Charles | Data Analyst

Numbers lied 42M in Revenue ❌ 29M in Revenue ✅ A clothing business recorded two years of sales data across 2023 and 2024. On the surface, the numbers looked promising 500 orders placed, products moving, customers buying. So when management sat down to review performance, they expected the revenue figure to reflect that activity. They expected a number that matched the energy of 500 orders. What they got instead was confusion. The revenue was lower than anticipated, and nobody could explain why. They had sold more or at least it felt that way but the money did not add up. The business was making decisions based on total order volume without understanding that a significant portion of those orders never actually converted into real income. Returned goods were quietly eroding revenue. Cancelled orders were inflating their order count without contributing a single naira. Pending transactions were sitting unresolved, neither confirmed nor lost. And nobody was separating these outcomes from the genuine completed sales. The numbers were not wrong. They were just being read without context. That is where the data analyst came in not just to organise the data, but to translate it. To answer the question the business had been asking without realising it: we sold a lot, so where is the money? Here’s what the data actually showed: → Real revenue: ₦29.4M from completed orders only → Denim Jacket was the #1 product in both sales and revenue → Outerwear alone generated ₦12.1M → Abuja was the strongest city at ₦7.3M → 40% of sales came from online → 90% of 341 customers never came back → 2024 made more money without selling more that’s the smart pricing strategy. Here’s what I told them to fix: 1. Find out WHY customers are returning goods ₦3.4M is recoverable 2. Follow up on 54 pending orders before they become cancellations 3. Build a retention system losing 90% of customers after one purchase is a leaking bucket 4. Push harder online lowest cost, highest return channel 5. Never let the Denim Jacket go out of stock The numbers were never the problem. Reading them without context was. What other insight or recommendations can you make out of this?

@aledueben Congratulations 🎉

I got the job

@WithBukky Congratulations 👏…. Saw where you commented that your laptop got stolen yesterday … more goodies to come 👏

Update: I’m excited to share that after the interview, I’ve been selected as one of the beneficiaries of the 10K Laptops Project. Thank you for this incredible opportunity. I’m truly grateful. Special appreciation to all the conveners of this project for making this possible. #10KLaptopsProject @DONJAZZY @silverpenydr @boye4christ2006 @SgtShow01

Just because a number is the highest doesn’t mean it’s the answer. In statistics, this mistake often comes from insufficient sample size when you don’t have enough data points to make a result reliable. I was analyzing an employment dataset, trying to find which education level produces the highest average performance score. Here’s what I got: PhD → 5.0 MBA → 4.74 MSc → 4.37 BSc → 4.02 HND → 3.63 OND → 2.97 On the surface, PhD wins. Easy conclusion, right? Wrong. I checked the sample size behind that 5.0. n = 1 One employee. That’s not a pattern that’s an outlier. You cannot build a business decision on a group of one. So I dug deeper. MBA holders had an average score of 4.74 based on 8 employees across multiple departments. Now that’s a result you can actually start to trust. The PhD category? Excluded from the final conclusion and flagged as statistically insignificant due to insufficient sample size. This is the difference between reading numbers and understanding them. Anyone can run a query and report the top row. A good analyst asks: “How many observations is this based on?” Because in data analysis: n = 1 is not a trend. It’s one person having a good day.

@pacify_kay Exactly

In statistics, even if you collect data on how the number of dogs in a country influences GDP, you will get results 😂 However, it’s the intuition behind those results that matters 🤝

@snr_elvisco Thank you

@CasmirCodesData Keep it up👏

A work in progress >>>>> See you soon ❤️

@Cruxoferudition Exactly

@CasmirCodesData It takes full attention to details to discover this

@chidirolex Congrats boss 👏

Another Upwork gig. This is an invite and it was just straight to contract and interview. Forever grateful Heavenly Father.🙏🏾

@WashidKhan_Dev Thank you 😂🙏…. You got a nice query line btw

@CasmirCodesData SELECT * FROM Charles_Progress ORDER BY Skills DESC; Results look promising

Coming together >>>

@HarryJ96997 Done ✅

@CasmirCodesData Check ur Dm

@Raji_Mariam01 Sure

@CasmirCodesData Can’t wait for the final outcome ☺️

@stio_biggy Thank you

@CasmirCodesData Nice one brother

@_ZainabOladokun All the best 👏👏

OMG 😭 I'm so happy.🥹 I was so scared with everything I've been reading online about people having to try again and again before Upwork verify them. You guyssssss, Upwork just verified me on first try😭😭

@AdetunjiAbassi Real like mad 😂

“The pain that you feel when food no dey na hin go make you ginger properly”

@TheIbibioGirl Both

What's your favorite data analysis process.. Data Cleaning or Visualizing the data?

@ObohX People wey suppose open shop Dey sell akara 😂…. This week I couldn’t send a job application through my email because of poor network

Airtel too useless make I no fit send ordinary mail .. Fxcking mail on a 5G network

@Cristiano @AlNassrFC @AlNassrFC_EN 🐐

One mission. We keep working. TOGETHER! 🟡🔵

@Fimiye LOVE IS A KINGDOM 👑

Name an album you can listen to every song from the beginning to the end and not have the urge to skip a song?

Most people overcomplicate SQL’s CASE WHEN… but it’s really just an IF statement in disguise. If you’ve ever used Excel, you already understand it. Here’s the simple breakdown: CASE WHEN → IF this condition is true THEN → return this value ELSE → otherwise return this END → close the statement That’s it. Example in SQL: SELECT full_name, CASE WHEN performance_score >= 4.5 THEN ‘Top Performer’ WHEN performance_score >= 3.5 THEN ‘Average’ ELSE ‘Needs Improvement’ END AS performance_category FROM employees; Same logic in Excel: =IF(P2>=4.5,“Top Performer”,IF(P2>=3.5,“Average”,“Needs Improvement”)) Different syntax. Same thinking. Where it gets powerful in SQL is when you combine it with aggregations: SELECT SUM(CASE WHEN exit_date IS NOT NULL THEN 1 ELSE 0 END) AS total_exits FROM employees; What’s happening here? You’re turning a condition into numbers: 1 for every row that meets the condition 0 for every row that doesn’t Then SUM adds them up. That’s how you count specific things without filtering your dataset. Why CASE WHEN is powerful: → Works inside SUM, AVG, COUNT → Handles multiple conditions cleanly → Scales across thousands (or millions) of rows If you’re coming from Excel, don’t overthink SQL. You already know this just a different language.