The Attribution Challenge – Part 1
Updated: Oct 18, 2019
I guess it’s no surprise that Google travel research surveys find average travelers research no less than 38 sites, plus social media activities, before finally making a booking. Which got me thinking: how can we track ROI of each sale when there are so many touch-points involved?
It used to be easy – the “last touch” method was ol’ reliable … by crediting the last place visited before they made a booking we had the source. But, today? Over 38 sites? What about SEO and social media, the two key ways we spend time and money engaging our guests? Do they get the short shrift?
HeR submits it is high time to adjust attribution methods and today we take the first of a two part look at this topic. Let’s review first the value of a “multi-touch” model to calculate ROI of our marketing programs: Point of Sale, Point of Decision, Point of Discovery and Point of Awareness are the keys to be integrated in the latest attribution analysis methodology.
In a “multi-touch” set-up, you allocate the conversion value across all of the marketing programs. The most common multi-touch model is very simple: all touch points receive an equal share of the credit. For example, a $200 booking with 20 touches would result in “a $10 attribution for each touch point.” More sophisticated marketers might weight different touch points differently – maybe the first and last split 60% of the value while the remainder gets spread evenly across the middle.
There are a number of considerations to review in setting a new policy, not the least of which are the peculiarities of the social channel(s) which make it difficult to cleanly and accurately attribute conversions.
Social isn’t considered a “direct response” channel, although there are increasing advanced functionality coming down the pike, specifically booking engines within Facebook and Twitter. Generally speaking though, aside from daily deals and coupons, most of the content social media marketers publish is about generating engagement and interaction with existing and potential customers. Some of the links you share might not even link back to your website.
As a result, social is very rarely the last marketing touch point before a conversion. This makes it even harder to accurately track the impact of social media marketing campaigns, especially if you are using traditional web analytics tools.
Adding to the challenge of attributing properly is the following: when the last touch comes from mobile, desktop applications, email clients and IM chats it is not as easy as it sounds! Many online marketing analytics platforms currently categorize visits by “site referrers.” Translation? the site that “referred” the visitor to your site. But are you aware that non-browser-based applications may not pass web referrer data?
The reasoning is quite simple. Web analytics platforms ask the question “What web page did you come from?” When coming from one of these applications, the answer is “I didn’t.” So, without web referrer data, your web analytics platform tracks this visitor as if they typed your domain.com directly into their browser or clicked on a bookmark.
Avinash Kaushik, Analytics Evangelist for Google, says that 78% of people consume Facebook and twitter content via applications. Let’s marinate on that number for a moment: 4 out of 5 people read your social content on an application that may not pass web referrer data, which is the core nugget of information that enables your web analytics platform!
So, yes, a big chunk of the reported direct visits to your website are people that click on links to your site on mobile devices and on other non-browser technologies that may not pass web referrer data. Cookies and conversion tracker tools help, but the challenge remains and marketers must be diligent to ask the right questions and include the right tracking codes for effective measurement.
The secret to really understanding the effectiveness of your efforts is identifying the influence value that touch-points 1-19 had on the guest before that 20th visit resulted in a booking. Here are 3 main steps to building a more effective attribution model:
1. Define the attribution window
Define how far back you want to measure influence on the booking. Rarely does a guest make 1-stop that results in a sale. Considering all influences that led to the booking will help you be a more effective marketer. Start wide and narrow as you learn your most effective window.
2. Identify the shape of the model
Attribution data is mostly read on a timeline. Viewing data this way helps you identify whether the influences are normally strong at the beginning middle or end. This may differ from campaign to campaign. Be flexible so you can properly identify and exploit each campaign’s unique characteristics in your future plans
3. Build the data system
As you capture the data, identify the paths and touch points at a macro level, you can begin to determine the proper weights to assign to each. These will help you answer the questions of how often your marketing campaigns will be measured and what defines a successful campaign.
In part 2 of this series next week, we will offer a more detailed steps for devising an attribution model for multiple channels that more accurately reflects your company’s business.
Meanwhile, share with us how you’ve adjusted your attribution model.
How has it improved your successes in ad placement and analysis of your guest profile?
#admeasurement #hoteltrends #lodgingproperty #HospitalityAnalytics #adattribution #marketingattribution #socialmediainfluencers #ROI #advertisingtracking #travelwebsites #advertisingattribution #hotelsalesforecast #revenuemanagement #hospitality #hotelmarketing #traveltrends