Differentiating prospects from members in marketing analytics reports is not only achievable, it is not hugely difficult, and is a game-changer for marketers, especially for SaaS companies.
If your marketing site is visited by existing members looking for a login link, you have probably at least scratched your head about how this extra noise affects your conversion rate; perhaps you have even tried to filter out this member traffic without satisfactory results. Here’s why most attempts to do this don’t have the intended result, and how you can achieve the ultimate in segmentation and filtering by prospect and member type, using Google Analytics as an example.
Who needs this level of segmentation?
This will help any online service where registered members can access their login page via the marketing site or just consume content on that site. These might be B2C, or B2B tech offerings, or perhaps most in need of this level of segmentation are two-sided marketplaces. Not to mention Ecommerce sites, which can benefit tremendously from member vs. new shopper segmentation.
The exact solution, while easily replicable, should be tailored to your particular business model and site structure. For example:
- Simple Membership: visitors consist of prospects and existing members (also applicable in traditional ecommerce)
- Trial > Subscriber Membership: visitors consist of prospects, trialees, and subscribers
- Distinct Audience Types: visitors may consist of small-business prospects and enterprise prospects, small-business members and enterprise members
- Two-sided Marketplaces – visitors may consist of buyer prospects and seller prospects, buyer members and seller members
In case you are wondering; yes, you should be able to segment all of the above.
What is the purpose?
At a minimum, you need a reporting view that filters out returning member sessions. With this, you can at least get a more accurate understanding of purely prospect traffic trends, conversion rate and behavior.
At most, you may want the gold mine of filtered views that isolate particular prospect types and even particular member types in addition to a unified or rollup view where you can analyze your traffic mix trends. This is particularly handy in two-sided marketplaces, where each audience affects the trends and behaviors of the other.
Example: Most of the recent growth from Branded PPC is driven by seller-prospects while other prospect types have stayed fairly steady and a moderate number of returning members, more buyers than sellers have used paid ads to return to the site.
Why don’t the usual approaches to this filtering and segmentation work?
The most common things we have seen businesses do to try to exclude returning members are:
- Using exclude filters to remove traffic to the members only area, usually a specific hostname or directory in the url.
This does not work because this sort of filtering is done at the hit level. The effect is that hits in that section of the site are excluded from the view, but their sessions are still included when the user visits the marketing site to login.
It actually makes matters worse because you can no longer detect, using segments for example, what portion of traffic to the marketing site is also going to the membership site. Furthermore, it removes your ability to analyze post-signup behavior for prospect sessions that convert. Not to mention, in the Ecommerce scenario, this is likely impossible as members and prospects are usually using the same site for their shopping experience. Lastly, it inflates session count if you ever combine the data across views.
- Using separate properties for marketing site vs members-only areas, creates the same problems as separate views, plus can actually prevent attribution for sessions that start on the marketing site and then login to the members-only zone.
- Filtering based on a user-scoped custom dimension that is set when the user reaches the membership area comes closer to the mark, but also does not have the intended effect on its own because it filters out ALL member sessions, including the sessions in which prospects convert which are arguably the most important sessions to include.
- Segments can sometimes be used to exclude returning member sessions in the simplest model, but even there it has minimal usefulness and distinct disadvantages. Any meaningful analysis over time will be subject to sampling (in GA Free) which diminishes its usefulness and this same logic cannot be achieved within Google Analytics to filter a view.
How can this be accomplished with Google Analytics?
The key is a session-scoped custom dimension (a custom dimension that remembers the last value received in a session, applies it to the entire session and can be filtered on). This can be done similarly using Adobe Analytics visit expiring, last value, eVars, or Adobe also allows you to use more complex criteria combinations without sampling, making segments a more viable option in that tool for some sites.
For the simplest sites, that might be as easy as hits anywhere on www are prospects (until proven otherwise). For the more complex sites with multiple prospect types, we add custom logic to deduce a prospect type based on the types of pages the user is viewing or actions they are taking during their session. For example, in a two-sided marketplace, a user visiting www might be treated as a nonspecific-prospect at first but then, when they access a seller specific page, they should be treated as a “seller prospect” from then on, even if they return to a more generic page after that.
Remember that session-scoped custom dimensions apply the last value received in the session to the entire session. The key in preserving the prospect status for prospect sessions that convert to become members is in NOT overwriting the custom dimension any time after the user is determined to be a member in their signup session. In this solution, when returning members access the membership area after first going to the marketing site, they might be treated as a candidate for their first few hits if they are on a new device. But as soon as they reach the membership area and are determined not to be a new signup, the custom dimension would be overwritten, identifying them as a returning member and that is the value that will stick for the session.
Whether your spread of audience types is simply “prospect” and “returning-member” — or maybe it is “seller-prospect,” “buyer-prospect,”,”seller-trial-member,” “buyer-trial-member,” “seller-subscriber-member,” “buyer-subscriber-member” — imagine what you could achieve if you could isolate each in a Google Analytics view and report the mix of these on a trendline!
Contact us to tailor a solution for your business. Empirical Path is a digital measurement consulting firm that partners with Google Analytics, Adobe Analytics, product analytics tools and CDPs to help clients better track and improve their online acquisition and user experience. We’ve helped membership clients with SaaS (Bill.com, Egencia), publishing (Business Insider, The Motley Fool) and marketplace (Roadie, Vroom) models.
how we can help you.