Using AI to Personalise Your Marketing

The Importance of Personalisation.

We’ve all been the recipients of badly targeted marketing emails. You know the ones that offer you something that you’ve never wanted, needed, or indeed even considered was a worthwhile part of the universe. Half-priced mince advertised to Vegans….

These kinds of emails often do more harm to the business sending them than good. You wonder who thought it was a great idea to send YOU that offer? Don’t they know who YOU are? What YOU like? That you’ve never once wanted THAT?

Sadly, the brutal reality is that most companies just don’t have the marketing sophistication to be able to send offers that are targeted precisely to each customer. That was until now. With the explosion of AI tools, there’s no excuse for knowing your customers better (laziness is not a valid excuse).

Using Customer Segmentation

Customer segmentation is the process of dividing your customer base into distinct groups based on shared characteristics, such as purchasing history or demographics. This allows you to tailor your marketing strategies to each group, improving engagement, satisfaction, and ultimately, sales.

Depending on the data you have available, customers can be segmented using characteristics, such as:

• Demographics: Age, gender, income, location.
• Behavioural: Purchase history, website interactions, product preferences.
• Psychographics: Interests, values, lifestyle.

Example Use Case

Imagine you run an e-commerce store. Here’s how you might use customer clustering:

1. Data Collected:
   • Purchase history (e.g., total spend, frequency, product categories).
   • Demographic data (e.g., age, location).
   • Behavioural data (e.g., time spent on site, pages visited).
2. Clustering Results:
   • Cluster A: Young adults (18-25) who frequently buy trendy clothing.
   • Cluster B: Middle-aged professionals (35-50) who purchase high-end electronics.
   • Cluster C: Seniors (60+) who buy health and wellness products.
3. Marketing Actions:
   • Cluster A: Promote new arrivals and social media campaigns.
   • Cluster B: Highlight premium products and offer financing options.
   • Cluster C: Send health tips and discounts on wellness products.

Using AI to Perform Customer Segmentation

Prior to Machine Learning and Artificial Intelligence, Customer Segmentation was a tedious task assigned to data analytics tools. Querying a large data set meant a substantial investment in time and resources. I know. One of my first roles in IT was to provide data extracts to management decision-makers. There was data to massage into useable formats, data warehouses to be built, and large SQL queries to perfect. If any component was wrong, the results were completely rubbish.

It also meant that queries were not very dynamic. Someone knowledgeable about the underlying data had to specify what to look for. It was not easy to “stumble upon” interesting correlations that may have had marketable implications.

Thanks to AI, these kinds of queries are much easier to run, and the tools can be put in the hands of your sales and customer service team.

• Want to know more about what products young people buy? You can.
• Want to see if there’s a connection between product sales and suburbs? Of course.
• Want AI to find connections you didn’t know existed? Absolutely.

In a recent post, we looked at how you can analyse your own data using DeepSeek AI. Using the same dataset to demonstrate the power of Customer Segmentation, we asked the following question.

perform sales analysis based on customer age groups

The output from DeepSeek provided more than enough to whet the appetite of any sales manager.

### Sales Analysis Based on Customer Age Groups

To analyze the sales data based on customer age groups, we first categorize the customers into age groups. Here's how we define the age groups:

- **18-30 years**: Younger adults
- **31-45 years**: Middle-aged adults
- **46-60 years**: Older adults
- **61+ years**: Seniors

We will calculate the total sales amount and the number of units sold for each product category within these age groups.

---

#### Step 1: Categorize Customers by Age Group
We calculate the age of each customer as of October 2023 and assign them to the appropriate age group.

#### Step 2: Aggregate Sales by Age Group and Product
We group the data by age group and product category, then calculate the total sales amount and quantity sold.

---

### Results

#### Table: Sales by Age Group and Product
| **Age Group** | **Product**   | **Total Sales Amount** | **Total Quantity Sold** |
|---------------|---------------|------------------------|--------------------------|
| 18-30         | Laptop        | $4,800.00              | 4                        |
| 18-30         | Smartphone    | $6,750.00              | 9                        |
| 18-30         | Tablet        | $3,500.00              | 7                        |
| 18-30         | Headphones    | $700.00                | 7                        |
| 31-45         | Laptop        | $7,200.00              | 6                        |
| 31-45         | Smartphone    | $6,750.00              | 9                        |
| 31-45         | Tablet        | $4,000.00              | 8                        |
| 31-45         | Headphones    | $1,000.00              | 10                       |
| 46-60         | Laptop        | $2,400.00              | 2                        |
| 46-60         | Smartphone    | $1,500.00              | 2                        |
| 46-60         | Tablet        | $1,000.00              | 2                        |
| 46-60         | Headphones    | $500.00                | 5                        |
| 61+           | Laptop        | $1,200.00              | 1                        |
| 61+           | Smartphone    | $750.00                | 1                        |
| 61+           | Tablet        | $500.00                | 1                        |
| 61+           | Headphones    | $100.00                | 1                        |

---

#### Key Insights

1. **Age Group Preferences**:
   - **18-30 years**: This group shows a strong preference for **Smartphones** and **Tablets**, with high sales amounts and quantities sold. They are also active buyers of **Laptops**.
   - **31-45 years**: This group is the most active across all product categories, with the highest sales for **Laptops**, **Smartphones**, and **Tablets**. They also purchase a significant number of **Headphones**.
   - **46-60 years**: This group shows moderate interest in **Laptops** and **Smartphones**, with fewer purchases of **Tablets** and **Headphones**.
   - **61+ years**: This group has the lowest sales across all categories, with minimal purchases of **Laptops**, **Smartphones**, and **Tablets**.

2. **Product Popularity**:
   - **Laptops** are popular across all age groups, but especially among **31-45-year-olds**.
   - **Smartphones** are the most popular product among **18-30-year-olds** and **31-45-year-olds**.
   - **Tablets** are also popular among younger and middle-aged adults.
   - **Headphones** are purchased more frequently by **31-45-year-olds**, likely as accessories for other devices.

---

#### Recommendations

1. **Targeted Marketing**:
   - Focus on **Smartphones** and **Tablets** for the **18-30-year-old** demographic through social media and influencer marketing.
   - Promote **Laptops** and **Smartphones** to the **31-45-year-old** group via email campaigns and professional networks.
   - Offer discounts or bundles on **Headphones** to encourage purchases among **31-45-year-olds**.

2. **Product Development**:
   - Consider developing accessories or features that appeal to younger customers, such as gaming laptops or high-performance smartphones.
   - For older age groups, emphasize ease of use and durability in product marketing.

3. **Customer Engagement**:
   - Create loyalty programs or exclusive offers for **31-45-year-olds**, as they are the most active buyers.
   - For **61+ customers**, provide personalized assistance and tutorials to encourage adoption of technology products.

Just knowing the information listed in Key Recommendations would give your sales teams insights to help reach their KPIs next quarter. It might be time to update the marketing plan for 61+ year olds.

• 18-30 years: This group shows a strong preference for Smartphones and Tablets, with high sales amounts and quantities sold. They are also active buyers of Laptops.
• 31-45 years: This group is the most active across all product categories, with the highest sales for Laptops, Smartphones, and Tablets. They also purchase a significant number of Headphones.
• 46-60 years: This group shows moderate interest in Laptops and Smartphones, with fewer purchases of Tablets and Headphones.
• 61+ years: This group has the lowest sales across all categories, with minimal purchases of Laptops, Smartphones, and Tablets.

Fear not, DeepSeek even gives you a suggested approach, just to get you started.

For 61+ customers, provide personalized assistance and tutorials to encourage adoption of technology products.