SEO Tips for CEOs and Marketers: Improve Your Digital Marketing Strategy by Changing How You Measure Conversions Pt. 1
Is your SEO team giving you the data you need to improve your company’s digital marketing strategy? Most SEO reports cover information on general traffic trends and conversion stats, often breaking them down by medium (i.e. visitors from organic search, referral sources, paid advertisements etc.) and providing some additional information on popular pages and top conversion sources. But those top conversion sources do not tell you the full story about how people find your website and turn into leads.
By default, Google tracks conversions using a “Last Non-Direct Click” model. This model gives credit for conversions to the medium responsible for the converting visit.. For example, if a visitor enters your site from organic search (Google) and completes a goal that you have set up in Google Analytics, then Google Analytics will give full credit for that conversion to Organic search, regardless of if and how the visitor got to the site on previous occasions. (Note: This model excludes the last interaction if it is a direct visit and gives credit to the previous visit for a variety of reasons that are not worth discussing in this post.) In some ways, this offers very valuable information in that it indicates which medium provides the final nudge to bring the visitor from being a passive visitor to a potential lead or customer. However, this model falls short if your business is more interested in seeing which medium initially connects visitors with your company or in understanding that the b2b industry often requires 7-12 touch points to turn a visitor into a lead. For this reason, you may want to work with you SEO team to select a conversion attribution model that specifically highlights the points of a visitor’s conversion path that matter most to you by considering the following questions.
What is your company trying to accomplish with its website and digital marketing efforts?
- Improve E-commerce sales?
- Increase lead generation?
- Identify qualified leads?
- Expand brand awareness?
How long is your sales cycle?
- By default Google Analytics looks at visitors’ interactions over the 30 days prior to converting; however, for longer sales cycles, you can extend that window to 90 days to get a better sense of how many times consumers visited your site and how they got there before converting.
What specifically are you trying to measure?
- The impact of marketing efforts to bring qualified traffic to the site?
- Which medium ultimately succeeds in getting visitors to convert?
- The role organic search and other mediums play throughout a visitor’s journey prior to converting?
Depending on your answers to these questions, the Last Click attribution model might not provide you the information you need to improve your digital marketing strategy. In order to help you better measure the effectiveness of your marketing efforts, I’ll review some of Google Analytics’ alternative attribution models that you and your SEO team may prefer to use.
Last Non-Direct Click
As mentioned above, this model gives the entire credit to the medium that contributed to the final visit.
The Last Non-Direct Click model is particularly useful for businesses with short sales cycles that need to know about what exactly turned their visitors into customers. For example, an ecommerce business that sells discount video games might not be interested in how visitors arrived at the site if they are just researching a new game and sometime later become a customer. Instead, they might be able to improve their marketing by knowing which medium delivers visitors who purchase products on that session. With this information, they could expand those efforts to bring in more customers or try to improve other strategies that bring in traffic that doesn’t convert.
As mentioned above, this model fails to capture the entire picture of a longer sales cycle and provides little insight into which medium initially attracts customers to a site in the first place.
This model attributes the entire value of the conversion to the first medium through which a visitor entered the site.
This model is particularly useful for evaluating the effectiveness of brand awareness campaigns by highlighting the tools used to get visitors to the site in the first place. Companies looking to review the impact of PPC advertising or social media outreach may choose this method in order to evaluate whether these efforts are driving qualified traffic to their respective sites.
Similar to last click attribution, this model does not provide insight into the full conversion funnel. Also, it does not give any credit to the last medium prior to the conversion, which is therefore not helpful to companies seeking to evaluate which medium finally pushes visitors to convert.
This model gives equal weight to each step along the conversion funnel, so if a visitor first enters the site through organic search, comes back through paid advertising, and finally returns via referral, then organic search, paid advertising, and referral would all be attributed 1/3 of the conversion.
This model is particularly helpful for better understanding the entire customer’s journey. It is useful for those looking to capture how each part of their marketing efforts play into a conversion, and more importantly it will highlight when a given medium is absent. For example, you might assume that your company’s social media activity might not be the first or the last click but is still helping direct visitors back to your site in the middle of the conversion funnel. Under this model, if social media is not listed for contributing to conversions, you might want to rethink the audience that your social media targets, for it might not be bringing in the correct traffic you need to meet your company’s objectives.
As the title implies, the model is linear and therefore gives the same weight to an interaction in the middle of the funnel as on the ends, which limits your ability to evaluate which medium specifically brought people to your site or helped them finally convert.
Like the linear model, time decay also divides credit for a conversion between the visitor’s interactions prior to converting. However, in this case, it gives more weight to the final interaction and decreasing amounts of credit to earlier visits based on how close in time they were to the conversion. For example, if someone is looking into purchasing plane tickets, he/she may check prices on a website, return several times to check different flights options and finally come back to purchase tickets after seeing a remarketing ad about a sale. In this case, the remarketing ad deserves a large portion of the credit for directly assisting with the conversion; however, the earlier visits from organic search and website referrals also played a large part into the visitor becoming ready to purchase those tickets. In this instance, each medium would get credit according to how close that interaction was to the final purchase.
This model combines some of the strengths of the last interact click and linear models, by highlighting the medium that ultimately resulted in the conversion without sacrificing some insight into the visitor’s previous entrances to the site. This is particularly useful for companies evaluating the impact of a promotional period for a product or looking at how each medium contributes to getting a visitor through a long research period to finally convert.
As this model is designed to value earlier interactions less and ones close to the final conversion, it is not particularly helpful in learning about the impact of initial outreach efforts that bring visitors to the site in the first place.
Position Based (U-Shaped)
The position based model can be visually thought of as a U. In this model, the first and last interactions each get 40% of the credit, and the interactions in between divide up the remaining 20%.
This model combines some of the best aspects of the first interaction, last click, and linear models, by giving more credit to the mediums that bring visitors to the site and finally help them convert, while still providing insight into the full customer journey. In this way, this model is most useful for companies and marketers looking to get a holistic understanding of which mediums bring in qualified traffic and turn them into leads or consumers. It is also helpful for companies with long sales cycles as it captures each touch point involved with bringing a customer back to a site while still highlighting the importance of the medium involved with visitors’ first and last interactions.
This model does not take path length into account, so for companies with long business cycles that tend to bring visitors back to their site numerous times, that middle 20% gets divided up into very small pieces, potentially over emphasizing the relative difference between the middle interactions and the ends. Also, as it does not distinguish between the first and last clicks, it doesn’t provide the specific insight into where in the conversion funnel a given medium is particularly effective. In this way, the model may not provide the feedback a company needs to better understand how to improve a given piece of the funnel.
Google Analytics Custom Models and Top Conversion Paths
In addition to the above fivemodels, you can also create custom models in Google Analytics, by taking one of the above models and modifying it. In this way you might change the ratio of the position based model to give the center more weight or you could add rules such as to exclude paid traffic or to include only traffic from Google. Although frequently one of Google’s existing models will fit your needs, this gives you and your SEO team more flexibility for tailoring a model to focus on your company’s needs and objectives. Moreover, Google Analytics also gives you the ability to look at visitors’ top conversion paths. So, in addition to getting a high-level overview from a select attribution model, you can also get a visual understanding for where select mediums are involved throughout visitors’ interactions prior to converting.
By reviewing your company’s objectives and how you expect your marketing efforts to contribute to meeting them, you can get a better sense of what information you need to evaluate the impact of your work. With this in mind, you may want to explore using a different conversion attribution model that better captures you marketing as a whole or specifically focuses in on a piece that you’re looking to improve. If you want help bringing your digital marketing efforts to the next level, contact Walker Sands to see how we can help you design, implement and ultimately measure the success of a comprehensive digital marketing program.