Personalisation – the mantra of Digital Marketing and eCommerce

Ivy Aug 11, 2015 No Comments

From being an obscure backstage performer to emerging as the hottest technology-in-use, personalization has come a long way. Deployed across multiple platforms, devices and on-site sensors, used by small digital marketers to ecommerce giants like Amazon, personalization has become the mantra of every business performing in the digital space.

While Google may have kick-started the process of the personalized algorithm with its tailored searches, Facebook gave it a face lift with the “completely personalized newspaper”.  Yet, it is digital commerce which has aggressively leveraged varying degrees of personalization for a choreographed staging of consumer activity, even before the consumer has scripted it! Yes, a user may have already bought furniture and home appliances for his new home, but is nevertheless tempted by ads and promo mails to buy an exhaust fan for the kitchen that was not included in his budget.

In the quest for engaging a customer for a longer CLV, satisfactory experience and higher conversions, user activity is closely monitored for a 360 degree view. A compelling need for hi-performance branding and marketing thus fuels a constant endeavor to devise the best-fit personalized algorithms.

What are the common technologies in use?

Personalization connects technologies and algorithms to data stored in various repositories for a meaningful connection to the customer profile. Customer behaviour, browsing patterns, transaction history, etc., are used to create an experience tailored to the customer.

• Cookies
• Collaborative filtering
• User profiling
• Data analysis tools
• A/B Testing
• Neural network
• Bayesian network
• Rules based engines
• Vendor driven technologies

Businesses like insurance and banking, which require a granular personalization algorithm to keep ahead of competition, have the capacity to invest in sophisticated technologies. Nevertheless, they too are deploying another method for successful implementation – that of Artificial Intelligence – where data is mined for insights driven delivery of personalized advice. The key to building an effective ROI while extracting the most value from the personalized algorithm, is to build a ‘lean’ personalization approach. .

How to make sure the same user comes back soon to make more transactions?

By adopting an intuitive approach to solve the problem in the following ways:

• Using experimentation, identify models that need to be refined or ‘tweaked’.
• Leverage guesswork within a tested / structured framework
• Exposing a trial or experimental personalization to a small sample of members, for a sense of its viability while limiting the cost of errors
• Feedback confined to staff, to understand the ‘whys’ behind product recommendation
• For improving brand mix or category cadence in a targeted or personalized sale
• Human experience and cognitive thinking (visitor who buys   is most likely
• A time –tested mechanism for recommendation arsenal
• Isolating the trial to prevent losses from an error
• User testing   to prevent costly errors and

• Employee only release feedback
• Human experience
• User testing – trial and error
• Use of artificial intelligence
• Expert systems

Regardless of the approach used, the end-all is to keep improvising the techniques and algorithm design. Improving the quality of data, creating a time-honoured “test and learn cycle” mechanism with targeted segment, and incorporating the same into the system, are the Holy Grail of personalization.