Type to SearchView Tags

Artificial Intelligence in Marketing Automation
Asish Kumar Patra Senior Technical Architect, Digital & Analytics | January 24, 2020
206 Views

In today’s competitive world, it is vital for a brand to build a successful marketing automation platform and strategy. It streamlines and simplifies most time-consuming tasks of modern marketing and helps in building contextual and efficient experiences based on each individual’s needs. Artificial intelligence (AI) in marketing automation, is playing an increasingly significant role to take marketing to the next level. It is now at the center stage to revolutionize digital marketing and radically change marketing automation strategy by implementing next-gen marketing automation platforms, to meet the demands of fast and hyper-personalized service for consumers. Marketers are adopting the capabilities of AI and machine learning to improve their digital marketing game plan. This enables greater customer experiences by leveraging valuable and actionable customer insights. AI automates campaign orchestration duties and decisions based on individual, unique profiles.

Key Drivers for Adoption of AI in Digital Marketing

With the introduction of modern digital channels and touchpoints like social media, smart mobile phones, voice assistant devices, smart TVs, and IoT devices, data pertaining to brand-customer interactions across different channels, has grown exponentially in recent years. Marketers must analyze and segment large amount of complex data sets to create true personalized experience. There is a need to automate the exploration of complex relationships in large amounts of data and a recurring need for predicting things that either cut costs or create value.

Each interaction with consumer should be consistent, continuous, and 1:1 relevant. The quality of great experience can be achieved if it is compelling, personal, useful, and everywhere.

Digital Marketing

Digital innovation processes are touching every industry throughout the world. There is huge enterprise-level interest in AI projects and their potential, to fundamentally change the dynamics of business value.

AI Capabilities in Adobe Campaign

Adobe introduced new AI capabilities in the Adobe Marketing Cloud platform that aims to deliver real-time personalization tools for marketers. AI features in the Adobe Campaign are meant to help brands automatically deliver the right experiences to the right individuals, at the most opportune time.

The three key AI-powered features of Adobe Campaign are:

  • Predictive Send Time Optimization
  • Predictive Churn: Intelligent Fatigue Management
  • Predictive Subject Lines

Predictive Send Time Optimization

The target audience list can consist of different sets of people with different lifestyles, habits, and geo- presences. It is very hard for marketers to decide the best time in a day or the best day of the week to send their daily, weekly, seasonal, or ad-hoc marketing communications. Some people like to check their inbox early in the morning and some, after working hours or school hours. The global presence of customers and their time zone difference adds more complexity and challenge to this problem. If we miss contacting the prospects and customers at their most opportune time, the email may get ignored and result in decline of engagement rate. However, in reality there is no single specific time and day with which you can treat everyone in your target audience group. There is a need to come up with a best time of engagement for each individual profile rather than treating them with a one-size-fits-all approach.

Adobe has recently launched a new AI feature in Adobe Campaign called ‘send time optimization’. It is the automated deployment of emails at the best time for each person. It determines the best time of the day and the week to send emails based on the continuous learning model of past response data.

Adobe Sensei takes the past Adobe Campaign delivery logs of contact history information, tracking logs with the response history data, and customer demographic data as input, to predict the best time of the day and the week to send the email for each customer.

This should considerably increase engagement with rise in open and click-through rate.

Predictive Churn: Intelligent Fatigue Management

Churn could happen due to many different reasons and churn analysis helps to identify the customers or subscribers that are more likely to leave or unsubscribe. This opens opportunities to implement effective retention strategies.

Adobe Sensei takes the past Adobe Campaign delivery logs, tracking logs, and customer attributes, and assigns customers to be engaged, at risk or fatigued segments as input, based on their likelihood to unsubscribe. It continuously learns and trains the model based on response data. These data can be used with manual segmentation or automated filtering to extend special offers.

With an effective retention strategy, marketers can decrease the opt-out rates and increase revenue.

Predictive Subject Line

With email being one of the popular and secure channels of communication, everyone tries to inbox their consumer with their marketing, promotional, or transactional emails. The subscriber may ignore, delete, or mark your emails as spam, unless your email is relevant, personalized, and compelling. The real challenge for marketers is how they can entice consumers to open their emails by making their subject line stand out, being appealing and meaningful among others. It is the subject line which the subscriber reads first when they review the email inbox.

Adobe Campaign, powered with Sensei, nurtures customer data and forms the bedrock for data-driven marketing.

Adobe Campaign’s predictive subject line feature powered with Adobe Sensei can do the following, to improve the performance of emails sent for a campaign:

  1. Predict the open rate
  2. Indicate the right length of subject line
  3. Recommend the specific words and adjectives

for an email subject lines to improve the performance of emails sent for a campaign.

Adobe Sensei takes Adobe Campaign’s datasets for subject line, sent count, open count, and date of previously sent emails as input, and applies a data science algorithm to it based on a trained model, to predict open rates when you enter a subject line to test. Adobe has pre-trained models for the cosmetic, the supermarket, and the medical industry. You can also train your own model with your contact and response history data.

This helps the marketers with greater personalization and saves time by automating the process of creating subject lines.

Rolling Out AI Features of Adobe Campaign For Your Organization

Before you roll out any of the above AI features for your organization, you need to establish a bidirectional data ingestion process between Adobe Campaign and Sensei. This is to collect the delivery and tracking logs from Campaign to Sensei and receive personalized predicted values from Sensei to Campaign.

You can conduct tests with few deliveries by setting up the test and control segment. Adobe Sensei optimizes the models by continuously training them with the information regarding the nature of the campaign you run for your organization, your target audience’s response to the campaign, geographical presence, and demographic data, among others.

Here are a few guidelines that can be used while you are performing control testing using AI optimization:

  • Do a random 50-50 split of the target audience to create test and control groups.
  • Control goes with no optimization.
  • Test goes with optimized sending using the AI features.
  • Compare the results against control after 5 - 7 days of delivery.
  • You should see lift in the open and click-through rates.

After a successful test, you can automate your campaigns to adapt new AI features for optimization.

Conclusion

These features of Adobe Campaign Standard powered with Sensei, nurtures customer data and forms the bedrock for data-driven marketing, under the umbrella of HCL’s ADvantage Experience portfolio. Customer data from the Adobe Campaign are continuously looked into, for a greater detail to train the models to infer the right results in Sensei and pass them over to campaign to target the customers with the appropriate subject line and personalized send times with optimal usage of the email channel with intelligent fatigue management.