Sitemap

๐—•๐—ฎ๐—ฑ ๐——๐—ฎ๐˜๐—ฎ ๐—ถ๐˜€ ๐™’๐™ค๐™ง๐™จ๐™š ๐—ง๐—ต๐—ฎ๐—ป ๐—ก๐—ผ ๐——๐—ฎ๐˜๐—ฎ ๐—ฎ๐˜ ๐—”๐—น๐—น

3 min readFeb 18, 2025

--

Press enter or click to view image in full size

Weโ€™re in the age of big data, no doubt, and thereโ€™s an unspoken assumption that ๐˜ฎ๐˜ฐ๐˜ณ๐˜ฆ ๐˜ฅ๐˜ข๐˜ต๐˜ข ๐˜ฎ๐˜ฆ๐˜ข๐˜ฏ๐˜ด ๐˜ฃ๐˜ฆ๐˜ต๐˜ต๐˜ฆ๐˜ณ ๐˜ฅ๐˜ฆ๐˜ค๐˜ช๐˜ด๐˜ช๐˜ฐ๐˜ฏ๐˜ด โ€” which is why companies hoard data, analysts crunch numbers, and executives rely on dashboards to guide strategy.

๐˜ฝ๐™ช๐™ฉ ๐™ฌ๐™๐™–๐™ฉ ๐™๐™–๐™ฅ๐™ฅ๐™š๐™ฃ๐™จ ๐™ฌ๐™๐™š๐™ฃ ๐™ฉ๐™๐™š ๐™™๐™–๐™ฉ๐™– ๐™ž๐™จ ๐™›๐™ก๐™–๐™ฌ๐™š๐™™?

Hereโ€™s the truth: bad data isnโ€™t just an inconvenience, itโ€™s a silent saboteur (take it from someone whoโ€™s dealt with this experience firsthand).

๐˜•๐˜ฐ ๐˜ฅ๐˜ข๐˜ต๐˜ข ๐˜ข๐˜ต ๐˜ข๐˜ญ๐˜ญ would at least force reliance on logic, instinct, or alternative sources of insight, but bad data, on the other hand, creates false confidence and can drive organizations straight into failure.

Let me put it this way: a business making decisions based on inaccurate, incomplete, or misleading data is like a pilot navigating with a faulty map. You know where that would lead, right?

๐—ง๐—ต๐—ฒ ๐—–๐—ผ๐˜€๐˜ ๐—ผ๐—ณ ๐—•๐—ฎ๐—ฑ ๐——๐—ฎ๐˜๐—ฎ

A Gartner study estimated that poor data quality costs businesses an average of $12.9 million per year, which isnโ€™t surprising when you consider how bad data plays out in real-world scenarios:

โ€ข ๐—›๐—ฒ๐—ฎ๐—น๐˜๐—ต๐—ฐ๐—ฎ๐—ฟ๐—ฒ: A misclassified patient record or incorrect dosage data can literally be a matter of life and death. Also, medical AI trained on incomplete or biased datasets can reinforce dangerous misdiagnoses.

โ€ข ๐—™๐—ถ๐—ป๐—ฎ๐—ป๐—ฐ๐—ฒ: A single incorrect decimal point or duplicate transaction entry can distort financial reports, leading to misguided investments or regulatory penalties.

โ€ข ๐— ๐—ฎ๐—ฟ๐—ธ๐—ฒ๐˜๐—ถ๐—ป๐—ด: Flawed customer data leads to irrelevant targeting, wasted ad spend, and loss of potential revenue.

โ€ข ๐—”๐—œ ๐—ฎ๐—ป๐—ฑ ๐— ๐—ฎ๐—ฐ๐—ต๐—ถ๐—ป๐—ฒ ๐—Ÿ๐—ฒ๐—ฎ๐—ฟ๐—ป๐—ถ๐—ป๐—ด: Garbage in, garbage out. Bad training data leads to flawed models, like recruitment algorithms that inherit biases or chatbots that respond inappropriately.

Get Olumide Akinboyewaโ€™s stories inย yourย inbox

Join Medium for free to get updates fromย thisย writer.

๐—ช๐—ต๐˜† ๐—•๐—ฎ๐—ฑ ๐——๐—ฎ๐˜๐—ฎ ๐—ถ๐˜€ ๐—ช๐—ผ๐—ฟ๐˜€๐—ฒ ๐—ง๐—ต๐—ฎ๐—ป ๐—ก๐—ผ ๐——๐—ฎ๐˜๐—ฎ

๐Ÿญ.โ  โ ๐—™๐—ฎ๐—น๐˜€๐—ฒ ๐—–๐—ผ๐—ป๐—ณ๐—ถ๐—ฑ๐—ฒ๐—ป๐—ฐ๐—ฒ ๐—Ÿ๐—ฒ๐—ฎ๐—ฑ๐˜€ ๐˜๐—ผ ๐—•๐—ถ๐—ด๐—ด๐—ฒ๐—ฟ ๐— ๐—ถ๐˜€๐˜๐—ฎ๐—ธ๐—ฒ๐˜€:
If data suggests a trend that isnโ€™t real, businesses might double down on a strategy thatโ€™s doomed from the start.

๐Ÿฎ.โ  โ ๐—œ๐˜ ๐—ฆ๐—ฝ๐—ฟ๐—ฒ๐—ฎ๐—ฑ๐˜€ ๐—Ÿ๐—ถ๐—ธ๐—ฒ ๐—ฎ ๐—ฉ๐—ถ๐—ฟ๐˜‚๐˜€:
One bad dataset rarely stays contained โ€” chances are it gets copied, shared, and integrated into multiple systems. Bigger issue? A single wrong entry in a CRM can ripple across sales, customer support, and forecasting models, causing widespread inefficiencies.

๐Ÿฏ.โ  โ ๐—œ๐˜ ๐—ช๐—ฎ๐˜€๐˜๐—ฒ๐˜€ ๐—ง๐—ถ๐—บ๐—ฒ ๐—ฎ๐—ป๐—ฑ ๐—ฅ๐—ฒ๐˜€๐—ผ๐˜‚๐—ฟ๐—ฐ๐—ฒ๐˜€:
Data analysts and engineers spend up to 80% of their time cleaning and fixing data. With bad data, instead of focusing on insights, theyโ€™re firefighting errors that could have been prevented with better data governance.

๐Ÿฐ.โ  โ ๐—œ๐˜ ๐—–๐—ฎ๐—ป ๐—•๐—ฒ ๐—›๐—ฎ๐—ฟ๐—ฑ ๐˜๐—ผ ๐——๐—ฒ๐˜๐—ฒ๐—ฐ๐˜:
Unlike missing data, which is obvious, bad data often goes unnoticed until itโ€™s too late as anomalies can be buried under layers of reports and assumptions, making it difficult to trace the root cause of a wrong decision.

๐—›๐—ผ๐˜„ ๐—”๐—ป๐—ฎ๐—น๐˜†๐˜€๐˜๐˜€ ๐—–๐—ฎ๐—ป ๐—™๐—ถ๐—ด๐—ต๐˜ ๐—•๐—ฎ๐—ฐ๐—ธ

As a data analyst, your job isnโ€™t just to interpret data, itโ€™s also to challenge it. So hereโ€™s how to protect yourself (and your organization) from the dangers of bad data:

โ€ข ๐—ฉ๐—ฎ๐—น๐—ถ๐—ฑ๐—ฎ๐˜๐—ฒ ๐—•๐—ฒ๐—ณ๐—ผ๐—ฟ๐—ฒ ๐—ฌ๐—ผ๐˜‚ ๐—”๐—ป๐—ฎ๐—น๐˜†๐˜‡๐—ฒ: Always question data sources, check for inconsistencies, and verify collection methods.

โ€ข ๐—”๐—ฑ๐˜ƒ๐—ผ๐—ฐ๐—ฎ๐˜๐—ฒ ๐—ณ๐—ผ๐—ฟ ๐—•๐—ฒ๐˜๐˜๐—ฒ๐—ฟ ๐——๐—ฎ๐˜๐—ฎ ๐—ฃ๐—ฟ๐—ฎ๐—ฐ๐˜๐—ถ๐—ฐ๐—ฒ๐˜€: Push for proper data governance, encourage clean data habits across teams, and educate stakeholders on why data quality matters.

โ€ข ๐—จ๐˜€๐—ฒ ๐— ๐˜‚๐—น๐˜๐—ถ๐—ฝ๐—น๐—ฒ ๐—ฆ๐—ผ๐˜‚๐—ฟ๐—ฐ๐—ฒ๐˜€: Triangulate insights from different datasets instead of relying on a single, potentially flawed source.

โ€ข ๐——๐—ผ๐—ฐ๐˜‚๐—บ๐—ฒ๐—ป๐˜ ๐——๐—ฎ๐˜๐—ฎ ๐—œ๐˜€๐˜€๐˜‚๐—ฒ๐˜€: Track and report data problems so they donโ€™t keep resurfacing.

In the end, itโ€™s not enough to just have data, you should strive for good data because when bad data sneaks into decision-making, the consequences can be far worse than working with nothing at all.

So, the next time youโ€™re handed a dataset, ask yourself: ๐˜‹๐˜ฐ ๐˜ ๐˜ต๐˜ณ๐˜ถ๐˜ด๐˜ต ๐˜ต๐˜ฉ๐˜ช๐˜ด?

If the answer isnโ€™t a firm yes, your first job isnโ€™t analysis; itโ€™s investigation.

--

--

Olumide Akinboyewa
Olumide Akinboyewa

Written by Olumide Akinboyewa

Data Lover. Learn and write about data. Everything about data amazes me. Data-driven decisions are very important.