To make the most of your MMM+ kickoff call and take the pressure off of the ingestion process, it's crucial to gather and organize the necessary data inputs.
This comprehensive guide will outline the key elements you should prepare ahead of our kickoff call. We can use at least 5-months of both online and offline data, but it is highly suggested to provide at least two years of data for more accuracy and effectiveness.
All of your data should be organized in the data input template provided to you via email ahead of your kickoff call. If you need further assistance accessing your company’s unique input template, please reach out to our Customer Success team by using this form to submit an inquiry.
Data Input Checklist
- Historical Data - to run the model, all historical spend and revenue is required. To adjust for seasonality and give an accurate report, two years of data is highly suggested. If connected to Northbeam’s MTA tool already, all connected platforms will automatically be taken into account. If Shopify is connected, all revenue will also load into the model, however if using the Server API, please ensure that two years of historical orders have been upserted as well. Your primary KPI will be the modeling output and adding in the count of emails and SMS sends per day gives the model another data point to trend against if there are spikes in the output.
-
- Backfilling custom spend
- Backfilling orders, if needed
- If connected to Shopify, all set
- Primary KPI (revenue or transactions)
- Emails and SMS sent by day (if applicable)
- Calendar of Events - Calendar of events data helps the model identify patterns that spend alone cannot identify. For example, your revenue may spike for a few days when you launch a new product. These touchpoints paint the picture of your annual revenue and spend trends, allowing the model to visualize why revenue increased/decreased at certain times and give a more valuable output. Things such as company events, product launches, company-wide price changes, and promotions (with discounts included) are all examples of helpful events that train the model.
-
- Company events
- Product launches
- Price changes
- Promotions and full promotional calendar with discounts
- Budget and Constraints - For each platform (or platform_category), we ask that you set a lower and upper bound of daily spend. This helps keep the model within realistic scenarios. Having set lower and upper daily spend boundaries for each platform / category allows for more precise suggestions, and gives the model a realistic landing point for recommendations.
-
- For Budget, breakdown defaults to platform-category (i.e. Facebook prospecting, Facebook retargeting, etc.)
- For constraints, make sure there is an upper and lower bound of what your company is comfortable spending. if you have highly fluid budgets and 0 constraints, please use same number for both.
- Other Marketplaces - We can customize the model to your specifications including adding in other marketplaces. If you want to include something like Walmart or Amazon in your model, it can either be separated out into different models or factored into one holistic output.
-
- Will need the data, but Ebay, Walmart, CVS/Walgreens, etc. all can be modeled
- Do you want to see Amazon included? - For example, most customers model blended revenue (including amazon), but we can make a model just for amazon revenue if that's important to your business.
- Intangibles - Understanding the non-financial sides of your business helps our Customer Success team best tailor the outputs to your business needs. For example, if you are in growth mode and have a more aggressive spend strategy, our team will be able to add these factors into the model to give better recommendations. Understanding any macroeconomic impacts on the business (i.e. inventory constraints, seasonal sales cycle) help the team give improved advice on how to proceed with your marketing mix.
-
- Business and budget goals
- Familiarity with MMM
- Desired use case for MMM
- Any macroeconomic factors affecting the company (seasonality of business, inventory constraints, growth goals)
- List of holidays important to your business
- Granularity and detail are your friends!
Other Important Notes
- Data needs to be by day, and at the latest every week so the model is most up to date and accurate. Parsing data by day also gives the best return and allows the model to run efficiently as opposed to having to translate and prorate from a weekly or monthly basis.
- If using Northbeam MTA, all connected revenue, spend, and metrics will be factored into the model (this includes data currently in the custom spreadsheet) - Please make sure the customer spend spreadsheet is up to date (podcast, direct mail, etc.). Any data you might see or provide with Northbeam MTA will not need to be provided for MMM+.
- Include the “Halo effect” of product launches to understand how a new product or marketing effort changes the revenue output is another data point useful to the model