Lies, Damn Lies, and Statistics. The Myth of Following the Data

People have shown increasing interest in big data across the entire world. Businesses rely on the data harvested by tech heavyweights such as Facebook to reach out to their customers and make important decisions in the corporate segment.

Data is the new oil due to the massive value that it has gained over the past few years. However, it is paramount to highlight that relying on big data as the ultimate source of truth might be a dangerous assumption because it may be misleading.

While big data can be helpful, its users should note that in-depth knowledge of big data analytics, the factors that restrict its reliability, and sound business experience are critical to deriving meaningful and correct conclusions from it.

Although there are lots of manipulations in advertising, retail media practices could be quite transparent. You can easily track how various indicators work and analyse the results. Moreover, retail media ads allow you to observe your customers’ behavior and make predictions based on your research. Stay tuned to find out how it works.

There are lies, damned lies and statistics

In the words of the world-acclaimed British statesman author Benjamin Disraeli, “There are lies, damned lies, and statistics.” The sudden popularity of big data has resulted in a scenario whereby conclusions purported to have been drawn from big data go unchecked and unquestioned in attempts to influence consumer or business behavior. It is crucial to recognize that sometimes statistics lie, and liars use statistics.

Conclusions drawn from big data are not invincible as they can be adulterated by several factors. What most people forget is that the entire big data chain is vital in yielding the correct output. It includes how that data is gathered, analyzed, and, eventually, summarized.

Due diligence is usually lacking in some of the processes that big data goes through before conclusions are made, making the entire effort prone to erroneous results. It is further worsened by the corporate culture in some companies that encourage fast-tracking developments and discouraging questions. It is critical to point out that gathering numbers is crucial, but ensuring that they are interpreted and reported accurately is more important.

Statistics lie and liars use statistics. Does it happen in retail media?

Can you think of a better way to manipulate consumer behavior than a bold proclamation that ‘Statistics show that 90% of the population use such a product daily’? Such a statement has the effect of drawing people towards that product. We should always be wary of the fact that statistics lie and liars use statistics.

Fortunately, retail media advertising allows for a much more significant degree of transparency. While old-school advertisers struggle to drive conversions via making the most out of trivial marketing channels, retail media experts apply more forward-looking techniques.

Let’s look at popular ecommerce marketplaces, such as Walmart and Amazon. They have managed to use closed-loop attribution, integrating sales with ad engagements. In other words, the company is simultaneously running ads and selling advertised items.

Moreover, wise data management allows the big ecommerce platforms to target the most interested customers. Modern tools successfully decipher the consumers’ intentions and predict their activity based on the deep analysis of their previous actions.

Big data is what attribution in retail media is all about. For instance, you can considerably improve your ad revenue strategy by analysing your metrics by sum, percentage change, and other indicators to get a clear picture of the campaign’s success.

It’s essential to understand how different providers measure attribution. For example, Google Ads now allows you to measure the impact of the ads in Display, YouTube, and Search and see how they correlate with conversions. You can also compare different attribution models and see which one works best for your business.

So it is paramount to critique conclusions or proclamations purported to have been derived from big data. Numerous reasons can result in erroneous conclusions from big data, and these should be assessed. These are described briefly below.

False or missing source of data

No doubt, you cannot expect to have correct conclusions if the origin of the data is false. There are numerous statements today made from big data that would have been harvested from the wrong sources.

Software errors

There is also the possibility that some of the software used to collect, filter, and store data could have been erroneous. We have had numerous reports of big technology companies coming out to say that they are patching up areas where their software would have been compromised. It is evidence that software is not foolproof.

Errors in data correction, data enrichment, and algorithms

In data analysis, there is the practice of leaving out the results that do not fit in the normal range. These are known as outliers, and the assumption is that there will be something wrong with them. This kind of data tempering further lessens the reliability of big data. We should also be wary that the algorithms that perform analysis and interpretation might be erroneous.

Tips to prevent damn lies statistics

During this age, it is easy to fall victim to damn lie statistics. It is essential to be fully aware of the tips to avoid them. This involves striking the much-needed balance in critical view and a practical approach. You also need to have sound knowledge of both business operations and data analytics.

To avoid damn lies statistics, it is paramount to verify the sources of your data. It is key as every other process is dependent on it. You must ensure that it adheres to the best data gathering practices. If you are using research, you must use it well, factoring in all the built-in slants.


There is no doubt that big data is useful. But it is not invincible. Conclusions drawn from it can be erroneous and misleading. With this in mind, it is paramount to ensure that you are aware of all the pitfalls when you depend on them to make crucial decisions.

Make the most of retail media advertising and use closed-loop attribution to good advantage.