Mevo:
The SaaS platform for planning, measurement, reporting and insights.
PHASE ONE
Data Collection & Ingestion
Predict user-level journeys using three sources of aggregate and/or person-level data.
CONNECT AND INGEST MEDIA DATA AND KPIs
We work with you to securely bring together all your marketing data, without the need for site tags, and define target KPIs to track against.
VALIDATE, CLEANSE AND NORMALISE
The 3 data types; marketing data, demographic data and KPI data are automatically validated, cleansed and normalised.
PHASE TWO
Modelling & Measurement
Next, two unique AI algorithms are applied to your data to complete the picture.
EXPOSURE AI
Exposure AI predicts the missing exposures. In other words, it fills in the gaps to understand what users saw (exposure data) and who they are (demographic data).
MEASUREMENT AI
Once we have user-level journeys, we measure the impact of each marketing tactic to show how exposures drove the consumer decision.
PHASES THREE
Analyse & Optimise
Finally, the Optimisation AI combines the results of the Measurement AI and Exposure AI algorithms to simulate and optimise you media budget.
CUSTOMISE
Customise and configure media budgets at tactical and strategic levels.
SIMULATE
The Exposure AI algorithm gives unique insights, factoring in customer reach, frequency, diminishing returns and media interactions.
OPTIMISE
Proprietary Optimisation Engine processes hundreds of thousands of placements in minutes.
DECIDE
Intuitive UI surfaces actionable recommendations across all media types and tactics so you can plan with confidence.
Mevo Approach Summary
Mevo's approach and utility is for media measurement, optimisation and scenario planning (forecasting prescriptive media optimisation) recommendations at the most atomic (granular) level depending upon customer data availability. External factors beyond media incremental impact on KPI (e.g. sales, bookings, web activity, brand love, etc.) are accounted for in the baseline. Data sets used for Mevo's approach are sufficiently large (millions of rows) that parameters of inference are significant at >95% even after adjusting critical values.
The Mevo Platform is inherently a testing and experimentation platform with Generative Attribution. All model training is based on the latest media, spend, impression, and conversion behavior, and custom labels enable breakout by campaign or testing deployment. An advertiser merely needs to label or otherwise track treatment vs control to be able to view experimental results in Mevo. Experiments require only 1 week of data, and as little as £10,000 of spend. Testing results are used in every optimisation in the 'Scenario Planner' application, where advertisers can compare baseline to optimization results. Experimental results are also available in the 'Media Reporting' application. Bayesian updating is the foundational aspect of Mevo's model training to incorporate all previous testing results into each model for measurement.