The last few years have seen a wide array of challenges before the Consumer Packaged Goods (CPG) Sector. Buyer expectations have changed, there’s been a pandemic and lockdowns that disrupted the supply chain, companies have faced innumerable logistics-related issues, they have been forced to invest heavily in new technologies and so on.
In light of these hurdles, data has proven to be critical to survival and growth. But, collecting data alone isn’t sufficient. It isn’t about quantity, it’s about quality. Data quality determines its value and potential benefits. Let’s look at some of the reasons why CPG brands must pay attention to data quality.
Maintain The Ideal Supply Chain
Every business plan is based on ideal scenarios and lists certain policies to give it a competitive advantage. But, these policies can have an impact only if the right questions are asked about the data behind them. Poor product data can affect inventory, fulfilment and logistics.
For example, if inventory data about shipping is not accurate and consistent, it becomes difficult to maintain predictable turnovers which in turn, could result in a slower turnover or the product being out of stock for customers though there may be adequate inventory. Similarly, inconsistent data could cause errors in distribution and shipping which further lead to unnecessary additional expenses.
Maintaining high levels of accuracy, completeness and consistency can help CPG brands push past these issues. Along with creating policies, companies need to link the policies to the potential impact they could have if they were not achieved. Assessing this connection and the cost incurred can provide a baseline for data quality measurement and help create targets for improvement.
Gain Reliable Insight From Data Analytics
Though studies have shown that data-driven businesses have a 23X likelihood of acquiring customers, only 59% of organizations use big data analytics. A lack of trust in the quality of data being used for analytics is one of the main reasons for this.
The trouble lies in inconsistency. Cornflakes in the sales database, corn flakes in the product inventory and corn-flakes in the order sheets may all refer to the same product. But, since they are written differently, machine learning and AI algorithms may be unable to make the connection. Typographic errors and taking into account the multiple variations available in terms of pack size, flavour, etc. could further complicate the issue. How does the company identify the top-selling SKU if data is listed so inconsistently?
Cleaning the data and standardizing formats can resolve the issue exponentially. Graphs and charts are simplified and brands can assess the performance of a product across sales channels, geographic locations, etc.
Have Data Ready For Enhanced Machine Learning
In continuation to the above point, CPG brands looking at leveraging data from internal and external sources to train machine learning and AI models must ensure that data meets all the quality control standards. As the saying goes, ”’garbage in, garbage out’ .
A typographic error that lists the available cornflakes stock as 1200 units instead of 200 units can throw the entire model out of gear. Similarly, if a record is simply written as cornflakes 100, the relevancy is lost does the number refer to the stock available or the number of units sold. Many similar data quality issues can cause trouble with machine learning models.
Data must be cleaned and standardized for it to be useful for machine learning. It must be complete, stored in a standard format, accurate, valid, correctly labelled and within the specified range. Quality is even more important for cases of unsupervised learning where an AI model needs to make connections between data for itself.
Deliver Personalized Experiences
Given the many platforms for retails available, today’s buyer is more concerned about the experience of shopping for a product than the product itself. Thus, CPG companies must provide personalized experiences that make customers engage with the brand. This is only possible when they have access to good quality data from the product details as well as the customer profile.
For example, let’s say the company’s sales chart shows that there have been 1000 units sold of strawberry jam through the website and 700 units sold on the app. It would signal that the former is a more popular shopping platform and the brand would accordingly direct campaigns towards promoting the website. But, if the order value for website sales had been duplicated, the actual sales volume would have been higher on the app. Thus, the sales channel preferred by the customers would not get the attention required.
Here’s another simple example of how data influences experiences – if there are multiple records for a single customer, he/ she will receive multiple copies of the same emailer addressed to different variations of their name. Rather than be piqued about the promotions, they may get frustrated by the excess emails and unsubscribe from the mailing list.
Data As The Foundation For Success
To set themselves up for success, data needs to be thought of not just as an asset but as a strategic initiative involving all the key stakeholders in the organization. The days of manually entering and verifying data are over. Instead, companies need to invest in verification tools that can keep poor quality data from entering the database and enhance data to improve the overall data quality standards. Not only does this have a lower risk of errors, it is also quicker and works in real-time to improve the overall customer experience.
For data to be useful, it must be accurate, complete, verified, up-to-date, listed in a standardized format and unique. Data needs to be checked against these criteria at the time it enters the system as well as on a regular basis. The latter helps identify data that may have gone bad with time.
For example, the city may have changed a street name and thus addresses will have to be updated. It is only when these criteria of data quality are addressed can CPG brands use it to evolve into intelligent enterprises.
