In this op-ed, Billy Loizou, APAC Area VP of Amperity, dives into the evolution of marketing personalisation and how the real-world applications of this.
The Evolution of Marketing Personalisation
For marketers, true personalisation at scale has long been an unreachable goal. The closest most teams have gotten is broad segmentation—creating content variants for large customer groups. However, the landscape is rapidly changing. The convergence of customer data platforms (CDPs), cloud-scale analytics, and generative AI is finally making individual-level personalisation possible, particularly by reducing the cost of content creation.
The industry is taking notice. Gartner forecasts that by 2026, 75% of businesses will leverage generative AI to create synthetic customer data, a dramatic increase from less than 5% in 2023. This surge in adoption signals a fundamental shift toward more intuitive, data-driven customer interactions.
Real-World Application: AI-Driven Personalisation in Action
A recent collaboration between Databricks and Amperity demonstrates the practical potential of AI-driven personalisation. Their team developed a generative AI workflow that transforms generic product descriptions into personalised content at scale—a breakthrough for retailers previously constrained by traditional content creation methods.
The workflow combines rich customer data from Amperity—including predicted lifetime value, product preferences, and geographic distribution—with Databricks’ AI capabilities. While the system can theoretically create unique variants for each customer attribute, the team strategically focused on 14 key segments based on lifetime value and product category preferences, proving that meaningful personalisation is achievable without overwhelming complexity.
What makes this approach particularly effective is its practicality. Using a large language model guided by carefully crafted prompts, the system generates unique product descriptions that:
- Maintain consistent brand voice
- Emphasise different product aspects based on segment characteristics
- Adapt messaging for specific audiences (e.g., highlighting premium features for high-value outdoor enthusiasts, or emphasising value and durability for price-sensitive segments)
The Foundation: Quality Data
The success of AI-driven personalisation hinges on high-quality data. First-party data—information collected directly from customers with their consent—is particularly valuable because it:
- Provides accurate insights into individual preferences and behaviours
- Enables generation of highly relevant content
- Reflects real customer interactions
- Forms the basis for meaningful segmentation
- Implementation Guide: From Data to Personalised Content
Here’s a systematic approach to implementing AI-driven personalisation:
- Data Integration: Start by importing first-party customer data into a unified data environment like a lakehouse platform. This will become the foundation for creating segment-aligned content.
- Customer Segmentation: Define customer segments, categorised based on attributes like predicted lifetime value, product preferences, price points, and geographic location. The more granular the segmentation, the more targeted the content will be.
- AI Prompt Design: Design AI prompts to generate personalised content. For instance, to create a customised product description of a winter powder jacket for the “avid skier,” the marketer needs to spell out specific parameters so that the software can generate a description that will resonate with this audience.
- Content Generation: After creating the prompts, use gen AI to create variations of content tailored to each segment. For example, if one customer segment prefers eco-friendly products, a prompt could instruct AI to highlight sustainability in product descriptions.
- Review and Optimisation: Despite AI’s efficiency, human oversight is necessary to ensure content quality. Marketers should review the output to ensure brand alignment and make adjustments as needed.
From Vision to Reality: Your Personalisation Roadmap
Generative AI is revolutionising personalised marketing by making individual-level content scalable and cost-effective. Success depends on combining the right technology stack with quality data and strategic implementation. Companies that invest in building strong data foundations, adopt appropriate tools, and thoughtfully implement AI capabilities will lead in delivering authentic, personalised experiences across channels.
For marketers, this represents an unprecedented opportunity to achieve the long-sought goal of true personalisation at scale. The key is to start small, focus on data quality, and scale strategically as you refine your approach.
You can find a detailed, step-by-step example of how to implement generative AI with Databricks to tailor marketing content here.
