The reasoning layer.
Not another dashboard.

LensOS is the intelligence engine inside MeasureLens. It sits above your existing data sources, evaluates each one, resolves conflicts between them, and produces a single answer with a confidence score attached.

In active development

What LensOS is

Most tools show you data. LensOS interprets it.

The fundamental problem with modern marketing measurement is not a lack of data. Teams have access to more performance data than ever. The problem is having no principled framework for evaluating which data deserves trust.

LensOS is being built to fill that gap. It is not a new attribution model. It is not a better dashboard. It is a reasoning layer that sits above your existing data, assesses the quality of each source, and synthesizes a defensible answer from the evidence available.

Think of it as the operating system for marketing measurement decisions. It does not replace your MMM or your attribution platform. It interprets them, weighs them against each other, and tells you what to do with the result.

Data sources LensOS evaluates

Meta AdsLast-click and view-through attribution
Google AdsSearch, display, and performance max
GA4Cross-device journey data
MMM OutputRegression-based channel attribution
MTA FeedsMulti-touch attribution models
ExperimentsGeo holdout and incrementality tests

Core capabilities

Evidence scoring

Not all data sources are equally reliable. LensOS assigns each source an evidence tier based on its methodology, from geo holdout experiments at the top to platform last-click attribution at the bottom. A holdout study earns more weight than an analytics dashboard. The hierarchy is explicit, not hidden.

Conflict surfacing

When two sources report different things, LensOS does not average them. It surfaces the disagreement explicitly, assigns a status flag, and explains the most likely reason for the delta. Attribution overlap, window differences, cross-device gaps, and model methodology differences each produce recognizable conflict signatures.

Confidence-weighted synthesis

After evaluating and reconciling sources, LensOS computes a Measurement Confidence Score between 0 and 100. This score reflects the agreement between sources, the evidence quality of each, and the recency of the data. A high-confidence result means multiple strong sources agree. A low-confidence result means you need more data before deciding.

Decision synthesis

The final output is not a chart. It is a recommendation: scale this channel, cut that one, or run a holdout test before committing. Each recommendation includes the confidence level behind it and what evidence would change the answer. It is designed to be useful in a meeting, not just in a spreadsheet.

Why "OS" is the right framing

An operating system does not do your work for you. It manages resources, arbitrates between processes, and provides a stable foundation for everything running on top of it. LensOS does the same thing for measurement.

Your MMM, your attribution tools, your incrementality experiments: these are all processes running in parallel. They each produce an output. They sometimes agree. They often disagree. LensOS arbitrates between them using explicit rules about evidence quality, then provides a stable answer that the rest of your decision-making can run on top of.

Without that layer, you are back to picking numbers by feel. With it, you have a defensible position, a confidence level, and a clear action to take.