Last-Click Attribution Is Not A Hard Habit To Break

jaystocki“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Jay Stocki, vice president of digital services at Experian Marketing Services.

It’s been said that the easier the habit, the harder it is to change. That must be why nearly every day I find myself explaining to marketers why last-click is rarely an effective form of attribution for display advertising yet it’s still being used like a bad habit.

I can understand how the habit forms. Last-click is the easiest form of attribution, it can be quickly explained to senior executives and it does not require cooperation from other parts of the marketing organization. But please just say no. Last-click is a terrible habit.

Case in point: Just 13% of marketers surveyed in an eConsultancy and Google study believed last-click was a very effective method of attribution. Yet 54% said it was their most common form of attribution.

“Last-click attribution is like hiring a sandwich board guy to stand in front of a big retailer, and then giving him credit for all sales from every customer while he is on duty,” said my colleague Brie Pinnow.  “Not the email campaign, loyalty program, TV spots, salesperson, display ads, search or direct mail. Just give the credit to the last ad the consumer saw – the sandwich board.”

When you look at it that way, relying on last-click attribution for your marketing budget allocations suddenly seems foolish. Even worse, giving credit solely for last-click has enabled less-than-scrupulous vendors to play games to boost performance. Instead of trying to influence potential customers down a conversion path, the bad actors simply do their best to get cheap ads everywhere, hoping to stick their virtual sandwich board guy (the last cookie) right in front of the consumer before they purchase.

This has led to wasting money on ads that are never seen or, even worse, on ads that are never actually displayed, such as those displayed behind web pages. As an industry, we need to put a stop to this reliance on such a misleading metric. The urgency to change this habit has become paramount.

But last-click attribution, like any bad habit, can be broken with the right mindset, a solid methodology and, above all, persistence.

The Right Mindset

Like the road to recovery from any bad habit, the first step is to admit that there is a problem. If your marketing spending is divided among multiple channels, you’re already aware that the consumer’s decision-making process is more complicated than a single click or action. Further, as consumers ourselves, we understand firsthand how the proliferation of devices has dramatically changed how we discover, research, compare and purchase products.

So why is all of the attribution based on that one click? Admitting that the attribution and purchase path are out of alignment is the first step in the journey to better attribution.

The good news is that improving attribution methodology is a lot easier today than even a few years ago. There has been dramatic growth in the amount of tools, vendors and experience. The recent update to Google Analytics now enables multitouch modeling through the Model Comparison Tool.  The right methodology, of course, is multitouch modeling, where spending is allocated to a channel based on the performance of its response or impact. The more impactful it is, the more attribution and greater spending it deserves. Agencies, data providers and attribution tool developers are a great source of ideas, technologies and methodologies, and we’re likely to see more investment in this space in order to integrate cross-channel identifiers with custom analytics capabilities in a single platform.

Just Say No

As advertisers shift into mobile and ramp up spend in that channel, those still clinging to last-click attribution will move farther down the rabbit hole of wasted ad spending and uncoordinated programs. If the vision of most marketers is to arrive at an omnichannel marketing capability, the first priority should be to elevate attribution from a functional to a strategic level. Yet, despite all the omnichannel rhetoric, we are still in the early adopter days with coordinated attribution programs.

Attribution is the Gordian knot. There will never be a perfect formula. It will require persistence and dedication, but the end result will be a significantly better understanding of what channels are most effective and deserving of increased marketing spending.

It is tough to break a bad habit. Changing attribution methodologies will cause organizational pain. Some hypotheses will fail. Reconciling data from multiple channels can be excruciatingly difficult.

Take the first steps and stick to it. “Bad habits,” as the Yiddish proverb goes, “are easier to abandon today than tomorrow.”

Follow Jay Stocki (@jstocki), Experian Marketing Services (@ExperianMkt) and AdExchanger (@adexchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. Great article, and I agree with you wholeheartedly on all points except that you’re still only referring to direct response attribution. And online DR attribution is still based on the flawed strategy of trying to move people around the internet via banner ads more or less against their will. This is still tip of the iceberg performance measurement–it’s just a better way of measuring the tip of the impact of an ad, but still misses the rest of the iceberg above the surface as well as below the surface.

    For ecommerce companies and pure DR advertisers, only measuring response makes sense. But for any other company whose strategy is to build awareness, familiarity and consideration, or drive store or showroom traffic, measuring only response “the easy way” via clicks will result in wasted time, effort and money. As an industry, we need to demonstrate how online ads affect awareness and opinion and even behavior on the publisher’s site (and advertiser’s site downstream). Some publishers are doing this today. Other publishers are either incapable or are feeding the click attribution monster. Until more leaders emerge to question the machine (in publishing, the agency world and especially on the client side) and create a better model, we risk online advertising being held back from its true potential as an advertising medium.

  2. Interesting article; one point that might be worth clarifying: by “last-click” attribution, do you mean that the user actually clicked on an ad, or merely that there was a (claimed) impression of the ad that dropped a cookie?

    Clearly if a user clicked on an ad, then many of the criticisms of the attribution method above do not apply; there is specific evidence of user interaction with that ad, so attribution in this case is pretty fair* – which is not at all like the sandwich board guy analogy. When a click occurs, this is more like the case where a coupon is cut out of a magazine and mailed in, which links a specific sale back to that magazine.

    * Attribution is fair, as we can be relatively confident that the interaction would not have occurred had the ad not been shown. There are certain caveats: the action may have been assisted by other, hidden influences such as brand awareness campaigns, or there could be some other issue with the attribution, such as robotic click fraud.

    Conversely, in cases where an impression occurred, but there is no specific indication that the impression was a conduit to an action, then all the points you make above are valid – we don’t know if the ad caused any subsequent actions, whether the ad assisted these actions, or was totally irrelevant.

  3. Affiliate Marketers – are they the stepchildren in the purchase decision? Somehow I missed mention of that channel in influencing a purchase decision. And, oh yeah, couponers – causing cart abandonment to search for a promo code…I think I missed that, too. 🙁

  4. This is a great conversation and I agree that excessive use of last click attribution may be seen as a bad habit. However, I read from the buzz around the world that efforts to try influence people for not having last click attribution may also be seen as a very bad habit.

    To take a more neutral stance in front of this debate, I would like to suggest that the right technique for measuring attribution may very much depend on the marketing objective behind the campaign and the Brand. Having said that, I also recognize that most of the Brands I have been working for over the past 20 years scarcely have one single marketing objective.

    So, thank you again for this passionate conversation which helps clarifying some of the best practices in measuring attribution. However, my dream remains the same :
    I wish that some day, there will be a clear state of the art:
    * Describing for each marketing objective what is the range of relevant attribution techniques,
    * And explaining why it is so important to have different analytics angles to consider rather than one.

    Before it happens, I will continue believing that several measures is much better than a single one when trying to understand the state of consumer engagement and how far Brands are from achieving their marketing goal. I will also continue being very cautious when analytics being used to measure consumer response only present a partial explanation of campaign successes and failures.

    Having said that, I am amazed by progresses recently made with analytics in digital communication. I will also continue setting a very high level of expectations, because I do believe we are not yet where the power of analytics will take us tomorrow.

    Thank you again for sharing your passion around the subject.