Decision making with data

Decision making with data

Too often we hear stories about how the “data” used to make decisions on performance digital strategy led to companies suffering in the aftermath, but clueless to the cause.  Companies rely on terabytes of data, smart analytics and smart people to justify their decisions. So how can they get it so wrong?          

The answer scarily simple: the data collected and assessed was not set up to answer the right questions. Evaluating and recommending the proper questions to ask when making decision based on your data is critical.

The Three Pillars of Data Analysis:

1.Answering the right question

Even when companies collect mass amounts of data, it may still not be enough to substantiate actions, because the data isn’t addressing the right questions. 

Questions to consider:

  • Why are SEM acquisition costs up in the first place?
  • Is regional targeting in SEM connecting with the wrong audience?
  • Are there ballooning SEM overheads in affected regions due to spend on the wrong digital assets (i.e. high cost, low performance terms)?
  • Have there been major changes to brand image and advertising in that region?
  • Have new competitors, resellers or affiliates entered this space and are they cannibalising search volume?
  • What do the consumers ultimately want and what online and offline vehicles do they use to get it

Jumping straight to the solution without first understanding the right question can be a drain on both resource and finance for organisations. These errors are avoidable if organisations spend 70% of the time making sure the right question is asked and understood, and 30% of the time gathering and interpreting the results – and not vice versa.

2.Understanding data interconnectivity

As many attribution experts will say, looking at channels in silos and not understanding how they interrelate and influence each other is a sure fire way to miss a large part of the performance picture.

Questions to consider:

  • Are the regions with the highest acquisition costs contributing to assisted sales in other geos?
  • Does offline in targeted regions affect online performance in others?
  • Do paid channels in target area curb competitor market share growth and consumer perception of brand nationally?
  • Are consumers in target region simply responding poorer to invested channels and conduct purchases via other channels? (i.e. prefer social instead of SEM)
  • Is core consumer behaviour different in this new multiscreen world? (I.e. consumers shifting from desktop to predominately mobile, while only the former is being targeted)

3.Understanding the “X” factors

Beyond attribution and addressing “why” when conducting data analysis, there are a few “X’ factors that can be immensely valuable in making smart business decisions. These are those variables pertaining to quality and reliability of the data.

Questions to consider:

  •  Can we trust the technology that was used to collect the data? Is it capturing all the right metrics?
  • Were data points tagged and constructed to measure the right variables? (I.e. if you are measuring completed conversions and are tagged to only measure starting points, then you are essentially equating apples to oranges.)
  • Was the data collected seasonally constricted to a low performance timeframe?
  • Are the products or services sold in growth, maturity or decline?
  • Does the product messaging online need a refresh? Have different messaging even been tested?

The difference between the next board meeting being one of frustration and one of jubilation is a more robust and well-constructed analysis methodology. The pillars above each contribute different elements towards our big picture and are a foundation to creating business heroes.

Follow the fundamental principles laid out above and ask the right questions because the right decisions based on the right data is what drives the right results.

Zeeshan Javed is business director at iProspect

Please login with linkedin to comment

Latest News