One of the key strengths of Recast’s model is that it operates within a fully Bayesian framework. Unlike traditional models, a Bayesian approach requires explicit assumptions from the outset. While all models rely on assumptions, Bayesian models stand out by allowing those assumptions to directly align with the knowledge and context that marketers already have.

The Purpose of Model Configuration

The purpose of the Model Configuration call is to capture these assumptions from your
team and translate them into inputs the model can work with. Here are a few important
things to keep in mind:

• This is YOUR model: We want the assumptions to reflect your unique understanding of how your business operates. It’s not about fitting your business into a model, it’s about helping the model fit your business.

• Confidence and uncertainty matter: For every value we ask for, we’ll also explore your level of
  certainty in the value you provide and can reflect that certainty in the inputs to the model. One of
  the model’s primary goals is to take the uncertainty you currently have and reduce it to improve
  your decision-making.

• Built-in validation: Our model includes checks to alert us if any assumptions don’t hold or are
  having an outsized impact on the results, ensuring that the outputs remain reliable and reflect
  reality.

Preparing for Model Configuration

A common question we hear from new Recast users is: “How can I best prepare for Model Configuration?” While it might seem counterintuitive, our usual advice is actually to avoid over-preparing, and here’s why:

Trust your instincts and keep an open mind: You know your business better than anyone else, and it’s important to approach the process without being overly influenced by other data sources, which often have their own limitations.

Avoid building on previous models: We don’t want the Recast model to be constrained by assumptions or data from other attribution models. Instead, we aim to create a fresh perspective tailored to your specific needs.

Incorporating experimental lift test results: The most critical data point, the results of experimental lift tests, will be directly incorporated into the model. Recast should already have these from the data collection stage and those will be factored directly into the model to help keep everything aligned!

You may not know, and that is ok!: While the Recast model is the most rigorous MMM on the market, the model configuration stage does not need to be an exact science. All of the rigor we put into building Recast is to help identify any issues with your assumptions! You may not have a strong perspective on a certain channel, and this is not a blocker! We set the model up to reflect your understanding of your business, and where you are less certain, the model will also start with less certainty.

What should you do to prepare?

Gather the Right Team Identify Other Information Sources Preparation Assemble a core group of 3- 5 decision makers to bring to the call Pinpoint relevant data sources such as site tracking, platform data and internal heuristics to kickstart discussions Review the Model Configuration Breakdown in advance, and ask Recast any questions before the call to streamline the meeting.

Model Configuration Breakdown

There are 3 main components to model configuration, below is a bit about each of them so you can use these to inform who you need on the call as well as what you need to bring:

1. Channel-Specific Details

Objective: Understand the nuances of each marketing channel.

Key Questions:

• What is the channel’s role in the customer journey?

• How scalable is this channel?

• Do other marketing channels drive traffic to this channel?

• How does the timing of recorded spend align with consumer exposure to the ads?

2. Time Shifts

Objective: Estimate how long it takes for marketing spend to impact the modeled key performance indicators (KPIs).

Key Scenarios:

• Time to convert when a customer is just beginning to explore the category.

• Time to convert when a customer knows the options and is researching.

• Time to convert when a customer is ready to purchase but needs final confirmation or logistics.

3. Contribution by Business Driver

Objective: Determine how much business is driven by marketing versus other factors.

Factors to Consider:

• Brand equity

• Barriers to entry

• Promotions (both spend-related and non-spend related) Model Configuration

Open

There are a number of steps Recast performs in order to ensure that your model is performant. Many of the steps are internal, but there are a few steps that will require your input:

1. Compiling the dataset

2. KPI Selection

3. Model Configuration Call: this is a conversation about your business where we’ll set inputs for the model. We’ll discuss how the marketing program is currently run and the team’s expectations around channel-level performance (priors), and the size of organic (or non-marketing-driven results, also called “Intercept” in Recast terminology)

4. Recast Model Checks

5. Benchmark Model Review: we'll review your initial results in the Recast platform and revise results based on your feedback

🔢 Compiling the dataset

See 📊 Data Guide for more information on compiling your data and sharing it with the Recast team.

KPI Selection

Your KPI is your outcome variable that Recast will model. This could include sales, acquisitions, conversions etc. depending on the nature of your business. The KPI is central to the model.

Select your KPI based on how you are measuring your marketing results. When you spend on your marketing, what metric are you tracking to measure effectiveness? This will be your KPI.

If you have multiple KPIs you want to model, reach out to your CSM or support@getrecast.com and we will talk through options for building multiple models.