OVERVIEW
Stakeholders for an online grocery delivery company want a marketing strategy to target different customers with applicable marketing campaigns.
OBJECTIVE
Perform data and exploratory analysis of Instacart’s customer data to derive insights and suggest strategies for better segmentation.
ASSUMPTIONS
Some departments are more popular than others, some customers spend more than others. Instacart requires targeted marketing strategies to reach specific customers.
SKILLS
- Data Wrangling, Merging, Aggregating, and Grouping
- Deriving New Variables
- Population Flows
- Reporting in Excel
TOOLS
- Python
- Pandas
- Seaborn
- NumPy
- SciPy
- MathPlotLib
GITHUB REPOSITORY
CHALLENGES
Instacart already has very good sales, but stakeholders want to uncover more information about their sales patterns. The company hopes to derive insights and to develop strategies for better segmentation based on the specific customer behavior.
ANALYSIS
Analysis of hourly order data suggests that the busiest days of the week are Saturday and Sunday, and the busiest time of the day is between 10:00 and 11:00. Customers spend the most money early in the morning.
The top performing departments, by number of sales, are: 1. produce, 2. dairy/eggs, 3. beverages, 4. frozen, and 5. pantry. These departments also account for 50% of all sales at Instacart. Produce alone accounts for over 30% of all of Instacarts sales.
“Loyal” customers are those who have ordered more than 40 times. Loyal customers make up 32% of Instacart’s customer base while 49% of Instacart customers are “regular” customers – which means they have ordered more than 10 times, but fewer than 40. Spending habits do not differ among loyal and regular customers. Further, the largest percentage of customers are in the Southern region of the United States.
Instacart customers are primarily married (70%) and parents (75%) who make between $50,000 and $125,000. The largest age group, however, is customers over 70 years old. Ordering time varies, depending on customer age and family status. Customers in the 18 – 20 year old group are more likely to order at 3:00, divorced and widowed customers are more likely to order at 2:00, and parents at 4:00. Outside of ordering time, there are not significant differences between customer by family status and age.
Subgroups have great potential when considering how to market to specific individuals, based on their interests. For example, analysis of the “Pet Owner” sub-group (customers who order from the pet department) and “New Parents” sub-group (parents who shop from the babies department) illustrate that these groups have unique habits. Pet owners, for instance, are more likely to shop at 17:00 and are over two times more likely to be high spending customers. Nearly 40% of new parents are loyal customers, compared to 32% of the general customer population.
RECOMMENDATIONS
Ideal times for advertising targeted at low order times and days is Tuesdays and Wednesdays, after 20:00. Ads, promotions, and recommendations should target the highest performing products. Consider adding higher priced or specialty items in the produce department (organic or locally grown, for example). Create incentives for regular customers to order more frequently – such as coupons. Target marketing campaigns to win new customers in new regions. The Northeast, for instance, has a very high population but the lowest number of customers. Consider customers subgroups when creating marketing strategies.
For loyal customers: The only thing that differentiates loyal customers from other customers is that they order more frequently. Instacart has identified its loyal customers internally, but these customers do not know they fall into this category. Create an incentive program – when a customer has made over 40 orders, let them know they are considered loyal. Offer special incentives or coupons, available only to these customers. Create marketing campaigns to make customers aware of the loyalty program and incentives for when they reach this bar.