Multi-touch attribution is the acknowledgement that multiple marketing campaigns may have contributed to the success of a conversion. There are a lot of different tools that let you track the journey of a visitor. You can see when they first received your e-mail, see a Google display ad on another website, visited a blog post on your site, before placing an item in your cart and completing an order on a newly released product. The problem is, you have:
- An e-mail marketing campaign (first e-mail)
- A display ad campaign (Google ad)
- Content marketing campaign (blog post)
- The recent release of a new product offering
Your new customer touched all four points in their journey before converting. So who gets the sale?
The challenge with attribution
There are multiple ways to attribute the sale to each campaign. You might give equal credit to all campaigns. Perhaps only the last touch receives the full credit. Maybe more weight to the final touch points of the customer journey than other touch points.
Each approach has their merits. However, there isn’t a perfect approach simply because the behavior of the irrational customer is simply unpredictable. That is, there are a lot of emotional factors involved, a lot of potential distractions or influences, and the newly converted customer might not be who we first map them out to be.
We also have to keep in mind that certain assumptions we make may be incomplete or false. You can run all sorts of survey funnels and focus groups you want. You can get as close to the truth as you possibly can.
In the end though, we can never be 100% sure. Do we simply ignore the idea of attributing a percentage of the sale to certain campaigns? No! But it is up to the marketing team to decide on an educated customized attribution method that makes the most sense.
For example, you wouldn’t attribute 100% of the sale to a display ad your visitor viewed from a single publisher. Nor would you attribute half of the sale to the last campaign in a sale when there are five other campaigns running simultaneously that might have influenced the customer.
Critical questions to ask yourself when considering multi-touch attribution
How significant is your budget for digital marketing campaigns? There are two numbers that play a role: 1) your overall total digital marketing budget; and 2) how much each campaign gets. Small budgets (<$5,000 per month) might not warrant any consideration. Your priority might be to build long-term assets with your content marketing and search engine optimization campaigns, while running Facebook marketing campaigns to pull in some love and eyeballs. But thinking too much about multi-touch attribute might not be the best course of action at this time. When you deal with larger budgets (>$10,000 per month), you can begin to allocate some resources to prepare yourself for setting up a multi-touch attribution model. $100,000 per month or greater budgets should definitely have resources allocated to put together acceptable multi-touch attribution models.
How many campaigns are you running in parallel with significant spend in each? If there are three or more, you can begin arguing why one campaign should received a larger allocation than the other. Any less doesn’t warrant attention yet. As long as you can continue collecting cohesive data (under one tool), you can look back to deduce what happened before. Keep in mind you want to avoid data that is siloed across multiple platforms. That is, data shouldn’t be segregated on different platforms.
How many micro-conversion events do you have in your customer journey that significantly ties to the bottom line? Micro-conversion events are little commitments made by a user before they reach a (macro) conversion goal. For example, signing up to an e-mail newsletter, viewing a video, and adding something to their shopping cart are micro-conversions. Although they don’t generate a true economic value, they are significant milestones that can ultimately lead to a purchase, a donation, or a potential lead. The more micro-conversion events you have, the more data you have to work with that can produce insights you can act on.
How large of a role do offline marketing campaigns play? This is a tough question to answer. Not only is it difficult to track offline campaigns, but the audience group might not be as targeted as your digital marketing campaigns. You will want to do your best to isolate visitors coming from your offline marketing campaigns. Say, with a new domain with redirects that have tracking parameters indicating they are coming from a particular source or a redeemable coupon code on a print ad! In the end, it’s another judgement call that may require further investigation to come up with an educated guess for attribution.
An example from Web Analytics 2.0
Web Analytics 2.0 is a book written by the famous Avinash Kaushik. A definite must read for those fascinated with the analysis aspect of digital marketing.
In his book, he posed the following example of a journey where his eventual customers were exposed to seven touch points, five of which were digital marketing campaigns.
| Campaigns | Credit allocation | Avinash’s Rationale |
| 20% | “because they might be a customer, I do a lot of work, and I deserve a lot of credit” | |
| Yahoo! banners with view-through | 5% | “because I don’t value view-throughs” |
| MSN home page promotions | 12% | “because this is a scatter-shot approach” |
| Google PPC | 18% | “because I can show up only for exact match relevant queries so these are very targeted” |
| Affiliate campaigns | 7% | “because the affiliates server customers who have already been exposed to other campaigns, so I value them a bit higher than view-throughs” |
| Bing organic | 20% | “because I love organic traffic, it’s free, and I love Microsoft” |
| Yahoo! PPC | 18% | “for the same reason as Google PPC” |
I believe this is really insightful on the general journey online visitors might run into. I could be wrong though (and am prepared to be). It was interesting to note that Bing organic’s rationale for receiving 20% is because it is organic traffic that didn’t require ad spend.
Side thought: I’ve done a lot of work in the affiliate marketing realm, and I was wondering why Avinash attributed only 7% to affiliate campaigns. Perhaps I’ve been exposed to too many affiliate programs and offers that pay 50%+ or are businesses that rely heavily on affiliates to go out and promote their products and services. I would argue the affiliate campaigns deserve more than 7%, or even more than Bing organic, because they might lend an invaluable “impartial” authoritative voice when recommending your business offerings. I suppose it would be difficult to really understand why 7% was allocated unless we knew precisely what this business was. We should note the sale that is being allocated amongst the touch point was a $75 purchase in the book.
Wrapping up on multi-touch attribution
There isn’t a perfect way to allocate credit when we run multiple campaigns. Coming up with an educated guess on how to credit different touching points can get complicated, especially when you introduce offline marketing into the equation. I highly recommend reading Web Analytics 2.0 by Avinash Kaushik if you want to learn more about this concept. He covers different models thoroughly and include his point of view (denoted as “POV” in the book) on how each can be applied sensibly. In case you’re wondering which chapter in particular, you can find the topic in Chapter 12: Advanced Principles for Becoming an Analysis Ninja.