A Pragmatic Approach to Attribution
Attribution is a hot topic in marketing that keeps just getting hotter. The problem is that, all too often, it gets treated as an all-or-nothing proposition.
As John Wanamaker observed almost a century ago, much of our advertising dollars are wasted. If only…if…ONLY!…we could figure out which dollars those are, we could finally stop dreading those meetings with the CFO! Surely, with the explosion of data and data platforms that have come with the shift of so much consumer behavior into the digital realm, we should finally be able to perfect our media mix. Right? Right?!!
The unfortunate reality is that, as long as we’re marketing to human beings in a chaotic, message-overloaded world, there will be some degree of unknowable waste in our marketing. Step #1 in an appropriate perspective regarding attribution is to treat any vendor’s claim to have “solved” attribution with a level of skepticism that matches the cost of their offering. That perspective leads us directly to those three magic words: return on investment.
ROI calculations have two components: the cost (investment) and the return (value delivered). In the world of marketing investments, there are some important subtleties:
- The return is the incremental increase in business results (revenue or others) that we are able to achieve as a result of the investment.
- The cost of using attribution as a tool for guiding our investment can vary wildly — from almost nothing to six-figure investments in complex platforms with highly involved data collection requirements
This means we are perpetually performing a balancing act — trading off the incremental value we will get from each improvement in accuracy or completeness of our attribution, with the cost of that improvement. With attribution, it’s easy to hit a point of diminishing returns — substantial increases in investment may yield only nominal improvements in value.
It is entirely too easy to get wrapped up in seeing “attribution” itself as the goal, rather than recognizing it as a means to an end: effectively and efficiently investing your marketing dollars. Or, even better, maximizing the return from your marketing investments.
What Is Attribution And Why Is It Messy?
Skip this section if you know the answer to this question already. There is no earth-shaking definition here, but some of the terminology dropped later in this post assumes a basic level of familiarity.
Attribution, put simply, means accurately and effectively assigning value to a marketing channel, campaign, or tactic. That way, we can compare the cost of that channel, campaign, or tactic to the value assigned to it, and derive the ROI described above.
So, why is this messy?
In a simplistic scenario, it’s not. Imagine that no one has ever heard of your brand: Super Awesome Widgets. And, imagine that your only promotion of the brand is through Google Adwords. And, imagine that consumers click through on your paid search terms, navigate the site, and place an order on the site during the same visit. Every time.
In that case, it’s easy and clean: 100% of your site’s revenue can be attributed to Google Adwords.
But, exactly zero brands operate in a scenario like that.
Another (still) fairly simple scenario would be:
- A user searches for “widgets.” She sees a paid search ad and clicks through on it, navigates the site a bit, and then leaves without ordering. But, now she has a good idea of what you offer widget-wise and and at price.
- A few days later — realizing she really needs to get around to buying a widget — she searches for widgets again. This time, though, she clicks through on an organic search result. She then navigates the site and makes a purchase.
In this case, which channel should get the “credit” for the revenue? Is it paid search, since that is how she first came to the site for her initial research? If paid search receives all of the credit for the revenue, then that is considered first touch attribution.
Or, rather, should organic search get the credit, since that was the last channel she used before making the purchase? That would be considered last touch attribution.
Or, should both channels get 50% of the credit (linear attribution)?
Perhaps both channels should get full credit (wildly irresponsible attribution)?
This scenario illustrates why attribution is complicated. And it can be very expensive. The rest of this post breaks attribution down a bit and lays out how, by keeping an eye on ROI prize, it’s possible to incrementally improve the sophistication and complexity of an attribution management program.
Three Different Dimensions of Attribution
One way to think about attribution is that there are three dimensions of the problem. Each dimension ranges from “basic” to “advanced,” and, to some extent, we can progress along each dimension independently from the others:
Keeping the ROI discussion above in mind, we can now think about which dimension(s) make for the most sensible incremental investment. There is no one-size-fits-all!
Channels: From One to Many
Each marketing channel is unique. As such, a typical starting point is to optimize within each channel independently. For instance, you may use ROAS or some other metric within your paid search efforts to adjust keyword bids. Contrast this with decisions you make regarding how often to send promotional emails — likely optimized for a combination of clickthroughs and conversions while keeping an eye on the opt-out rate.
This approach isn’t inherently “wrong.” It’s actually a fine place to start — certainly much better than simply throwing investment at each channel and not assessing and tuning the investment within that channel.
On the other end of the spectrum is attempting to manage the interplay of all channels with each other. This is…complex and expensive. So, is there some middle ground? There is!
The middle ground involves a little thought and an incremental increase in complexity and cost:
- Assessing which channels can most directly be impacted (“Our direct traffic has the highest conversion rate! We need to drive more direct traffic!” Um. Good luck with that.)
- Assessing which channels, logically, we expect to be the most intertwined. Paid search and organic search are two obvious ones here: two different channels (in the marketer’s mind, if not as clearly so in the customer’s) with differing ability to directly and immediately impact. But, focusing on just those two channels, including possibly doing some controlled experiments, can yield a better understanding of both channels and how they interact, which can lead to more informed and effective investments.
That’s one dimension of attribution, and that dimension, alone, offers many opportunities for incremental improvement.
The Length of the Attributed Journey
Similarly, there is a “basic” starting point for the breadth of the consumer journey that we consider: the journey starts when a customer arrives on our website, and we consider it completed when the customer “converts” on the site.
This is relatively straightforward, in that this basic level can be performed entirely within the web analytics platform (although it has dependencies on the quality of the data, including accurately tracking the sources of traffic to the site).
The length of the journey can be expanded both upstream and downstream:
- Ad impressions occur upstream of the actual clickthrough. (For that matter, attribution efforts always ignore “legacy brand feelings” — I was raised in a Colgate household, so how much does that impact my toothpaste purchasing decision? As it turns out, I now use Crest…because my wife was raised in a Crest household and generally does that shopping. It’s going to take a lot of marketing to overcome habit!)
- Focusing on the immediacy of an on-site conversion may or may not be the main value of the customer relationship. Does the first purchase lead to additional purchases? Do customers who look nearly identical at the time of the online conversion actually have vastly different lifetime values?
Like our first dimension, it makes sense to start simple and then progress from there — weighing the cost of the increased complexity with the value that that progression provides in return.
Model Approach: From Heuristic to Algorithmic
This final dimension may be the one that causes the most confusion. It’s also the dimension with the fanciest terminology.
Fundamentally, there are two different ways to approach attribution:
- Heuristics — this simply means you, the marketer, chooses how you want to assign value along a multi-touch path. 100% of the value to the last touch? 100% of the value to the first touch? Evenly spreading the value across all touch points? Weighting the the value so that later touchpoints get more value? When Google Analytics first rolled out their multi-channel attribution solution (prior to their acquisition of Adometry), this is how they defined attribution. It’s not inherently “wrong” to attribute value this way. It just means that value is being assigned based on the judgment call of the marketer.
- Algorithmic (also referred to “data-driven”) — this is much more complicated, requires more data (and more detailed data), but, in theory, is more objective (and, in theory, more “right,” but “right vs. wrong” is a slippery slope perspective that we should try to avoid!).
In both Google Analytics and Adobe Analytics, the pseudo-default view of traffic source is the heuristic approach of last touch. Both platforms make it very easy to compare the data from a first touch and last touch perspective, and that’s a comparison that it’s worth making:
- The results will be different, BUT
- Are they so different that you would act differently if using one approach over the other? This isn’t a question with a single answer: it really depends on the nature of the business being assessed.
Google Analytics makes it easy to go one step farther and evaluate additional models. Which, if the first- vs. last-touch comparison raised an eyebrow or two, it’s worth exploring.
IF this heuristic comparison turns up wildly dissimilar results, then it may be worth exploring algorithmic attribution. Google’s new Google Attribution platform will actually include “data-driven” as a default option (but will require sufficient data volume in order to work). It’s too early to say for sure, but I suspect that option will generally wind up just being a starting point for algorithmic attribution. Things get complicated quickly when moving to the more advanced end of that dimension!
In Short: There are Many Options
Consider a mid-sized online retailer with a limited budget for digital marketing and limited staff to manage that budget. The retailer has historically spent all of its budget on Google Adwords, but is considering experimenting with shifting some of that spend to Facebook advertising. Using the approach described above, the retailer might:
- Start by evaluating the spend within the single channel of paid search to identify the lowest performing keywords based on the cost and then the revenue generated.
- As a quick check, assess, at a channel level, the differences in the revenue attributed from first click attribution versus last click attribution.
- Assuming that this simple check does not return two drastically different stories, shift budget from the lowest performing keywords to Facebook advertising.
- If the overall results improve, then, potentially, simply focus on optimizing within each channel.
- In addition, consider some controlled experiments: turning off Facebook advertising in a few geographic regions and compare the results to regions that had performed similarly, but where Facebook advertising has remained turned on. Perform this same experiment with paid search. In both cases, evaluate different (heuristic) attribution models: first touch, last touch, etc.
While this is a relatively “simple” scenario — it sticks generally to the “basic” end of all three dimensions of attribution — it still requires planning and diligence, and it can yield meaningful information about the value being delivered by each channel, as well as how each channel can be adjusted to improve overall results.
Contrast that scenario with a large, multi-channel retailer that has significant investments in both offline and online advertising. The team has already done quite a bit of intra-channel optimization, and has also performed a range of experiments over time to get a good understanding of the interrelationships between different pairs of channels: when a TV campaign runs, they know what sort of a bump to expect in paid search, and they already have a mechanism in place for attributing that bump back to the TV campaign. The team has also performed experiments with turning off and on different ad groups — brand terms vs. non-brand terms — to get a good sense on the interplay between paid search and organic search. And, they have adjusted their paid search spend accordingly.
The retailer is planning to increase their digital advertising spend significantly over the next 1–2 years and wants to ensure that their media mix remains appropriate as they do so. In this case, the organization may be ready to move more to the “Advanced” end of the attribution dimensions:
- Rather than working off of the revenue generated from each order, estimate the customer’s lifetime value at the time each order is placed, and use that as the primary attribution metric.
- Shift from heuristic attribution to algorithmic attribution (which may require investment in a platform and/or internal resources with more advanced analytics capabilities)
- Work to include impression data — at an individual level for digital marketing and at an aggregate level for offline channels — in the inputs being used for the algorithmic attribution.
In both cases, the goal is to find the right level of complexity (aka, cost, investment) to make effective decisions. For one organization, an investment of several hours a week — and no investment in additional technology — may be the “right” approach to attribution. For another organization, a potential six-figure investment in people, process, and technology, as well as a multi-year roadmap, may be warranted.
At the end of the day, any assessment of traffic/order/lead/revenue sources implicitly includes some form of attribution. That may be very basic, but it may be good enough. It’s almost certainly worth some exploration of increasing the sophistication of the attribution approach, but that doesn’t mean an immediate commitment to spend six figures on an advanced, algorithmic, multi-channel, offline/online solution. That may be where you ultimately wind up, but it’s best to wade steadily into those waters rather than taking a blind leap off the 30-meter platform.
At Search Discovery, we’ve ranged from the shallow end to the deep end of the pool depending on the client and their needs. We’d love to suit up with you, so contact us if you’re interested in learning more about our approach and experience.