On a quest to find the Mountains of Kong
The Mountains of Kong have been described as a giant, impassible mountain range, stretching for 100 miles along the 10th parallel from eastern Senegal across to Nigeria.
They first appeared on a map in 1798 by the English cartographer James Rennell. Rennell was one of the most highly respected cartographers of his time and he depicted the topography of the region based on the surveys conducted by the Scottish explorer Mungo Park, who talked of a “belt of mountains, which extends west to east […] and occupies the parallels between ten and eleven degrees”.
Rennell was an expert. A revered figure in contemporary map-making. No one questioned Rennell’s maps because there was no reason to. As a founder of the Royal Geographical Society and one of the developers of oceanography, there was no need. He was unquestionable. Which was a shame, because the Mountains of Kong do not exist.
For over 100 years, maps of Africa featured Rennell and Park’s fabled mountains, which were eventually joined with the equally fictitious Mountains of the Moon.
It wasn’t until 1887 and the French explorer Louis-Gustave Binger, who was searching for the source of the river Niger that the myth of the Mountains of Kong was finally debunked.
So how does a myth endure for nearly a century?
To put it simply, the myth endured because no one challenged it. There are even accounts from visitors to the region who claimed to have seen a giant and entirely fictitious mountain range. Rennell’s initial hypothesis was taken as gospel – because he was the expert and the myth was handed down from cartographer to cartographer until the myth became a “fact”.
Our quest for the Mountains of Kong
We’re very lucky to be able to represent several design and development agencies as their analytics and reporting partner or white label service. One of these agencies called us about a problem that one of their clients was having.
The client is a national business who relies on their ecommerce performance. To them, small drops in performance and conversion rate, equal millions of pounds. During a critical period in their year, their site’s sales performance dropped dramatically.
Internally, they have an industry leading team of digital marketers, led by Marketing Director who is an industry ‘heavyweight’. They are supported by one of the largest and best-known marketing agencies in the UK. Together, they had looked at the analytics data and decided that there was clearly a fault with the website. They were the experts. They knew and they weren’t to be questioned.
Staring at potentially heavy performance penalties for a faulty website and convinced that their work wasn’t at fault, our partner took a gamble. They engaged us to find the error on the website, understanding that our report may find them to be at fault. They were also pretty certain that if it wasn’t their issue, they’d still have an unsatisfied client. After all, who challenges the experts?
We spent a week taking apart every area of the website. We looked at every update that was made to the website, every release by our client, every code update, every change request and every bug fix. We spoke to third party payment providers and security partners and logged everything they did which would affect the website too.
We found nothing technical. We found nothing that could be linked to the website at all. We couldn’t find the fault that a highly qualified, hugely experienced and exceptionally knowledgeable team told us we’d find. Until we looked at the CMS logs. There, we weren’t able to find just a slight correlation between the performance drop and a change. We found the precise change, the exact time of day it was made and the team member who made it.
We discovered that their sales performance was intrinsically linked to their homepage performance. The changes that they made to the homepage on that day, at that time slashed their homepage conversion performance and therefore damaged their sales performance.
There were no Mountains of Kong.
What was the result?
That website now has a structured A/B split testing program in place, which ensures that these problems won’t happen again. They have a ‘backbone’ homepage created using analytics data. Any changes that are pushed live via the CMS are published to a control version of the homepage. Changes that improve conversion performance are rolled in to the ‘backbone’ page and everyone learns from the things that failed.
The lesson from all of this? Don’t be afraid to challenge the experts. There can be fictitious mountain ranges anywhere and they need to be conquered.