Methods
How a number
gets onto the page.
Three things make Precise different: contribution math, depth and speed, and traceable outputs. This page shows the method behind each, using the same sample analytics as How it works.
This page describes method, not results. Every figure is a synthetic example.
01 / Contribution math
Contribution math.
Multi-patented methodology calculates contribution value relative to cost. Scores individual attributes and their combined effects.
An impression carries many attributes at once: the creative, the audience segment, the publisher, the supply path, the platform, the daypart, the frequency. Each can contribute value on its own and change value in the company of the others.
Contribution value is the share of the outcome an attribute can fairly claim, alone and in combination, against the cost it added. The allocation is a Shapley-style fair share: each attribute is credited with what the outcome would have lost without it, averaged over every combination it appears in.
The unit is the same everywhere: estimated revenue contribution per $1 of media spend and fees, so a publisher and a data layer can be read on one scale.
Where two attributes reached the same people, their contributions are not summed. The overlap is named and, where only platform reports exist, the number is labeled gross.
One creative. Each added layer of context.Illustrative
Cumulative contribution value as context is added; each row includes the rows above it. The same example as the drilldown on How it works.
02 / Depth + speed
Depth + speed.
Evaluates hundreds of attributes, individually and in combination, in near real time. Recalculates contribution as new campaign data arrives for immediate recursive learning.
Precise works at the grain each platform reports. Where impressions are visible, the unit is the impression. Where they are hidden, as on social and search, it is the ad set, campaign, keyword group or day. Each figure says which grain it stands on.
The cycle follows the data. Delivery and price are checked the same day; outcomes are scored as their windows close, daily or weekly. Each recalculation feeds the next recommendation.
A value is not yet a decision. Each combination can absorb only so much spending before the next dollar returns less; the response curve says how much. Precise reads the curve for the cell that would give and the cell that would receive, and prices the move: how much, and what is predicted at the horizon.
Every predicted figure carries a range. The range comes from scored history: moves of the same kind whose outcomes have already been measured. Fewer scored moves, wider range. Where there is not enough evidence, that is the answer, said in words.
Each combination absorbs only so much.Illustrative
The two cells that move the most in the $1M recommendation on How it works. Money leaves the flat curve and arrives on the steep one until the next dollar returns the same on both.
Scored history for one kind of move: what a predicted range stands on.Illustrative
Each point is one scored move: the gain in return per $1 at the 7-day horizon against its holdout. Synthetic fixture, n = 14.
03 / Traceable outputs
Traceable outputs.
Preserves the data and logic behind every result. Produces structured decision records for agents to act on and people to audit.
A recommendation is written down with what was known, what was chosen, what was predicted with its range, and, later, what happened and what it taught.
Sent, reported and observed are three separate facts. Scoring happens at the agreed horizon against a holdout that kept the old setting, so the next range stands on more evidence than the last.
The same record is readable by people and by agents. The data and logic behind every figure travel with it, so a result can be re-checked later by someone who was not in the room.
One decision, on the record.Illustrative
The lifecycle and the five fields of one recommendation, from the same synthetic campaign as the chart below.
Our recommendation: predicted and measured returns
| morning | actual | predicted | low | high | left alone | holdout |
|---|
Illustrative schematic. The area to the right of today shows projections, not observed results. Return means revenue per $1 of media spend and fees.
Getting started