of your product-market fit. But just as the moon’s visibility depends on whether the observer is on a ship in the Mediterranean or a low-lying valley, your business performance depends on the context of the user. A feature that looks like a "must-have" to your early adopters (the high-mountain observers) might be invisible to your mass market (the valley-dwellers).

If you don't account for the "atmospheric conditions"—the cultural, economic, or technical friction—of your users, you will make the mistake of "sanctifying" a month based on incomplete testimony. You will pivot when you should stay the course, or scale when you should be optimizing. As a founder, your job isn't just to look at the data; it’s to understand the positionality of the data. Are you reading the numbers, or are you reading the geography that produces those numbers? Let’s stop confusing the calculation with the event.

Analysis

Insight 1: Truth is Relative to the Observer’s Vantage Point

Maimonides writes, "For the moon will not be able to be sighted by a person in a low place, even when its crescent is large. Conversely, it will be possible for a person on a high and lofty mountain to sight [the moon], even though its crescent is very small."

In business, this is the Observer’s Fallacy. Founders often dismiss user feedback that contradicts their internal dashboards. If the dashboard says the feature is working, you conclude the user who says it isn't is simply "wrong." But Maimonides forces us to accept that the truth of the observation is inextricably linked to the position of the observer. If a power user sees value, they are on the "lofty mountain." If a churned user sees nothing, they are in the "valley." Both are telling the truth. Your job is to map the terrain, not to pick which truth is more valid.

Insight 2: Cross-Examination as a Validation Protocol

The text dictates: "The judges should ask the witnesses, 'Where were you when you saw the moon?'... we suspect [the veracity of] their testimony and subject them to much cross-examination."

Most founders take "witness testimony" (customer feedback) at face value. If three customers tell you they want a specific integration, you build it. That is amateur hour. Maimonides teaches that you must qualify the testimony by the context of the source. Before you allocate engineering resources, ask: "Where were they standing?" Were they in a "rainy season" (a market downturn)? Were they in a "low place" (a legacy infrastructure that limits their ability to see the value)? If you don't cross-examine the context of the feedback, you are building for the exception, not the rule.

Insight 3: The Necessity of Calculated "Buffer" Months

Maimonides explains that the court relies on tradition and calculation to maintain order when observation fails: "Never should there be fewer than four full months in a year, nor should there ever be more than eight full months."

This is a masterclass in Risk Management. You cannot rely on "sighting" (real-time, anecdotal data) alone. If you only operate on what you can see, you will eventually face a "ludicrous and demeaning situation"—the calendar drifts, and you lose synchronization with reality. You need a system of structural, calculated predictability—what we call KPIs and OKRs—to anchor the business when the "clouds" of market volatility make real-time feedback opaque. You need a baseline of "full" and "lacking" months (growth vs. efficiency phases) that you adhere to, regardless of whether the "moon" (the latest trend) is currently visible.

Policy Move

The "Context-First" Feedback Review Policy

To move from impulsive decision-making to disciplined execution, you must implement a formal "Cross-Examination Layer" in your product development process.

The Policy: No significant product pivot or feature release can be approved based solely on "user request" data. Every major feedback loop must be accompanied by a Context Audit.

1. The Source Identification: Before any feedback is presented to the leadership team, the Product Manager must tag the source. Are they a "High Mountain Observer" (Power User/Early Adopter) or a "Valley Observer" (New User/Low Engagement)?
2. The Environmental Check: Similar to Maimonides’ focus on "rainy seasons" vs. "summer," identify the external conditions. Is this feedback coming from users who are in a specific industry sector undergoing change, or those using a specific, outdated integration?
3. The Verification Protocol: If the data is ambiguous, the team is mandated to perform a "Second Sighting." You must find three users from a different geographic or technical "valley" to see if they encounter the same phenomenon.

Metric/KPI Proxy: Context-Weighted Feedback Density (CWFD).

• Definition: The percentage of product changes backed by evidence from diverse user cohorts rather than a single, homogeneous cluster.
• Target: >70% of features must show validation across at least two distinct user personas (High-Mountain and Valley) to move to the roadmap.

By forcing your team to identify the "altitude" of your data, you eliminate the bias that leads to building products for only the loudest, most elevated voices in your user base.

Board-Level Question

"Are we making strategic decisions based on the 'mountain' of our power users, and if so, how are we factoring in the 'valley' of our churned or non-activated users to ensure our calendar doesn't drift?"

This question forces the board to confront whether the strategy is overly optimized for the elite 5%—who enjoy the view from the peak—while ignoring the 95% who are stuck in the "valleys" of the product. It shifts the discussion from what is happening (the data) to why it is happening (the context). If the leadership team cannot account for the "valleys," they aren't leading the whole company; they are only leading the top of the mountain.

Takeaway

Stop being a victim of your own perspective. Maimonides teaches us that reality is a combination of observation and calculation. Your data is the observation; your strategy is the calculation. If you confuse the two, you will eventually find yourself in a "ludicrous" position, chasing a market that doesn't exist. Be a Mensch—own your perspective, cross-examine your sources, and build a system that accounts for the clouds.