Creating a Model

Lean how you can create your first Marketing Mix Model!

Apoorva Wate avatar
Written by Apoorva Wate
Updated over a week ago

To begin creating a model:

  1. Click on the “Create Model” button on the MMM page.

  2. You can create a model either by uploading your MMM data in a CSV or by using integrated data in a spreadsheet.

  3. To upload your custom dataset, select the option labeled "Upload your data." Either click on "click to upload" or drag and drop the file into the designated area.

  4. Name the model and click "Next" to proceed.

Guidelines for CSV file upload:

  • Column headers should only contain alphanumeric characters and underscores.

  • The heading should start with an alphabet only.

  • The date format should be YYYY-MM-DD.

  • In the case of weekly input, all dates should be Sundays or Mondays.

Schema Mapping

If you have uploaded a custom MMM dataset, map the schema by associating your input data with Lifesight's data types. If you are using integrated data, this step is unnecessary.

  1. Select the column containing the date and map it to the type of data. Lifesight allows you to input data in daily, weekly, and monthly formats.

  2. Select the column containing the KPI you want to measure and map it to the metric intended to serve as your business outcome data (KPI).

  3. Define the marketing spending, organic impressions, and contextual variables. You can also add impressions and clicks from paid and organic campaigns.

💡Increasing the number of variables may enhance the model's accuracy.

Configure your Model

  • Enable Pre-configured Contextual Variables, like Holidays and Seasons, that may affect your business outcome

  • Refresh: To update the model with the latest data regularly, you can configure the refresh frequency. The refresh frequency can be selected from 7, 15, 30 and 90 days

  • Advanced settings (Hyperparameters)
    Adstock - Adstock refers to the prolonged or lagged effect of advertising on consumer behavior. Simply put, it’s the impact of an advertisement that extends beyond its initial run. It allows marketers to account for the delayed and sustained impact of advertising campaigns on sales.

    Saturation Curve -The saturation curve shows how the incremental impact of a marketing effort decreases as the level of that effort increases. Saturation points represent the point at which additional investment in a particular marketing activity, such as advertising or promotions, will no longer result in a proportional increase in sales or revenue. Simply put, at a certain point, all marketing investments reach a saturation stage, making them unprofitable. The shape of the curve can vary depending on factors such as the industry, product, or market conditions.

    Changing the different values of shape parameters and inflection point parameters changes the shape and scale of each of the above curves. These variables are numbers where the user defines lower and upper ranges for each of the variables. Users would also be able to see the curves side by side of the values they are passing. You can input the value based on the results of your previously completed experiments.

  • Calibrate: You can calibrate the model using the results from incrementality experiments, lift studies, or post-purchase surveys conducted on any advertising platform. Ensure to add separate and multiple calibration inputs for each platform as needed.

Click on ‘Create Model’ to proceed. Once the model training is completed, the status changes from “Training in Progress” to “Success”. It takes about 30 minutes to an hour, in general, for the platform to prepare the model. You will also see a confidence level, which indicates model accuracy from 0 to 100% once the model is generated.

If you have any further queries, please write to us at [email protected] and we'll get back to you at the earliest.

Did this answer your question?