Basic issue and ‘‘wrong‘‘ thesis‘:
The question is ‘‘why is our mobile conversion and the absolute amount of transactions (conversion) so low?‘’
A lot of answers could have been correct. An agency told the customer ’’The products are purchased via PC (desktop) instead of purchasing them mobile. Another answer contained the idea to change the buying process. However, analysis of purchase/transaction funnels accounted that low conversion rates are not an issue of the shopping cart process. Of course you may try to optimize the diverse Micro- Conversions, but in order to be able to accomplish this web analytical challenge, you need to provide further analysis.
The following description explains how you may build a web analytic Analysis in order to receive insight and subsequently infer appropriate guidance which affects the increase of the absolute conversion positively.
Finding the root of a problem – Bounce Rate and average loading time
Step 1: create a personal report (inside of Google Analytics)
If it is known that the conversion rate of a terminal is lower than the rate of another device, it would be worthwhile to set a comparison report. Due to the fact that the customer uses the Tracking-Tool Google Analytics, a new report has been compiled in ’’customization’’.
This report displays what is already known: the conversion rate of mobile devices (Dimension: Device Category) is considerably lower than requests via desktops (Metric: Bounce Rate).
The report showed though, that – inside a space of time- the bounce rate of mobile devices is higher at a similar amount of entrances (Metric: entrances).
In order to analyze if the bounce rate was low over a several time space like weeks/ months or if it decreased due to a change of setting on the website, the dimension ‘’week of year’’ was added to the reporting.
Considering the fact that there were merely any fluctuations at the bounce rate in the temporary evolution (except after the display campaign which affected the desktop performance negatively and therefore have not been considered) it was concluded that it was a permanent issue and not contingent on a technical adjustment.
Step 2: Expand your report – experiments are not amiss!
The bounce rate of mobile devices right after their entrance was higher and the drop outs were impinging on the conversion rate and therefore influenced the amount of transactions.
Now the question occurred ‘‘Why is it this high’’? After some time of experimenting (testing of diverse dimensions and metrics) it unfolded that the average loading time for mobile devices was much higher than for desktops.
Via the download function of Google Analytics, an Excel Sheet containing a scatter diagram which visualizes the dependence of the metrics bounce rate and loading time.
Another graphical display of the conversion rate depending on the loading time in seconds (x-axis) shows again the issue of a slow loading time and the effect of bouncers on the amount of transactions and conversion rate.
The longer the loading time is, the lower the conversion rate. The convergence behavior approaching zero shows the enormous importance of the loading time regarding the sales performance.
A further analysis showed that very fast loading times – less than 2 seconds- also have a low conversion rate, regarding the fact that those pages have wrong/ bad content like category/information pages.
Step 3: Further classification is not amiss– keep on experimenting!
The bounce rate for mobile devices is higher than for desktops, the reason is the long loading time. Does it include mobile phones? The extension of reporting called dimension ’’Mobile Device Marketing Name’’ shows the designation (name) of the mobile phones may show you if any mobile devices has problems loading contents or elements of the website ’’fast’’.
If you want to connect exotic or mobile devices, which are rarely used by visitors and do not start sessions, you may use the extended filtering of Google Analytics.
Only those marks will be displayed that triggered at least 1000 entrances in the selected time space. (This amount was useful for this case, though you may use another lower limit for another customer or for another time space).
Via the download function of Google Analytics, you may visualize a bar chart (entrances) and a line chart (bounce rate) which manufacturer/brands of mobile devices show issues with loading time and as a result with bounce rates.
This kind of analysis showed the customer that certain operation system do have issues loading the website in a proper time which results in a high drop rate.
After a technical adjustment and improvements, the conversion rate of mobile devices rose 1.25% for this customer.
Note: This is just a section of the complete analysis, of course more dimensions, metrics and visualisations were created. However, this brief introduction is supposed to show the effectiveness of even small and easy analysis inside of web analysis.