The Coca-Cola example on SaliencyLab is not there to claim a performance uplift. It is there to show how a team should read a decision-ready report: the core output of pre-spend creative intelligence, a scored, benchmarked verdict on a creative before any budget is committed.
What this example is, and is not
It is an authored walkthrough using a public creative. Its scores, quotes and timings illustrate report structure; they are not a stored live analysis. It is not a client engagement, not a case study, and not evidence of a business result.
We are deliberate about this because the alternative is common and misleading. A fabricated uplift number attached to a recognisable brand reads as proof while proving nothing. The structure of the report is the thing worth teaching.
What the walkthrough is designed to teach
The example report is meant to show four things in one place:
- the asset
- the KPI pattern
- the benchmark and confidence context
- the next action the team should take
That structure matters more than any invented case-study number ever could.
What to look at first
Read the KPI pattern, not just the headline score
The most useful signal is usually the spread between the strongest and weakest KPI. That spread tells you whether the ad is consistently strong or whether one issue is dragging the rest of the performance down.
This is also what the verdict rules encode. Scale requires a composite of 70 or above and no KPI below 55, so a wide spread can hold a creative in Sharpen even when the average looks healthy.
Look for the decision cue
A report should help a team scale, adjust, pause, or replace. If it cannot do that, it is still reporting rather than decision support.
Read the benchmark label with the report, not after it
Benchmark context becomes useful when it changes how urgent the next action feels. Check the sample size alongside it, since a percentile from a thin category slice is orientation rather than a verdict. How to read benchmark context covers this habit in more depth.
What makes the example valuable
- It shows how evidence connects to the score
- It shows how a report can be read in a live stakeholder conversation
- It shows how a diagnostic becomes a concrete next action
How to use this example internally
Use the example report as a review aid when onboarding new teammates or clients. It helps people understand the reporting language before they upload their own work.
For the general skill this example demonstrates, how to read a RoastIQ report walks through the same reading order on any report, not just this one.
The right expectation
RoastIQ’s five KPIs are diagnostic model outputs. The separately reported platform-native attribute model has its own outcome-validation record; its correlations must not be attributed to the five-KPI composite, attention heatmaps or BuyerLens. See the dated methods and validation scope. None of these outputs establishes sales, ROAS or a measured human response to your ad.
Walk through the current example
- Read the identity first. The public page describes an 8.4-second Meta cut. It is different from other Coca-Cola versions that may appear elsewhere.
- Read the proposed action. Keep the opening atmosphere; consider an earlier readable pack cue. This is an example recommendation, not a completed human review of a client report.
- Inspect the source status. No verified outcome cohort is attached. Do not turn the displayed example into a benchmark or campaign claim.
The numbered reading steps above are annotations to the actual public interface. Open the full example to read the text at its original size. A real client deliverable needs its own run, creative version and review record.