Common Analytics Mistakes
In order to make precise decisions, success can only be measured using accurate methodologies. However, Australian businesses commonly make analytics mistakes in the way they measure their success. We discuss the most common mistakes made by businesses using analytical data.
1. Assumptions off Incomplete Data
Australian businesses collect an awful amount of data. The problem with data collecting too much data is many businesses don’t have adequate resources to properly analyse it all.
As a result, they either cut back on the data collected or make incorrect assumptions on the data that is analysed.
A clients pages shows a very high time on page result.
Does this mean the content on this page is a success? OR Does it mean it is not a success? It could mean the page has a number of issues from various possibilities. However, a wrong assumption could easily be made by not completing a thorough analysis.
Don’t draw conclusions from a piece of data that only presents a portion of the entire story. Accurate data means it has been tested and beyond all reasonable doubt, a conclusion can be made.
2. Lack of Predictive Analytics (foresight)
The most commonly requested data our clients ask for is ‘Customer behaviour’.
Which include: The numbers of users, which button they clicked, what product most looked at, how long they read an article for, which page they left at etc.
The basic goal of any business is to exist and not go bankrupt. Therefore, it is crucial for any businesses to be able to recognise and predict an event that could occur and have the best-planned response from it. This is achieved through predictive analytics.
Predictive analytics is done very well with many large corporations who study the current market with a microscope. When predictive analytics is executed very well, a business can not only be a sideline observer and know when something will occur but also prime and condition a market to actually do what they have prepared for.
Analytics helps us know what future investments will pay the most using a combination of analytical data.
The major component required to achieve this is measuring the customer journey within the market. (Both consumers and competition) Why the customer journey?
The customer service journey is crucial in knowing when, where and how the satisfied a customer is. If at any point throughout the journey a customer feels dissatisfied they will behave accordingly. It is basic human nature behaviour and one business do not pay adequate attention to. People switch on and switch off subconsciously. This data is easily obtainable through predictive analytics.
Focus on customer satisfaction as it is a great tool while analysing predictive analytics. The analysis of this data can provide insights that result in financially beneficial decisions.
3. Focus group testing
The thousand dollar-million-dollar business question is will small panel users in a controlled environment accurately represent your realistically broader target market?
Here is a good example of how people function in a controlled environment.
Have you ever had your blood pressure measured. You can feel the pressure around our arm increase and subconsciously you concentrate on your breathing.
Consequently, your results are skewed by the ‘unnatural’ stresses of the controlled environment.
Focus groups are still useful for observation. However, many businesses use a usability group when they are facing a problem they find from data.
They assume a focus group will help them gather more practical knowledge to find a solution which is ok if the problem really is
The number one failure in using usability group testing is not the ability to resolve a problem but the ability to find the right problem which actually doesn’t require a usability group.
Usability groups can steer a business in a direction learn more about what their customers are trying to accomplish not what the business is trying to accomplish or solve. And this information is often nowhere near a controlled environment.
Both analytics and practical group testing complement each other if the tests are done without customers being too aware they are being observed.
Ask your customers
We need to reiterate how important customer study is. In the beginning, there was the self-taught web designer.
Do you remember when everyone thought they could design the best website?
We all soon realised this was not the case and then came the web designer, after that came the pre-designed web theme, then came the UX designers and now it seems we are in the era of ‘just ask your customer’.
Humans will always do something very well. Complicate something. Simplification for success actually requires a lot of backstepping and thought so we subconsciously complicate wherever we can.
In a time of deafening chatter, content overload it is more important than ever to hear your customers voice because it is the most important one.
All business decisions should stem from customer needs and not what an expert or professional can do with their newly taught skills. Take a look at some of the largest and wealthiest websites. Some are awfully simplistic yet incredibly successful because they achieve exactly what the customers’ needs.
All business decisions must be based on research connected to customer needs.
Analytics Cause and effect:
Even the most intelligent people can’t help but use their gut instinct, intuition, and logic. It seems harmless to do so right? Wrong.
Most businesses opt for this decision making avenue without even knowing it. Why? It is the very human thing to do and most of all it is very easy and it seems very resource efficient but is actually financially costly.
Very intelligent clients we have worked with in addition to experts on their board have made mistakes by confusing the cause of a piece data and how that data correlates to it. It is a very easy mistake to make.
Sydney has experienced a rainy month. We see that a Sydney competitor for our client has had a great sale in gumboots. So it would be right to assume we need to stock up too? The causal factor is not that Sydney has a new fashion trend of gumboots due to the rain. Wrong. The shop was close to a music festival for 2 weeks and the festival goers needed them as the grounds had turned to mud after the first day.
Can you see how it is easy to make a fairly common-sense correlation and be wrong?
A popular honey our client sells wants to make it available in a bulk size. In order to promote it we put it on the homepage with a sale price. The sales skyrocket. The client says ‘ The customers like the new size’ We cannot assume the conversion rate went up because of the product size. The causal factor here could be
1. The homepage promotion banner or
2. The ‘sale price’
You cannot determine which actions correlate without generating and analysing meaningful data to confirm it.
Only then, will you understand why a customer made the decision to make that purchase.