There are two fundamental variables that marketers need to identify prior to embarking on the journey, or rather odyssey, of solving for multichannel attribution. Even more importantly, there is one critical first step necessary to succeeding at the effort.

Before we dig into these two variables and critical first step, let’s first clarify the phrase ‘multichannel attribution’. We know the consumer is influenced by numerous touch points across paid, earned and owned media on their way to conversion, and those touch points occur across a myriad of digital devices and physical experiences.  We also know that each touch point has some level of value and cost to marketing efforts, and therefore has some role in your ROI arithmetic. Multichannel attribution is the effort to assign a monetary value to the investment costs and revenue returns of those touch points.

Here’s the tricky part: Several of these touch points are not observable with standard issue reporting platforms, and so marketers must either try to infer these interactions with inferior tools, or invest in superior tools to make more accurately informed inferences.  It’s analogous to how scientists started with a concept called ‘Dark Matter’ so they could make their equations about the expanding universe net out, but then had to go and build a 17 mile underground tunnel in Switzerland to observe the particles they had theorised in their equations – not exactly a walk in the park to execute.  
Thankfully, we marketers don’t have to solve for equations that Einstein invented.  We just have to invest in more of what works and less of what doesn’t in order to hit our ROI targets within our budgets. Which leads us to the fundamental variables…

Cost of multichannel attribution

Multichannel attribution comes in many shapes and sizes.  There are high end technologies that only the top 1% of budgets can afford.  There are also increasingly mainstream multichannel attribution platforms like Convertro, Adometry and C3 Metrics that cost about as much as buying a car per month. I use the ‘car benchmark’ because car costs range between high mileage, high risk vehicles you’ll find in the back of Auto Trader, to high end luxury vehicles that sponsor major golf tournaments. Depending on your platform selection, media volumes and range of channels you seek to track, the cost of a pure play multichannel attribution technology can vary similarly to my deliberately vague ‘car cost benchmark’.  

That said, the car benchmark is useful if your budget doesn’t allow for a car of any sort per month.  At that point, you’re likely limited to Google Analytics multichannel funnel reporting. While going that route is worthwhile, understand that it will only give you a read on click path attribution, and won’t give you insights regarding offline-to-online attribution or cross-device tracking. It also can’t help you to model algorithmic ‘what-if’ scenarios for media mix modeling.  There are additional point solutions in the market that allow for a read on specific channel interactions, like view thru contribution of display advertising to paid search, which you can achieve by consolidating on a campaign management technology like DoubleClick.

Incremental lift required to shift investment

As Einstein famously had his ‘thought experiments’, before you decide to invest dollars and time in a multichannel attribution platform solution, I recommend you first conduct the following thought experiment:
You likely know how much revenue your whole business is generating, which means you know the upper limit of what you are trying to attribute.  Similarly, you also know what credit the last click or bottom-of-the-funnel channels are being attributed. What you are likely not sure about are questions such as:  How much of my direct load traffic or branded search traffic is driven by higher funnel engagement with other channels or campaigns? Or, if I increased my investment in upper funnel push channels like display, broadcast or even direct response channels like direct mail and DRTV, what would be the halo effect into bottom funnel channels?  This is where our ‘Incremental Lift Required to Shift Investment’ variable comes in.

What do you need your lift to be to merit increased investment in those upper funnel channels?  If it is a percentage lift between 5 and 25%, you are a strong candidate for investing in a paid-for multichannel attribution technology (some combinations of business verticals and marketing channels could have 25% to 100% potential halo effects, although this is less common).

On the flip side, if you need to see a multiplier halo effect of your channel spend to convince the powers-that-be to increase or shift budget, the entire multichannel attribution exercise will probably quantitatively disprove the return on that increased channel investment.

And that leads us to the critical first step for driving ROI with multichannel attribution. Consumer interaction across multiple touch points along the funnel is very real, and so are your costs and time for complex marketing technologies and analysis. By definition, this exercise can only be ROI positive if the time and cost invested in the effort shapes the marketing decisions your company will make.

So, before you break ground on your own cost-intensive multichannel attribution science experiment, challenge your business stakeholders and yourself to establish what the acceptable costs of implementation are, as well as the necessary attribution thresholds to shift your budget allocation strategy. By identifying these values, you will have made a great first step to positioning your multichannel attribution project for success, as your organisation will be better prepared to act in response to the possible data-driven outcomes.

And that first step was free.