Revisiting the Funnel – A Practical Application
Applying the funnel to an American airline that uses paid search to drive demand to its booking engine across Asia.
Applying the funnel to an American airline that uses paid search to drive demand to its booking engine across Asia.
In my previous column, we looked at the all encompassing funnel and the pertinent metrics that one should consider while passing each milestone in the funnel. This column focuses on a practical and relevant example of how having pertinent metrics along the journey helps and how to leverage those parameters to identify critical aberrations.
Our focus here is a leading American airline brand that’s been leveraging paid search marketing extensively to drive demand to the online booking engine across Asian markets. Needless to say, a robust external and internal analytics system has been critical in extracting the best possible ROI out of such a large-scale programme.
One day, we noticed that the efficiency in a particular Asian market was declining and when historical data was quickly gleaned, the rate of revenue decline was of such alarming nature that a sudden dash to the consoles was the inevitable outcome.
Creating Conversion Funnels
We created a conversion funnel trending using analytics tools that highlighted the scale of decline (as indicated by the graphic below).
At this juncture, I urge you to remember the metrics from my last column that included clicks to visitors, visitors to lookers, lookers to bookers, et al, as these are essentially the trigger points at each transition point of the visitor journey that trigger alarms, should anything be amiss.
Back to the scenario, once this drastic fall in revenue and bookings were noted, the next step was to examine the key transition points in the funnel to understand whether there was any abnormal attrition that was taking place, and if so, what could the reason be for such attrition. The following graphic shows the funnel trending across two consecutive months and you are urged to note the change in attrition levels across key transition points.
Examining Conversion Funnels
It is not a daunting task to examine the transition points from the moment a click becomes a visitor on the website and goes on to become a final conversion or otherwise. Therefore, it was noted without much effort that visitors who were landing on the site during month two (funnel two) were mysteriously not conducting ‘flight searches’ (rather, there was a huge drop in the number of visitors conducting flight searches). Due to this huge attrition point, a very low volume of visitors was arriving at the cart build up stage and hence revenues and tickets were declining at a drastic pace.
What has caused such an aberration in a website that would otherwise be home to proliferating, frenetic flight search activity? Is there something that was not allowing users to search for flights or was it something else?
Our team tried to identify the various factors that could culminate in this unlikely scenario. Elimination of the usual suspects was methodically executed given that there were no fluctuations in pricing, product, service, or any other internal or external factor that could potentially affect in such a fashion. The only stone that was left unturned was the ‘user experience’ within the website. Had anything changed?
Striking at the Root Cause
The airline website was subjected to a ‘live booking test’ and voila… there lay the troublemaker that was causing huge attrition of visitors who entered the site without engaging in flight search!
Since this problem occurred in a local language Asian website, it did not take long to dig out the fact that the ‘From’ and ‘To’ fields of the airline website was not recognising the ‘typed in’ characters which made up the cities and was noted to be giving consistent error messages. The only way to enter the flight search was to use a separate drop down to select airport codes for the ‘From’ and ‘To’ points and this drop down was not clearly discernible as well.
Suffice to say, it took a simple step to quickly resolve the issue – to include a message directing users to choose the drop down while the problem was being sorted at the back end. This circumvented the problem situation to a large extent and the original situation was restored in a short time frame thus rounding off the whole incident in a very pleasant note. The valuable takeaway points in the direction of having robust tracking and analytics systems powering your online demand generation programmes – there is no substitute to rigorous measurement and tracking as we have seen time and again, and not investing in such a system might save you a few dollars in the short term but the missed opportunities will outweigh this saving by a huge margin in the long run.
As we come to the close of yet another column, let me urge you once again to internalise the importance of the funnel and more importantly, the analytics system that tracks the funnel.
There are many brownie points to be won if you are a ‘funnel tracker’!