a

Lorem ipsum dolor sit amet, consectetur adicing elit ut ullamcorper. leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet. Leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet.

  /  Project   /  Blog: Email — Subject Line Personalisation (AI/Machine Learning)

Blog: Email — Subject Line Personalisation (AI/Machine Learning)


Go to the profile of Ram

Concept

An organization, BigCo, builds and sells an email marketing automation product to enterprise clients that is used to send millions of emails to their users. The variety of emails sent by clients range from information emails to newsletters, promotions, lead nurturing, and shopping emails, etc.

Goal

Automate and personalize email subject lines, with an aim to improve the performance of emails sent by their clients.

Benchmarking & Competitor Analysis

Performed desk based study to analyse the competitors’ offering and problem space in existing solutions. Various solutions already exists.

1. Customized {FirstName} {LastName} & fallback options if data doesn’t exist can be configured in some products.

2. Dynamic subject line {if}….{else} conditions for personalised message.

3. Engagement metrics are part of the email marketing products. Offering insights on previous campaigns and success rates.

Few other examples

User persona

Wireframe

Explored the design solutions around subject line input field. While a marketing professional is in the process of creating a campaign mail ‘Subject Line’ is one of the key information to grab user’s attention.

Design

Screen 1
  • From the left menu item ‘Campaign’ > user begins configuring the ‘From’ & ‘Subject’ fields as first step to begin the process.
  • TESS is an AI assistant with machine learning (ML) capabilities to predict AI score for each personalized ‘Subject’ line measuring the open rates and other details of the campaign.
Screen 2
  • When a user types a subject line and clicks ‘Next’, TESS helps predict the Open rate score.
  • AI with predictive text using ML recognizes phrases, emojis etc with deep learning.
  • Marketeer can add additional preset tags or remove (on hover) irrelevant tags to arrive on a realistic score, based on previous campaign data.

Inference: ‘Welcome’ email to all potential leads, where emoji is not a bad idea to begin with.

Screen 3
  • At an advanced stage of a campaign, when marketer wants to send a promotional email.
  • Based on the type of brand, product, how deep you are in the campaign journey builder, past data analysis user can see different charts for e.g., open rates & age group classification.
  • After checking the open rate score, if a user doesn’t change the tags the CTA updates to ’Next’ button.
  • ‘Last subject’ used is displayed if it’s not the first campaign mail.

Concept Validation

There are a lot of predictive text analysis tools in the market for e.g. Google Cloud Natural Language has ingredients to bring the subject line personalisation concept to reality.

Syntax Analysis

Extract tokens and sentences, identify parts of speech (PoS), and create dependency parse trees for each sentence.

Entity recognition

Identify entities and label by types such as person, organization, location, events, products, and media.

Sentiment analysis

Understand the overall sentiment expressed in a block of text.

Content classification relationship graphs

Classify documents by common entities or 700+ general categories

Source: https://cloud.google.com/natural-language/

If you like the concept , please 👏

Source: Artificial Intelligence on Medium

(Visited 13 times, 1 visits today)
Post a Comment

Newsletter