How they Differ, and How they Work Together
If you’re up-to-date on your alphabet soup, you may know that multi-touch attribution (MTA), media mix modeling (MMM) and incrementality work together to give you a full picture of your campaign performance. They all provide useful answers as you assess what’s working and what’s not — just to different types of questions.
And if you’re not as familiar with these terms, let’s talk about it. Read on for more information about MTA, MMM, and incrementality so you can understand how they work in concert to give you the information you need to make informed marketing decisions.
Multi-Touch Attribution (MTA)
MTA is major; we have an entire guide dedicated to it.
The TL;DR is this: MTA is a method of marketing attribution that accounts for all of the different touchpoints and activities in a customer’s journey — not just the first or last touch.
While first or last touch attribution gives all of the credit to a single touch on the customer journey, MTA understands that the customer’s path to a conversion isn’t linear, and aims to distribute credit (and attribution) across all relevant touchpoints.
MTA is your go-to tool for the day-to-day. It gives you attribution information on a granular level about campaign performance so you can make decisions on where to spend and adjust your budget on a given day in order to optimize for profitability.
What worked last week may not work any more; market and consumer variables are constantly shifting and making your ad performance fluctuate. MTA lets you stay on top of those shifts across all of your digital channels.
Checking out your MTA on a regular basis helps you keep your finger on the pulse so you can make regular adjustments and know what’s happening based on hourly or real-time performance data.
The cons of MTA? It’s primarily focused on digital touchpoints, so if you do a lot of traditional marketing, it may not capture attribution for all of your activities. That’s where MMM comes in.
You should use MTA if:
- You want to know the contribution of each touch point to a final conversion
- You want granular, real-time data
- You want a go-to resource to check daily to inform your spend decisions
Media Mix Modeling (MMM)
While MTA goes deep on individual touchpoints, Media Mix Modeling (MMM) goes deep on individual channels and tactics. MMM gives you a top-down look at how each channel is performing so you can fill in the gaps of your MTA reporting.
MMM goes beyond the day-to-day to incorporate all of your past and present data as well as seasonality, market trends, and more to inform your strategy and help you allocate resources.
Rather than a tool to help you make everyday decisions, MMM can be thought of as a powerful forecasting and budgeting tool in the longer term. It helps you understand the marginal efficiency of each channel, and where you might start to see diminishing returns.
Let’s say that one of your channels is performing at an 8x ROAS (return on ad spend). With just that data point, you might scale your ads indefinitely and pour more money into that channel. But proper MMM will tell you that if you spend more than a certain amount per day there, you’ll stop seeing returns.
Both pieces of data are needed to make the best decision possible: MMM puts integral parameters around your MTA insights.
A major pro of MMM: it can give you information about both digital and traditional channels, both online and offline efforts. Not only can you see how your ads and influencers are performing, but you’ll get data about your billboards and commercials as well.
Another pro: MMM shows you the limits of what’s possible with the resources you have available — regardless of what your boss says.
Marketing teams are often given budgets that are difficult to work with. With MMM, you can input your budget and see if your goals are possible based on historical channel performance so you can set realistic goals and make your case using data.
The cons? MMM tools are advanced, and you may need support to set up or interpret your outputs. Depending on your provider, they may also take longer to run — while Northbeam provides regular MMM reports, other providers can often only deliver these insights on a quarterly basis.
Still, MMM is powerful. The best MMM tools will give you suggestions on how to shift your budget to optimize for performance based on your goals. You can do multiple MMMs by regions or marketplaces, metrics, and more. You can rework the numbers to see how your channels are performing against new customer acquisition, total revenue, or any other metric you want to measure. MMM is a flexible and dynamic way to understand your marketing attribution.
You should use MMM if:
- You want to understand when you’ll start seeing diminishing returns and/or room to scale in a given channel
- You want to establish and track a long-term strategy that involves a variety of different marketing channels
- You want to capture offline channel performance
Incrementality
Incrementality can be thought of as a calibration tool — a sanity check. It provides a more controlled environment in which you can test and control for different variables, like spend, that may affect how your campaigns or channels are performing.
Here’s an example: let’s say that you run a subway ad and receive 5,000 new users that week. But what if 50% of those users would have signed up anyway, regardless of whether or not they saw the ad?
Incrementality can help you answer the question of organic cannibalization by better controlling for variables and measuring the actual impact of each activity on your core metrics.
While attribution matches two events, like a touch and a conversion, incrementality quantifies the relationship: what is that relationship actually worth?
But measuring incrementality can be time- and resource-intensive to do consistently. It involves a test group and a control group, and a variable you want to test for, like geography. When you compare the performance of both groups, you’ll be able to quantify the incremental lift provided by each campaign.
Because they capture a moment in time, incrementality tests get outdated quite quickly. The tests you ran in January will yield results in February or March, and by then, they may already be outdated.
And because incrementality involves a test group, you will have to turn off marketing for a subset of your audience. This could result in lower sales and marketing performance — maybe that’s worth it to gain a deeper level of understanding, but maybe it isn’t. Keep in mind that a channel has to be large enough to begin with to garner statistically significant results, so this decrease in marketing will not be insignificant.
When running an incrementality test, you also need to be hands off and let things run for weeks or months at a time. That means very little room for changes, tweaks, and adapting on the fly if you want to get clear and accurate results.
It’s important to note that MTA and MMM, when used together, can provide a similar or much higher level of detail than incrementality without the hands-off requirement. MTA can show you all the different touches that happened on the way to a conversion, and MMM can add a layer of strategy and insight about where you should funnel budget to see incremental returns.
Bonus: incrementality results can be incorporated into your MMM model; Northbeam offers this service to train your MMM on your specific and unique data set.
You should use incrementality if:
- You’re worried that your paid channels are cannibalizing your organic growth
- You want to improve your MMM tool by feeding it incrementality test data
- Your marketing team is not in a time-sensitive or resource-sensitive phase of growth
What do we suggest?
The best marketing strategies create room for all three of the tools above at different stages of a company’s growth and maturity. The data that each tool provides is different in key ways, and they each have their own distinct pros and cons, so you can best assess which makes the most sense for your current marketing priorities.
%25201.png)
.png)
.jpeg)
.png)
.png)
.png)
.svg)
