Getty Images/iStockphoto

Headless CMS vendor Contentstack integrates generative AI

Incorporating consulting services and flexible accommodations for different LLMs, developer-focused Contentstack offers its own take on GenAI for CMS.

Contentstack, a developer-friendly headless content management vendor, released a generative AI toolkit for its customers to integrate into CMS and marketing automation systems along with expanded consulting services to make sense of the myriad large language models users can deploy.

Included in the release are Personalize, a personalization and A/B testing engine native to the Contentstack CMS and Brand Kit, which sets up user-defined guardrails to keep GenAI-created content within users' brand "voice" parameters. Competitors such as Sitecore, Magnolia and startup Storyblok also have added generative AI features to their CMSes.

The idea behind Contentstack's integrated features is to maintain a piece of content's effectiveness derived from A/B tests as well as SEO while, at the same time, using GenAI to automatically create new variants in different languages, of different lengths, or to target different niche marketing segments, Contentstack vice president of product Conor Egan said.

"Brand Kit sits between whatever LLM you choose to use in our system, and that's where you're enriching requests to the LLM with your content, in your voice," Egan said. "Most customers will typically bring their own account, and they can do that with all the major LLMs. We have access to all the most up-to-date versions for all of them."

While Contentstack is LLM-agnostic, they can provide hooks to LLMs on AWS, Google Cloud Platform and Microsoft Azure, as they serve customers in all three of those hyperscaler environments. Contentstack also supports many other popular LLMs such as those from Anthropic, Meta, Cohere and Stability AI as well.

Contentstack also rolled out AI Accelerator, a package of services that includes Contentstack Academy self-paced courses, Brand Kit, consulting sessions, software connectors to build AI automations and 5,000 pre-built automations. It also comes with self-paced courses designed to help customers figure out which LLM and automations are most effective for their type of marketing or e-commerce as well as what their business needs dictate. Different LLMs perform better for different applications, Egan said.

AI Accelerator has, so far, helped Contentstack customer Golfbreaks figure out how AI could create and update web pages, said Jack Simpkins, Golfbreaks digital product owner.

Golfbreaks is still in the preliminary stages of applying AI to tasks such as keeping the 2,000 golf course maps up to date for the tourists who book golf trips through the company's site. But it's clear to Simpkins that Contentstack educational modules and live help sessions has helped set the company's direction in effectively applying AI to their content operations to fulfill business needs.

"Together, we built out the [proofs of concept] for the implementation of AI and automation to specifically address some of the pain points we knew we had, with the hope of saving time and with the end-goal of boosting revenue," Simpkins said. "We need to be efficient, and the AI Accelerator program has moved us many steps closer to fine-tuning that efficiency."

Types of marketing segmentation.

GenAI could be huge leap forward for personalization

Personalization technology has been "clunky as a whole," Deep Analysis founder Alan Pelz-Sharpe said, whether it's coming from marketing automation or content management vendors. Contentstack's developer-centric approach is a step forward in solving that issue.

"It's really hard to do. It is heavily reliant on IT, and what that means is, over time, nobody actually really does it. They just have a few personas," Pelz-Sharpe said. "Contentstack is doing something good here by bundling everything together, taking this out of the hands of IT and techies and giving it back to the actual marketers."

The key to successful GenAI deployments for marketing organizations comes long before switching on the technology. First, the team must locate all the content relevant to brand voice to train the LLM. For many Contentstack users, it's marooned in silos such as emails, on someone's hard drive, or printed on paper somewhere. That is why despite the availability of powerful, advanced GenAI tools, few large enterprises have yet adopted them.

While Contentstack might appear late to the GenAI party, it's still early days for implementation among the buyers and users of the technology.

Brand Kit, Personalize and Academy AI products were created during an internal hackathon soon after ChatGPT was released. Contentstack employees globally submitted ideas for GenAI uses, founder and CEO Neha Sampat said.

"Quickly, they came up with some really cool and interesting use cases on how we could improve efficiencies inside our org, how we could better connect with each other, and how we could use AI to be more competitive and more expedient," Sampat said. "The winning team was actually a mix of marketing people, the people team, and someone on the product team who coached them."

The Contentstack tools were released in conjunction with the company's ContentCon user conference in Austin, Texas, on June 5.

Don Fluckinger is a senior news writer for TechTarget Editorial. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.

Dig Deeper on Content management software and services

Business Analytics
Data Management