Profile
Data-centric digital marketing generalist.
Experienced in web and app product analytics, data visualisation, conversion rate optimisation, SEO, SEM, marketing analytics.
Domain
Digital Healthcare
Technical Skills
GA, SEO, SEM, GTM, Search Console
Looker Studio, Tableau
GCP - Fuction | Scheduler | BigQuery
Python, SQL
Wordpress
View My LinkedIn Profile
Description: This illustration shows ways to pick out the best performers in different scenarios and would result in different priorities. One way to perceive conversion funnel optimisation is to pick the most efficient step at each step to maximise the outcome, for example, as simple as allocating marketing budget based on conversion rate in leads would have a better multiplying effect. Extending the concept to the whole funnel such as prioritising product optimisation for the highest revenue product or picking out higher potentials for SEM based on both conversion rate AND search impressions gives better total leverage.
Here we try to pick out best performing landing page with the highest conversion rate and total conversions for organic traffic. The conversion rate we are using here is conversions/sessions
. We see organically Dimension B in pink and brown dominates the better blocks of outcome.
Now let’s focus on the brown inverted triangle on the top-right. Say for the next step we want to pick out landing/products with the highest potential for SEM, the inverted triangle would logically be one of our top choices.
However if we add SERP data, the inverted triangle is a lot less sexy with its low search volumes and (ironically) better SEO with 38% of the maximum traffic we’ve already taken.
This would lead us to simply focus on impressions and conversion rate since it potentially gives a greater amount of leads. Here we see many of the landing pages in the best block (top-right) is having a low sessions-organic ratio, which is great for SEM since we want to hijack the rank in SERP.
The data we use here are on leads only, but we can mimic a scenario includes the price of the product by assigning a random value to each data point. With price, we can generate a metric as maximum expected revenue with price * Impressions * conversion rate
. This figure for sure will never be achieved as we need to work on the CTR on SERP with SEM but it should be enough at this stage.