No matter how well data is presented, there will always be room for people to doubt it due to their particular perspective or opinion. There’s an old saying: “Lies, damned lies and statistics” and it has never been truer now that we’re surrounded by clouds of data every day in our work and home lives.
Whether you’re wanting to build a website and include statistical Infographics or you’re an academic researcher or perhaps an accountant for a large corporation, your challenges will contain some similar elements.
For example, imagine that you’re a government employee tasked with presenting figures on how many people of ethnic background were employed in various industries across different sectors. You might discover that 60% of catering & hospitality jobs in lower-paid positions in the Southern states of the USA were taken by African Americans, but when you publish those figures, someone is quick to point out that around 56% of the entire population of African Americans in the 48 contiguous states live south of the old Mason-Dixon line.
So the proportionality of your figures is already skewed, because over half of the African American population live in those states you’re surveying. In order to correct such statistical ambiguities, there are many mathematical and scientific strategies that can be applied to ‘normalize’ percentages, but in effect, your figures are already effectively inaccurate before you’ve even started a meaningful analysis. Let’s look at some more of the problems involved in presenting data for marketers and others.
Top 7 Problems Involved in Presenting Data for Marketers
Some of the most significant challenges for accurate data presentation include:
1. Data Security: Safeguarding sensitive business data whilst it’s being used for analysis is essential. Most researchers or employees handling sensitive nature data, such as the ‘special categories’ under the European General Data Protection Regulations (GDPR) would do well to use at least a free VPN as a matter of course during their online research and activities.
Ensuring that only authorized personnel from given IP addresses have access to specific data sets and reports can be problematic, especially when dealing with very big quantities of data, so protection against hackers should be foremost in data handlers’ minds.
2. Data Governance: Instigating and policing data governance policies to keep compliance with political and legislative norms can be complex. In effect, balancing data accessibility (i.e. who can find out what from where) with privacy and security is a tightrope walk. If you present figures, it’s essential to say where they were sourced and who harvested them. Bias appears in all industries, so it’s important to be transparent.
3. Data Integration: Governments, corporations and businesses often have data stored in various different formats and systems. Integrating data from these various sources, even when using artificial intelligence (AI) to place data into a single reporting format can be technically challenging, but they can be faced by using effective ETL (Extract, Transform & Load) processes.
4. Data Accuracy and Quality: Ensuring that any data being presented is consistent, accurate, and trustworthy across any system is a regular hurdle. Inconsistent or inaccurate data can lead to poor business decisions and even catastrophic outcomes for individuals.
5. Data Visualization: This is especially important for marketers analyzing such data as, for example, Facebook advertisement metrics. Furthermore, there are now software platforms emerging that can effectively analyze ‘what if’ scenarios from spreadsheets, without harming or altering the original data in the source spreadsheets themselves.
For example, imagine that you’re a salesperson for a Venture Capital (VC) investment house. You’re given a spreadsheet with some predictions for profit and loss over the first three years of a start-up company, and you create some graphs and infographics to drop into PowerPoint slides, or perhaps publish them on a WP website.
You may be asked to source data from a content management system, then, during a presentation, perhaps change the possible outcome of the graphs by reducing the profit figures in year 2, but by doing so, you either forget the original figures in the spreadsheet or omit to change them back after the presentation is over.
The data used for the pitch is then effectively ‘broken’ - with potentially disastrous consequences for future financial reporting. But nowadays, software packages are available that can change the look of graphics without having to alter the source data itself. In this way, none-financial employees and marketers can play in a data ‘sandbox’ without compromising security or accuracy of the source material.
6. Cultural Diversity: Different departments and organizations may have wide-ranging priorities in data presentation and aligning these with security (so that sufficient people have access to the source figures), and a unified reporting framework needs technology compatibility to make it all fit together. Again, this is where marketers can alter the way that data is presented with a couple of clicks of a mouse using appropriate technology.
7. Cost of Resources: Keeping an enterprise-wide reporting system can be expensive, both in terms of salaries and technology updates. Again, a cost-effective web reporting platform can keep these costs down whilst maintaining data accuracy.
In summary, the use of basic security techniques such as a VPN and ensuring bias-free and up-to-date data is essential in any situation where marketing metrics are crucial. Adherence to these basic rules need not be too costly, providing that the right technology is employed, and that the workforce is properly trained on how to use it.
COMMENTS