When navigating the vast sea of technical information, credibility is the compass. As a Senior IT Consultant, I have always emphasized the importance of basing every statement on concrete, verifiable data. This weekly review is not just a rundown of news, but an opportunity to reflect on a crucial aspect of technical communication: the responsible use of statistics. I conducted an internal review of four recent articles, focusing specifically on the statistics and percentages used. The objective was clear: identify and, if necessary, correct all numerical claims lacking an official and verifiable source. The result of this analysis highlighted the need for greater rigor, a lesson I share to strengthen the quality and reliability of the content we provide. This is not just about accuracy, but about trust: IT professionals expect precision and rigor, and providing unsupported data undermines the credibility of the entire message.
Tested on: Ubuntu 24.04 LTS · N/A · August 2026
Prerequisites for Data Validation
Before integrating any statistic into an analysis or technical article, it is fundamental to establish validation criteria. This is not an academic exercise, but an operational practice that prevents the spread of misinformation and poor decision-making in critical environments. Fundamental prerequisites include:
- Official Source: The statistic must come from a recognized and authoritative organization in the sector (e.g., CISA, NIST, Gartner, IDC, official vendor documentation). Read also: NIS2 Compliance: 6-Month Review, Operational Insights
- Contextual Relevance: The data must be pertinent to the specific context of the article and not a generalization. A statistic on cloud security for SMBs might not be applicable to an enterprise on-premise infrastructure.
- Timeliness: The statistic must be current. Data in the IT sector ages quickly. A 5-year-old data point could be completely obsolete today, especially in areas like cybersecurity or cloud.
- Transparent Methodology: If possible, the source should explicitly state the methodology for data collection and analysis, to understand its limitations and applicability.
The Article Review Process
The review process I applied to the four articles followed a systematic approach. For every statement containing specific numbers, percentages, or metrics, I actively sought the original source. If the source was not cited or, once found, was not considered sufficiently authoritative or verifiable, the data was removed or reformulated in qualitative terms. This is a crucial step to ensure that the content not only informs but is also reliable.
Consider an example. If an article stated that “70% of ransomware attacks start with phishing emails,” I verified if this percentage was supported by a security report from a government agency (such as the Cybersecurity and Infrastructure Security Agency) or a leading research firm. Without such support, the statement was modified to “ransomware attacks often use phishing as an initial vector.” Read also: Active Directory: 5 Anti-Breach Controls
# Pseudocode example for statistic validation
function validate_statistic(statement):
if contains_numeric_data(statement):
source = find_cited_source(statement)
if not is_authoritative(source) or not is_recent(source):
log_warning(f"Unvalidated statistic: {statement}")
return "REMOVE_OR_REFORMULATE"
else:
return "VALID"
return "N/A"
# Application to article text
article_text = [
"70% of ransomware attacks start with phishing emails.",
"The new NIS2 directives strengthen IT security."
]
for sentence in article_text:
status = validate_statistic(sentence)
if status == "REMOVE_OR_REFORMULATE":
print(f"Action required for: '{sentence}'")This rigorous approach is essential for anyone operating in IT, where decisions have a direct impact on security and operational continuity. Read also: Disaster Recovery Testing: What Broke First
Remove or Reformulate: A Practical Case
During the review, one article contained the statement “an average company suffers X DDoS attacks per month, with an average cost of Y euros for each hour of downtime.” Without a specific and verifiable source for X and Y, these numbers were removed. In some cases, it was possible to replace the statistic with a qualitative reference, for example: “companies are frequently targeted by DDoS attacks, which can cause significant disruptions and high costs.”
Another example concerned the effectiveness of a particular security solution. A statement like “tool Z significantly reduces false positives” without a case study, an independent benchmark, or a vendor report attesting to its measured effectiveness in a controlled context, was considered invalid. In these cases, it is preferable to describe the tool’s functionalities and expected benefits, rather than quantifying an impact without proof. Read also: Backup Tools: Restic vs BorgBackup on 2TB
# Example of reformulating unsupported data
def clean_statement(statement, original_data, replacement_text):
if original_data in statement:
return statement.replace(original_data, replacement_text)
return statement
original_statement = "An average company suffers 5 DDoS attacks per month, with an average cost of 10,000 euros for each hour of downtime."
cleaned_statement = clean_statement(original_statement,
"5 DDoS attacks per month, with an average cost of 10,000 euros for each hour of downtime.",
"frequent DDoS attacks that can cause significant disruptions and high costs.")
print(cleaned_statement)
# Output: An average company suffers frequent DDoS attacks that can cause significant disruptions and high costs.Common Errors and Troubleshooting in Using Statistics
The most common error is the temptation to use “impressive” numbers to lend weight to an argument, without worrying about their origin. Other errors include:
- Over-generalization: Using global statistics for specific contexts (e.g., US statistics for the European market).
- Cherry-picking: Selecting only data that supports one’s thesis, ignoring contradictory data.
- Lack of context: Presenting a number without explaining what it represents, how it was collected, or its limitations.
To avoid these problems, it is good practice to adopt a critical approach to all numerical data. When reading or writing, the question “who said it and how do they know?” should always be present. Read also: IT Handover Checklist: Essential System Transition
FAQ — Frequently Asked Questions
Why is the source so important for an IT statistic?
The source provides authority and verifiability to the data. In the IT sector, where decisions can have significant impacts on security and operational continuity, relying on unvalidated information is an unacceptable risk. A credible source allows readers to delve deeper and confirm the information, building trust in the content.
What is considered an “authoritative source” in the IT context?
An authoritative source is an organization, government agency, research institution, or recognized vendor that produces data based on solid and transparent methodologies. Examples include CISA, ENISA, NIST, Gartner, Forrester, annual reports from major security companies, or official documentation for a software/hardware product.
Can I use statistics from blogs or third-party articles?
Yes, but only if the blog or article itself cites an authoritative primary source. It is always preferable to go back to the original source to verify the context and methodology. Citing a blog that cites another blog that in turn cites a report introduces too many steps and potential errors.
If I can’t find a source, is it better not to use the statistic?
Absolutely yes. If in doubt about the origin or validity of a statistic, it is always better to omit it or reformulate it in qualitative terms. Precision and credibility are more important than a potentially misleading number. Better zero statistics than an invented or unverifiable one.
Conclusions with Operational Takeaways
The article review reaffirmed a fundamental principle: data integrity is non-negotiable in technical communication. For IT professionals, this translates into clear operational takeaways:
- Cross-verification: Before citing a statistic, seek at least two independent sources that confirm it.
- Transparency: If you use data, always cite the source with a direct link.
- Context: Always explain the context of the data. What is the sample? When was it collected? What are its limitations?
- Prioritize Quality: A well-argued qualitative statement is better than a spectacular but unfounded number.
Adopting this approach not only improves content quality but also strengthens professional reputation. Trust is built on precision.
Sources
Updated: August 2026
