• Navid Nassiri

The role of data automation in subscription marketing


The death of third-party cookies is forcing companies to overhaul their approach to marketing and attribution spend. With the loss of cookie-based attribution, many are relying on a small subset of data to make decisions about a much larger dataset.


But basing decisions on a shrinking number of traceable users, rather than considering their audience as a whole, can lead to the over-amplification of certain signals - not least in subscription marketing.


In this article, we explore how content marketers are using alternative methods to gain a holistic view of their customers and explain how data automation can help achieve this.


Gaining a more holistic view of the marketing model


As one of our customers, a global publisher, can attest, they’re beginning to think much more holistically about the impact of their daily operations on audience engagement. For instance, it’s not just direct marketing activities that drive publication subscriptions; it’s factors such as the design of their website, the quality of their content, and organic growth.


In fact, this customer found that 80% of their marketing spend was on branding, with very little investment required in subscriber conversion. So, linking every penny of marketing spend to final acquisition produced a flawed view of their customer journey.


Getting to know personas (while respecting privacy)


Instead of relying on individual user data, our customer decided to aggregate users based on eight common personas, which they had identified as constituting their subscriber base. They then tracked each cohort’s user journey rather than the individual’s.


Working backwards from a target number of new subscribers at the bottom of the funnel, they estimated how many users showing intent signals (searching for the publication, registering on its website, or landing on product offer pages) would be required to achieve this level of conversion.


Next, they calculated the scale of habitual website visitors within which there would be the necessary number of customers with intent. Then the number of casual website visitors, and finally, the number of potential professional readers who would need to be reached by a paid media campaign.


Verifying your audience


Once the funnel model has been established, you need to verify your audience by collecting demographic data about your customer base to work out whether they fit your targeted cohort. For example, use website surveys to gate content, or a form to gather demographic data about new subscribers.


Next, determine the accuracy of that particular funnel model for that cohort. If the accuracy is too low, examine which aspects need to change. For example, optimizing your website to engage users for longer, or pushing more relevant content through your social channels.


It’s important to use the right metric for each stage of each funnel. If you only know the conversion goal is missed at the time it is missed, it’s too late to correct your model, so find appropriate indicators to forecast final conversion numbers. For example, the number of quality website visitors might cause conversion numbers to fall in three months’ time. This level of visibility flags problems before they occur.


The importance of data fidelity


As your audience moves further down the funnel, the fidelity of your demographic data increases. For example, a relatively anonymous user from social media visits the website and provides further information, which allows you to determine their profile by iteratively adding new first-party data points at each stage.


The challenge with developing your own funnel model is disparate channel data. For example, you may not know how well paid media spend is performing according to your goals. Since there is no universal cookie replacement, you can only measure how particular cohorts engage with particular parts of the funnel.


So what does all of this have to do with data automation?


How data automation can help


While the lack of third-party cookies means it’s harder to follow complete customer journeys, we can use data to help make multiple partial joins. For example, we can measure the click attribution of the professional reader audience who follow a paid media segment to a website, then compare this to other audiences.


Switchboard helps create data pipelines and modeling to produce a centralized data asset in the customer’s chosen data warehouse. This can include affinity-based data from social media platforms, which is not linked to individuals, and can be aggregated for better analysis.


By measuring potential customers based on their engagement, and using data automation to help with analysis, you can gain higher-quality leads, grow your customer base, and better understand how your subscription marketing budget works. Get in touch to find out more.