Data observability raised a fortune to watch the number. Defining the number raised nothing.

Watching a number is fundable. Defining it is not. One activity has a category, a pitch deck, and a valuation; the other has a recurring meeting that keeps getting moved, and the gap between them is where trust quietly goes to expire.

What happened: data observability became a funded category. Monte Carlo, which coined data downtime for periods when data is wrong, missing, or inaccurate, announced a $135M Series D in May 2022 at a $1.6 billion valuation, led by IVP. The premise is monitoring borrowed from software: detect when a table goes stale, a row count collapses, or a schema shifts.

Why it matters: the capability is real. Freshness, volume, and schema breaks are exactly the silent failures that poison a dashboard while every status light stays green. Monte Carlo's own framing is full visibility into the health of your data so you are first to know when it broke. That is detection. It is useful, and it is not the same thing as meaning.

What it reveals: observability answers did the number change. It does not answer what the number is supposed to be. You can monitor active customer with flawless freshness while two departments hold rival definitions of customer, and the platform faithfully reports that a figure nobody agreed on arrived on time. The money flowed to the symptom that is easy to instrument, not the cause that needs a decision.

What to watch: whether organizations pair monitoring with definitional ownership — a contract, a glossary entry, a named steward — or buy detection and treat semantics as solved. The reality tax is the cost of skipping that work: alerts that fire about a quantity no one can explain. Vendors found a category. Definitions are still waiting for one.