What is it?
Creative Mix Modeling ends the guesswork about what makes a creative perform. From your media history, it learns which creative and execution characteristics drive clicks, and turns that into two things media and design didn't have before: a clear list of what to prioritize and the ability to predict a piece's performance before launching it, with one click, without running research. In our models, creative tweaks guided by the right drivers raise CTR by 70% to 140% on average.
Do your media and design teams argue over which creative to bet on? CMM tells you what to prioritize and predicts each piece's CTR before it goes live.
Who it is for
Media, performance, creative, and design teams (and agencies) that run many creatives and need to decide what to prioritize and validate pieces with evidence before spending.
What we answer
Clarity on what to prioritize: CTR drivers ranked by impact
Predictive pre-test without research: predicted CTR per piece before airing
Media and design decisions grounded in evidence
Less budget spent "learning live" and more hits from the start
More clicks and conversion for the same media investment
Statistical CTR model (logistic regression) with the weight of each driver
Ranking of drivers by impact and by occurrence
Impact × occurrence matrix (where to act first)
Piece-by-piece CTR prediction before launch (no field research)
Best combination of attributes (with 95% CI) and estimated CTR
Model quality (R², MAPE) and actual-vs-predicted validation
Interactive performance simulator
Two superpowers media and design didn't have before: knowing what to prioritize and predicting performance before launch.
Instead of an endless list of best practices, you get CTR drivers ranked by impact: format, device, day, color, theme, presence of people, product photo, highlighted price, text size, message framing. Media and design start briefing and producing knowing where the gains are.
The model predicts the CTR (and funnel effect) of any creative + media combination before airing, with no fieldwork and no sample. You test the piece in the simulator, compare variations, and only run what has the highest chance of performing.
Formats: Quantitative modeling on your media data + creative labeling (computer vision/manual). Delivered as report + predictive dashboard/simulator. One-off or continuous (creative tracking). Once calibrated, pre-testing new pieces is instant.
Assemble the piece from creative and media attributes and see the predicted CTR before it runs. The model propagates the effect to conversions, ticket, CAC, and LTV. Illustrative values, calibrated per project with your data.
Predicted CTR
0,578%
+0,0% vs. baseline (0,578%)
▲ green = improvement vs. baseline. CAC shows green when it drops. Baseline = account average.
We turn data into decisions, with method and transparency.
Brands that trusted Okiar to answer their business questions


Media Optimization & ROI
Models (MMM, Lens CTR, VBB) to discover what drives your media results and optimize every dollar invested.
Explore the solutionEconometric Models
Econometric modeling that measures each channel's contribution and optimizes budget allocation, with attribution beyond last-click.
Explore the solutionProprietary Intelligence
Proprietary historical series on the Brazilian financial consumer, ready to answer without starting from scratch.
Explore the solutionWhat challenge can we help you with today?
Request your quote and transform your business based on data. With tailor-made solutions, our team will help you turn information into strategy.