One of the most important outcomes from Marketing Mix Modeling is the Return on Investment (ROI) for as a total for all media investments, but also separately for each channel. ROI is sometimes also referred to as ROMI (Return On Marketing Investments) or mROI (Marketing Return On Investment). ROI is a key metric for measuring media performance, and enables comparing financial performance of ad platforms, channels, and campaigns.

Most common approach is to express ROI of media as "Sales ROI", i.e. how many Euros/Dollars of sales did one Euro/Dollar invested in marketing bring back. Sales ROI is calculated as follows:

An alternative for Sales ROI is "Margin ROI", which takes into account the gross margin of the sold products. Sales ROI is simpler to calculate, because it doesn't require data or assumptions for gross margins, but the downside is that it doesn't take into the profitability of the advertised products.

Margin ROI, i.e. how many Euros/Dollars of margin (e.g. gross margin) did one Euro/Dollar invested in media bring back, can be calculated as follows

ROI-calculation is dependent on robust estimation of media-driven sales (or gross margin for Margin ROI). There are four main alternative methods for estimating media-driven sales:

1. Marketing Mix Modeling (MMM) is currently the only method that can estimate incremental sales impact of media for all advertising channels in a continuous manner. MMM platforms also provide you with optimization tools that help you improve media ROI with more optimised media budget allocation. Read more on Sellforte website and in this introduction to Marketing Mix Modeling work.

2. Incrementality tests or experiments are one-off studies that attempt to estimate incremental impact of media. Examples of experiments include conversion lift tests, geo tests, and shutdown tests. They are not practical to use in continuous ROI measurement, because they can be expensive and laboursome to implement, typically have wide confidence intervals, and are focused on one channel at a time. However, experiments are commonly used in Marketing Mix Model calibration.

3. Google Analytics 4 last-click, or other cross-channel attribution tool. Many performance marketing teams track the effectiveness of and conversions driven by digital media with Google Analytics 4 or other attribution tools. However, these attribution tools have significant biases and don't measure the true incrementality of digital media. As an example, last-click typically under-reports the effectiveness of Paid Social by a factor of 2-9x, as shown by this study analysing the effectiveness of Meta channels. Additionally, attribution tools don't capture the offline sales impact of digital media.

4. Ad platform attribution. Each ad platform has its own tools and approaches for estimating conversions driven by the ad platform. These conversion reports are typically inflated, because the attribution method used by ad platforms is not able to distinguish between a conversion that would have happened without media, and a conversion that was truly driven by the ad platform. The share of truly incremental conversions driven by the ad platform is low for retargeting campaigns to existing customers, and higher for prospecting campaigns targeting new customers.

ROAS (Return On Ad Spend) is a metric that can seem similar to ROI, but there is a clear difference.

ROAS is typically used when referring to effectiveness estimates originating from attribution, such as Ad platform attribution, Google Analytics 4 last-click or multi-touch attribution. ROAS is calculated by dividing attributed conversion value with advertising spend. When using ROAS, it is acknowledged that "conversion value" does not reflect the true incremental sales driven by the media. As an example, as platform conversions can include conversions that would have happened without media, and conversions driven by other media.

ROI is used when referring to effectiveness estimates originating from methods trying to capture the true incremental sales driven by media, such as Marketing Mix Modeling.

If you look at demo of Sellforte's product (https://demo.sellforte.com/), you can find ROI in multiple places.

Firstly, the "KPI"-tab (in the picture below) provides you with

Since many performance marketing teams are used to tracking attributed ROAS from GA4 and Ad platforms, we are including their ROAS metrics also in the Sellforte dashboard, so that they can be compared to MMM-reported ROI.

Secondly, the "Table"-tab shows the Sales ROI for each Ad platform:

In the "Chart"-tab, the user find visual representation of the Table-view.

By pressing "Show all charts"-button on the top-right corner in the same view, one can also find an even more comprehensive visual representations of the table.

"ROI Bubbles"-tab is one the most popular views in Sellforte for comparing media or campaigns. In this view,

ROI measurement can be conducted on multiple levels. Below are typical levels of granularity that you can find in Sellforte's Marketing Mix Modeling:

The ROI calculation in this article reveals the average ROI for a given time period. This is helpful when comparing historical performance across different activities. However, the ROI metric does not tell which activity provides the highest return if one more euro/dollar is invested in it. To analyse this question, one should become familiar with the concept of Marginal ROI.

ROI can vary over time, depending on various factors, such as seasonality. Additionally, ROI of marketing typically decreases if marketing investments are increased. This is called the Diminishing Return effect.