Working as Data Analyst in FMCG Company
It was June 2019, two months prior my graduation day. The offer from this FMCG company came in like a surprise. I didn’t expect it to be a “Congratulations!”, let alone coming faster than I expected. I was thinking, “The job description is crystal clear, but why I couldn’t grasp what I would do there?”.
FMCG (Fast Moving Consumer Goods) company is a company that produces products that are sold quickly and at a relatively low cost. Examples include non-durable household goods such as packaged foods, beverages, toiletries, candies, cosmetics, over-the-counter drugs, dry goods, and other consumables. For example, Unilever, Nestlé, P&G, L’Oréal, etc.
I could not understand FMCG company’s relevancy to tech, how data will help the business, how the company manages the data, and how they decide with data. It’s something I haven’t learned before.
Fast forward 2.5 years later, I got every single answer that popped up in my head that day.
- What do data analysts do in FMCG company?
This depends on the type of products the company produces. It can be personal care products such as soap, toothpaste, or shampoo. It can be beverages. It can be home care products. All these types of products will have different behavior of consumers usage.
However, all businesses ask the same question, including FMCG.
“Does our product work? Do consumers love our product?”
How do data analysts answer this question? Thanks to the internet era, we could get any reviews and opinions about our products easily on any site, social media platform, etc. However, this does not apply to some types of products.
For example, skincare products have very high numbers of reviews on social media. Why? Because skincare products are highly impacting consumers’ appearance. It has a very specific niche with consumers who are highly vocal about it. It is something that’s normally talked about in social media. Thus, if we want to answer the “Does our product work? Do consumers love our product?” question, the data analyst answers it by generating reviews data from social media, then analyzing it using sentiment analysis.
Let’s compare it to home care products, such as dishwasher liquid. Its users are not frequently reviewed by consumers on social media. Hence, if we are asking “Does our product work? Do consumers love our product?”, data analysts usually set up a qualitative interview with consumers to get their opinions on the product. Also, a data analyst could extract data from platforms that specifically provide reviews of home care products.
2. What tools are commonly used by data analysts in FMCG?
- Social listening tools: These types of tools provide a hassle-free and legal process of extracting data from social media platforms. They also provide a dashboard-building feature, in which data analysts could directly build dashboard in-app. The dashboard usually shows the overall insight of consumers’ reviews of certain products. Some examples: Brandwatch, Awario, Synthesio, etc.
- Spreadsheet: Microsoft Excel and Google Sheet are helping us a lot to crunch numbers, build charts, and make a pivot table. It’s important to learn and understand Excel formulas, specifically in manipulating string data types.
- Python: If the analytics process could not be done by social listening tools and spreadsheets, we’ll use Python. Python has stable and up-to-date development in the field of sentiment analysis. They have rich libraries to process text data, such as NLTK, scikit-learn, CoreNLP.
3. How is the workflow? Who do we report to?
Most of the time, data analysts report directly to brand managers. Brand managers are the “brand owner”, which means they have authority in directing the brand. Thus, usually, the flow is as follows:
- Brand managers have a question regarding their product. It could be “Do consumers love the product? Will they repurchase? How is the proposition of our product? How consumers perceive our brand? What will happen if we do A or B?”. They will write a brief to submit to the data analyst.
- Data analysts review the request, then set up a quick call to align on the request. By this, they can estimate the ETA of the analysis result and the deliverable expectation.
- Data analysts do the work, then review the work with the higher-level analyst to refine and recheck the result.
- Once the result is ready to share, the data analyst delivers it to the stakeholder (brand managers), which is usually followed up by an hour-long discussion call.
4. What is the ups and downs commonly faced by data analyst in FMCG?
- The rooms to grow! For me, data analytics in FMCG is still seen as an underrated topic, as FMCG is not the first that comes to mind when talking about “big data” or “data science”. Majorities will think of tech companies when they talk about big data. Thus, this is a perfect opportunity for anyone who wants to explore more about text analytics and its impact on FMCG industry. However, as it is something that’s still in development, it’s hard to compare FMCG’s data analytics culture to the tech company.
- Data availability. Because we mainly use consumers review as our source of data, it’s hard to analyze when the sample size is under the required. There are two solutions; the first is conducting a qualitative interview with consumers, while the second is doing an advanced resampling process to populate the data.
- Quantitative vs Qualitative. The data analytics process in FMCG is not always purely quantitative, as our products are mostly reviewed by text, so qualitative analytics is one of our best friends. It means that having a balanced portion of quantitative and qualitative analytics is the key to generating insightful learnings from the data.
It was a challenging yet rewarding year to be a data analyst in FMCG. It was the best place to learn about mass-products business, critical thinking, and how to take strategic action to drive the business.
Please kindly drop a comment if you have any questions or stories related to this to share. :)
