Everything You Need to Know About Attribution Modeling

No business has an infinite marketing budget. It is impossible to simply throw money at campaigns across a variety of media in the hope that something will stick and conversions will be made.

Even if this were possible, it would still not be a great use of marketers' time. In this competitive landscape, marketing campaigns need to be effective and strategized from the start, driving revenue each quarter.

If you're unable to unequivocally say what triggered a conversion, resulted in a sale and ultimately made your company money, how can you repeat the process?

What is attribution modeling?

If you've been continually relying on older methods like click-throughs and open rates but feel as though you aren't getting the best ROI, then it might be time to take a look at attribution modeling.

This is defined by Google Analytics as:

Put simply, it represents a complete map of a buyer - from unknown prospect to potentially loyal customer and then brand advocate. It allows marketers to measure the impact of each piece of communication they issue, whether that's a tweet, blog post or pay-per-click (PPC) ad.

It's a way of giving credit to each marketing activity based on its effect on the ultimate conversion and business goals, whether this is revenue or customer retention.

How attribution modelling can solve workplace problems

Marketing attribution modeling doesn't look at macro-level activities, but rather showcases the entire journey of engagement with your brand to clarify which particular factors are influencing users at any given time.

Importantly, this allows marketers to be able to invest more on what really works, readjust or even cut what isn't effective, and ultimately deliver more value to customers and the business - with the same overall marketing budget.

Broadly speaking, attribution modeling can help you achieve three things:

1. Help you quantify the value of every customer

2. Better understand customers

3. Provide tangible ROI

Where to start with Marketing Attribution

The first step to effective attribution modelling is determining what model best suits your marketing strategy. There are a number of different attribution models to choose from and ever-more being created as the phenomenon progresses.

Here are a few of the most commonly used:

1. First Click attribution

This attributes 100% of the credit to the first touchpoint your customer interacts with. So if someone finds you via Google and then eventually goes to your website and purchases, the search engine would receive all of the credit for that conversion.

2. Last Click attribution

This is one of the most commonly used methods and rewards the last customer touchpoint before a purchase, perhaps a web page featuring costings.

3. Last Non-Direct Click attribution

This model would see the final click in the buyer's journey that isn't a direct visit rewarded with the credit for a conversion, based on the idea that the brand's website shouldn't be included because the customer already made a decision to buy before they typed in that URL address.

4. Linear attribution

This model assumes that every stage of the marketing process is equally responsible for a sale - but it might not help marketers looking to make budget cutbacks. Also, is one campaign Tweet more important than a carefully-crafted landing page? Linear would argue yes.

5. Time-Decay attribution

This model is based on the idea that the closer in time it is to a conversion, the more effect a marketing activity has had on that decision. It is perhaps most useful in short-term marketing efforts.

There are many more models, because marketers love to try to quantify their efforts - especially when they are niche. However, the important thing to know is how to implement the technology to gather the information that will allow you to carry out attribution modeling.

How to implement Marketing Attribution Modeling

The first and most important thing to acknowledge is that for attribution modeling to work, you must have all of your marketing efforts - from display ads to tweets - tagged with campaign tracking parameters so that each part of a buyer's journey can be analyzed.

You'll then need to perform an analysis, which can be done with Google's Model Comparison Tool (free inside Google Analytics). This allows you to see your digital marketing channels and work out where to assign credit using the Assisted Conversions report. The default models will be Last Interaction and Last Non-Direct Click, but the exciting thing is that this can be changed.

You can create a custom attribution model for your own business using rules that you specify, measuring up to ten parameters at a time and layering them in as you go to work out what's important.

Try doing a review each month and see the results you could achieve. According to a recent iProspect report, the typical outcome of implementing this type of modelling is a 15% to 35% gain in media efficiency.

Doesn't that alone make it worth giving this technique a try?