Main content

The next marketing frontier - personalization at scale

Barb Mosher Zinck Profile picture for user barb.mosher June 5, 2024
Summary:
Contentstack brings real personalization to the digital experience.

personal

Brands are facing a relevancy crisis. Although digital experience has become the competitive frontline, it's so noisy and crowded that traditional digital channels have lost their effectiveness. Forget about staying ahead; just staying afloat is a challenge for most brands, and executive leadership isn't helping by slashing budgets. So what's the answer?

There are three main challenges to personalization at scale, according to Conor Egan, Contentstack’s VP of Product:

  1. A fractured authoring experience forces teams to work in multiple interfaces to define and implement personalization (especially in a composable environment). A heavy reliance on IT to help also makes it more challenging.
  2. Personalization requires a lot of content - more content than usual, but budgets are not increasing to support this additional content. For those who think AI is the answer, it's not that simple.
  3. Scaling personalization requires you to operationalize the entire process in a repeatable way.

 

To overcome these challenges, Contentstack has unveiled several new premium features that support this push to deliver real personalization at scale, and it should be no surprise that AI plays an important role.

Personalize - a built-in A/B/n testing and segmentation engine

Contentstack Personalize brings a new A/B/n testing and segmentation engine to the platform and a new authoring experience. The new decision engine allows you to layer audiences on top of each other to create highly unique experiences. The testing experience provides an infinite combination of content to use in those experiences. You can personalize based on A/B testing, a segmentation rule, or a combination of the two, giving you many combinations.  Egan says:

We may have a repeat customer that is a reward member, and is also in Europe. We can take elements of all those different audiences and combine them into a permutation or an experience that is tailored to those combinations of audiences.

All this is completed in one authoring experience and workflow within Contentstack, removing the requirement to go between tools to create segments, run A/B tests, and create content.

The Contentstack Brand Kit

As brands look to AI to help them generate content, they also look for ways to leverage AI to ensure that content is on brand. Contentstack has extended its investment in AI (Chat GPT was integrated into the editing experience last year) and now offers a Brand Kit.

Before the Brand Kit, Contentstack worked only with stock Large Language Models (LLMs). Now it has developed a retrieval augmentation AI system. This type of system connects to LLMs but can also tap into additional data sources without needing to retrain the model.

In Contentstack's case, two additional data sources are connected that ensure content is always on brand:

  • Knowledge Vault: This repository of a brand's information assets, such as websites, blogs, product information, technical documentation, and other content. A brand can have multiple vaults.
  • Voice Profiles: Brands can create multiple voice profiles (e.g., technical, formal, casual) that shape the tone, style, and language of AI-generated content. You can provide writing samples to help train voice profiles and indicate what can or cannot be said. Voice profiles can also be adjusted to audience, intent, or channel.

Contentstack pitches this as AI that is useable out of the box.

By combining the Brand Kit with Personalize (all selected within the same authoring environment, content teams can create genuinely personalized content and experiences. Think about a website visitor and how they move through the site. As they look at different content, visit different pages, and consume content, a brand can now automatically change the experience.

One key point to make here is that all this work is done by the content team/marketer. Egan said a developer sets up a small amount at the beginning, including installing the SDK and importing audiences from connected data sources, but then everything is controlled by the marketing team.

Along with Personalize and Brand Kit, Contentstack announced new AI connectors for Automate. Now, brands can connect with all the major LLMs, including Azure GPT, OpenAI, Google Vertex, Amazon Bedrock, and more.

Helping brands accelerate their use of AI

Contentstack has introduced the AI Accelerator Program as part of the Contentstack Academy. The new program is designed to help a brand go from zero to fully deployed with AI in 45 days or less.

The program combines software and training.

The software includes AI Assistant, AI Connectors for Contentstack Automate, 5,000 Automate executions, and the Brand Kit with 1 Brand Vault. There are five self-paced Academy courses and five AI-enablement sessions with Contentstack’s Technical Services Organization (TSO).

My take

The Contentstack team demoed the new A/B testing and personalization experience, and it was very cool, if not a little overwhelming. Working with testing, segmentation, and a brand kit in one tool is good. What slows marketers down is working with multiple tools that don't always work nicely together.

Designing the experiences felt overwhelming. Granted, it was a fast demo that attempted to show a lot of functionality at once. As with any digital experience, much work needs to happen upfront to determine what tests to run and what visitor activities would result in adjusting the experience on the fly. However, what was clear is that marketers have so many options with Contentstack to create the right personalized experiences at the right time.

And that’s a good thing. A recent Twilio State of Personalization report said that personalization helps improve customer retention as well as customer acquisition. There is so much noise now, especially with AI content leaping off every website, that breaking through requires doing more to show customers you get them. This is where AI really shines through.

Loading
A grey colored placeholder image