All resources

The knowledge hub

Why Data Breaches Happen and What Your Business Can Learn

Business owner examining exposed customer records and the operational consequences of a data breach.

“Millions of accounts exposed” is the kind of headline that makes you stop scrolling. But what was actually exposed? Passwords? Phone numbers? A recycled list of email addresses? Those details can change the story completely.

If you want to understand why data breaches happen, look past the biggest number. The useful lesson is how the information became exposed, what someone could do with it and which business decision might reduce the same risk for you.

1. Breaches Often Start With an Ordinary Weakness

A stolen password. An application nobody has updated. A sharing link set too broadly. Breaches don’t always begin with an elaborate attack. Sometimes the opening is a routine responsibility that slipped through the cracks, or a supplier account with more access than it needed.

The path often involves more than one weakness. A stolen password may provide entry, overly broad permissions may expose additional records and weak monitoring may delay discovery. Finding the entry point matters, but fixing only that point can leave the wider problem intact.

2. A Customer List Can Be More Valuable Than You Think

Picture the information you use in a normal week: customer contact details, staff records, payment information and account credentials. Personally identifiable information can identify someone on its own or in combination with other data. Your pricing, designs and business plans can also be valuable even when they contain no personal information.

The consequences depend on what was exposed and what someone can do with it. A list of phone numbers may support targeted scams. A set of reusable passwords presents a different account-access risk. Exposed health or identity information can create lasting harm.

3. Read the Number, Then Read the Small Print

Before treating a breach figure as a headcount, check what is being counted. Records, accounts, phone numbers and unique people are different things. Duplicates and overlapping datasets can inflate a total, and investigators may revise an estimate as the evidence develops.

Costs can include technical investigation, operational disruption, customer support, recovery, communications and lost business. An average cost per record from a particular study isn’t a reliable quote for your business. Your dependence on the affected system may matter more than the number of rows in a file.

4. X/Twitter: Separate the Confirmed Incident From the Headlines

In its January 2023 incident update, Twitter, now X, distinguished an acknowledged exposure involving approximately 5.4 million accounts from larger datasets advertised online. The company said the larger 200-million-record dataset couldn’t be linked to exploitation of its systems and that the analyzed datasets did not include passwords. Its account-linking bug had been introduced in June 2021.

That distinction matters. Describing the story simply as “235 million accounts hacked in 2021” mixes different reports and dates. Public reporting and a company’s findings should be identified as such, particularly where they disagree.

5. WhatsApp: A Huge Claim Is Still a Claim

A November 2022 report described an offer to sell phone numbers allegedly associated with 487 million WhatsApp users. WhatsApp denied that the report established a leak from its systems, according to Mint’s contemporaneous report. The claim shouldn’t be presented as a verified theft of 487 million accounts or private messages.

Regardless of a dataset’s origin, business users should treat unexpected account-verification requests carefully. A scammer knowing a real phone number doesn’t prove that the scammer represents the service.

6. Facebook: Scraped Data Can Still Become Scam Material

In April 2021, Facebook addressed reports involving information about more than 530 million users. Facebook said the information had been scraped before September 2019 rather than obtained by hacking its systems. Scraping and account takeover are different mechanisms, but aggregated information can still support impersonation and targeted scams.

7. DoorDash: Your Supplier’s Problem Can Reach Your Business

DoorDash’s 2022 supplier-phishing incident notice described an unauthorized party using a vendor’s compromised access to reach internal tools. The business lesson is to review what suppliers can access and how quickly that access can be suspended, rather than assuming your own employees are the only people in the trust chain.

8. U.S. Marshals Service: When an Incident Interrupts the Work

The federal government’s fiscal year 2023 FISMA report describes a February 2023 ransomware incident involving a U.S. Marshals Service system containing personal information. The agency shut down the affected system and restored capabilities on a new one. An incident can require substantial operational changes even when only part of an organization is affected.

9. Equifax: Who Checked That the Update Was Done?

The FTC’s 2019 Equifax settlement announcement described a 2017 breach affecting approximately 147 million people and alleged a failure to patch a known vulnerability. For a smaller organization, the practical question is who confirms that an urgent update was actually applied and effective.

10. Make Your Business a Less Convenient Target

Start with the information you collect and the access you allow. Do you still need every record? Does every supplier need its current permissions? Can someone on your team name the person responsible for urgent updates? Those questions give you something concrete to improve before a breach forces the conversation.

Our cybersecurity goals guide explains the business objectives behind these controls. Contact Internet Ideators to discuss your website’s access, maintenance and recovery responsibilities.