What Does it Mean to be ‘Marketing-Minded’?
Despite the desire for objectivity given all of the facts we work with; there are processes in data analytics that cannot escape some degree of subjectivity on the analyst’s part. As data analysts, it is our job is to collect meaningful data, monitor it for trends in the agreed upon KPI’s and interpret our findings to stakeholders. In carrying out this analysis, we have to make subjective decisions. Even though analysts work with facts, numbers and data, there will always be an important strong interpretive component to data analytics.
To some extent, an analyst’s interpretive processes can make or break the performance of their work. It’s what makes each analyst different from the next. Therefore, data analysts need to be extremely cautious, because all humans are very much susceptible to cognitive biases. In marketing and data analytics, these cognitive biases can wreak havoc on the performance of an account. Making business decisions based on pre-existing beliefs, limited data or just irrational preferences will quickly derail your campaigns and whittle away your profits.
Below are four of the most common types of cognitive biases and how to keep an eye out for them in your own thinking:
Survivorship Bias: Looking only at the data that’s in front of you without considering the data that is not. Working with incomplete data can dangerously skew your findings and lead your entire strategy astray. In the marketing and data analytics world we see this most commonly happening through the use of filters, segments and templated market insights. Sometimes it pays to strip away all the scaffolding and world to capture as un-altered data as possible.
Sunk Cost Fallacy: This may be the easiest cognitive trap to fall into and unfortunately it can also be one of the most disastrous. Everyone has been in a situation where they ended up wasting more time because they were trying to salvage the time they had already invested. In marketing, this can often happen when investments have been made into promoting a new product/service/ etc. and the results don’t match the forecasted performance. It can be difficult to know how long to keep testing the market and when to keep making pivots in an attempt to salvage the venture. In our experience, measured A/B testing is the best risk management technique to safeguard against the sunk cost fallacy.
False Causality: Correlation does not imply causation. The human brain is wired to see patterns – even when none exist. In marketing and data analytics, this can be witnessed in situations where data is explained away as being ‘seasonal’, ’caused by a particular trend on social media’. The next time you’re examining an acquisition chart and see two variables moving in tandem – remember that this does not necessarily mean the one-caused-the-other.
Confirmation Bias: “If you torture the data long enough, it will confess”. The nightmare of statistics and data analysis is that with enough willpower, you can force the numbers to support whatever preconceived notion you’re trying to prove. We see this – all the time in marketing analytics. Presuppositions about targeting audiences made up of people belonging to particular income brackets, genders, interest-areas etc. being used as the foundation of entire marketing strategies. This is one of the reasons why, at PINTAYA, we insist on continual unabridged data collection and constant hypothesis analysis. Thankfully, stakeholders are waking up to confirmation bias in all sorts of contexts and it’s a good thing too! It’s a troublesome cognitive bias to get trapped in.