Marketing mix modeling vs. Multi-touch attribution
Welcome to this thrilling boxing match between the giants of B2B marketing analytics!
Tonight we’ll finally know which one is king of the ring when it comes to analysing and measuring the effectiveness of your marketing efforts.
So, which one will it be?
Marketing Mix Modelling (MMM) - the math-heavy, channel-focused, top-down approach for analysing your B2B efforts.
Or Multi-Touch Attribution (MTA) - the granular, journey-focused, bottom-up approach for optimising your B2B marketing efforts?
Overwhelmed by data from an ever-growing go-to-market tech stack, it's hard for B2B marketers to navigate the analytics landscape.
This complicates the task of assessing what content is reaching your audience, which ads are resonating with them, and what marketing channels are driving revenue.
Luckily there are some tried and tested solutions to help overcome this analytics challenge.
In this blog post, we’ll be busting up the two main methodologies to see which one comes out on top for measuring the performance of B2B marketing endeavours.
The main methods in B2B marketing analytics
As any experienced marketer will know, making data-driven decisions is the key to success.
Buyers are becoming more and more selective when it comes to choosing products.
In a world where there’s so much noise from an endless supply of products, knowing which campaigns drive leads or what touchpoints are the most important in catching the attention of your audience is the key to success.
I mean, who wouldn’t want to know what channels, campaigns, and content are actually driving growth?
But to get there we have to make sense of the vast ocean of data accumulated by the modern B2B go-to-market team.
So how do we do this?
Well, by running your data through one of the available marketing analytics methods, obviously!
And there are many methods out there - the options are almost endless.
Customer Segmentation
Customer Profiling, Lead Scoring
Data Analytics and Reporting
Single-touch attribution
Multi-touch attribution
Customer Lifetime Value (CLV) Analysis
Predictive Analytics
Competitor Analysis
Marketing Mix Modeling
Customer Surveys and Feedback Analysis
ROI Analysis
But out of these, there are two that have come out as the most popular and powerful options in the B2B go-to-market analytics space.
Marketing Mix Modelling and Multi-touch attribution.
What is Marketing Mixed Modelling?
So what is Marketing Mix Modelling?
As we have already established, MMM is an analysis method for marketing data.
It measures the impact of marketing campaigns to determine how various elements contribute to conversions and ROI.
And by also comparing this to what the baseline would like without the use of any marketing efforts, MMM is generally considered to capture the incremental effects of marketing better.
To do so, MMM needs aggregated data on all the things that could be driving outcomes.
The data can be derived from a multitude of channels, both traditional and digital; internal and external. Channels such as social media, podcasts, television, magazines but also out-of-home advertising (i.e. billboards)
It even takes into account variables such as seasonality and promotions.
From this, it is able to give a variety of insights based on a wide range of channels enabling the marketer to make data-driven decisions.
But how does it work!?
So by employing either linear or non-linear regression between both dependent variables (sales or engagements) and independent variables (ad spend on different channels), MMM can measure the overall influence of marketing efforts on final sales.
Now, for this to work properly, you really need to carefully select and prioritise the data you want to measure.
That means: No outdated data, no inaccurate data, and no incomplete data sets.
You really need quality data for MMM to work.
Now, the amount of data that’s normally needed for MMM to function ranges from anywhere between two to three years of historical data.
However, once this data has been aggregated, thoroughly cleaned and modelled, it enables marketers to measure the effectiveness of their channels, allowing them to access future ROI, spending allocation, and sale forecasts on channel level.
This data-driven approach serves as a potent tool in the marketer's arsenal, empowering them to achieve their ultimate objectives!
What is Multi-touch attribution?
Now, what is multi-touch attribution?
Rather than reviewing marketing activities by looking at aggregated data from a high-level view, multi-touch attribution looks at individual customer journeys and takes a bottom-up approach.
MTA uses all the recordable touchpoints to form customer journeys. From the first anonymous visitor to the paying customer.
By assigning value to each touchpoint in the customer journey, MTA offers a more granular view of the data behind it, ensuring that you are able to link revenue to the specific channel or marketing effort.
What seemed to be a set of dots randomly sprinkled out between the first and the last touch of your customer’s journey now lies in an ordered path, with all the gaps bridged together.
No more shooting in the dark.
Instead, you can now easily connect touchpoints directly to pipeline conversions and ultimately revenue contribution.
Now, when it comes to assessing performance, MTA uses a variety of different models.
Single-touch attribution models are used to pin critical value to single touchpoints - either by giving credit to the first touch or the last touch by being the significant impact giver.
On the other hand, multi-touch models distribute value across multiple touchpoints. For instance, Linear attribution distributes value equally amongst the touchpoints. U- and W-shaped models, distribute value equally between the first and second touchpoint (U-shaped) and between the first, middle, and last touchpoints (W-shaped).
Lastly, we have the data-driven attribution which uses statistical algorithms and machine learning to analyse historical data, determining the relative importance of each touchpoint in driving conversion.
A B2B marketer selects the model which best meets their unique business need - although most are now gravitating towards data-driven and letting the data do the talking.
Of course, describing the data modelling behind each and every MTA model out there would probably take a whole series of blog posts, so you’re welcome to check out either this one or this one if you’re interested
If you’re curious as to whether multi-touch attribution might be the correct fit for your company you should read this article.
MMM versus MTA: What are the pros and cons for the B2B marketing team?
No more beating around the bush. It’s finally time for the showdown!
Will it be MMM, or will it be MTA?
While they’re both ultimately tackling the same issue - analysing marketing data - they approach the problem differently, and as such should be viewed less as competitors than complementary methods.
On one hand, MMM provides the user with a top-down perspective of their marketing efforts enabling them to see how different channels contribute to revenue.
MTA, on the other hand, provides the user with a bottom-up, customer journey-focused approach. Instead of evaluating aggregated data, MTA focuses on the individual touchpoints of the customer journey, accessing the value that each one gives to the final outcome.
But that is just scratching the surface.
So let’s take a closer look at the pros and cons of each method!
Strengths and weaknesses
Both methods feature different strengths and weaknesses when it comes to their take on analyzing marketing efforts.
Marketing Mized Modeling (MMM)
Pros:
Handles things that you cannot "track" or directly tie to a user or company. Like views on Facebook, TV ads, etc.
Inbuilt incrementality (Extremely dependent on accurate data usage - missing drivers will overestimate incremental effects)
Business-wide optimizations is a more direct outcome
Cons:
Lacks granularity i.e. you are likely stuck at channel level
You don't have a customer journey as it operates on aggregated data
Low update frequency (monthly but normally quarterly)
Hard time capturing long-term effects ie. revenue that happens a long time after the action
Very data dependent
Math heavy
Multi-touch attribution (MTA)
Pros:
Granularity - allowing tactical optimisations eg. what campaign performs best.
Customer journey insights
Easier to draw conclusions from
It’s time efficient
High update frequency
Same data needed as you need for personalisation
Cons:
Some channels don’t allow you to track the actions of single users or companies (like TV or Radio ads).
Incrementality isn't a direct output; it requires separate measurement, but it offers a more accurate assessment of incrementality with correct experimentation methods
Lacks direct business-wide optimisation outputs.
Requires customer or company-level data to create its customer journeys.
Final round!
So in summary, we can now say that, while MMM does excel in both connecting offline and online conversions and incorporating incrementality, it works well with a range of data due to its versatility.
By including things such TV-adds, phone calls or PPCs through signing off on trial offers, you can argue that MMM gives you a wider range of things to include in the analysis.
But for these to work, MMM demands an enormous amount of quality data. It can’t provide the user with granularity making it difficult to effectively change certain aspects of a given campaign, it's extremely math-heavy and it struggles with giving insights into long-term effects.
On the other side, MTA stands out with its granular insights, tactical advantage, and clear view of the customer journey.
With MTA you know exactly what that company did at that exact moment on their customer journey. Allowing you to personalise and optimise the individual customer journey.
On top of that, MTA is just simpler to work with. It’s not as time-consuming as MMM and the update frequency is higher, giving you a more actionable solution with shorter feedback loops.
Yet it still faces challenges tracking offline conversions, and not being able to provide the whole picture. It requires the implementation of incrementality, and it lacks a universal optimization solution.
Both methods show strengths and weaknesses, so which one is actually the best when it comes to a B2B SaaS context?
Let’s see the judges’ verdict, shall we?
Applying MTA in a B2B SaaS context
It’s a technical knockout!
B2Bs are drowning in data. And often enough their go-to-market tech stack is both siloed and disjointed.
This leaves marketers with a vast amount of data. And yet, they are still lacking the necessary insights needed for them to take action!
But who can blame them? The B2B SaaS customer journey is convoluted and complex.
Marketers want to know what works. They want to know which channels, campaigns and content are actually driving revenue growth, especially performance marketers.
And multi-touch attribution can provide exactly this.
By providing granularity, you not only gain insights into which strategies are working, but you also know exactly how to amplify them!
'
This detailed information is lost when you aggregate the data with MMM and so is only provided by the MTA approach.
As a paid marketer, granularity provides you with the necessary details to either optimise the individual marketing channels, appreciate the importance of niche channels, or understand customers' journeys.
Of course, there’s MTA’s Achilles heel - offline tracking. As we’ve already described, MTA cannot easily track offline touches, although it is still possible (check out how you can include offline touches in Dreamdata’s ROI reporting).
This however isn’t a huge stumbling block, as most B2B SaaS companies are mainly operating on digital platforms, and aren’t as dependent on say TV ads as B2Cs might be.
Besides, B2B companies are also very dependent on individual account tracking, as a lot of communication happens 1:1 via sales or customer success.
This makes MTA a more natural fit.
But wait, there’s more!
As the customer journey for B2Bs often reaches an average of 192 days, the need for MTA becomes even more apparent.
For its part, MMM has trouble capturing effects when there is a long time between exposure and revenue.
The requirement to handle long sales cycles is again making MTA the more natural method for B2B SaaS companies.
You essentially know, what to scale, how to scale, and when to scale. Increasing your return on investment on every marketing effort you make. Ensuring you get exactly what's coming to you.
By applying multi-touch attribution platforms, you’ll know everything. From the first minute action all the way up to your entire marketing spend.
All in all, multi-touch attribution platforms have developed a lot and have become more versatile.
So when it comes to a B2B SaaS market, MTA is the best fit.
Start your Multi-touch attribution journey today with Dreamdata!
