Case Study
GoodBed Turns a Static Mattress Quiz Into a Conversation with Amazon Bedrock
Solutions
Industries
About GoodBed
GoodBed provides objective, personalized guidance to help consumers find the right mattress for them. The platform runs the largest consumer review site for mattresses on the web, backed by an in-depth product testing methodology and a matching quiz that searches across brands and stores, both online and in person. GoodBed reaches millions of shoppers each year through its website and social channels, connecting brands and retailers with an audience of shoppers who know that no single mattress works for everyone. Founded in 2008, GoodBed has pushed for truth and transparency in an industry often clouded by paid placements dressed up as unbiased advice. Learn more at goodbed.com
Challenge
GoodBed’s Key Business Objectives
GoodBed’s quiz hands shoppers a mattress match in minutes, but they didn’t want the conversation to end there. They wanted a scalable way for customers to interact with the results, ask questions, and better understand why each match is a good fit for them.
Their key goals included:
Key Amazon Web Services Used
- Amazon Bedrock
- Amazon S3
- Amazon RDS for PostgreSQL
- Amazon CloudWatch
- AWS Infrastructure
- AWS CloudFormation
Third-Party Tools
- GoodBed Recommendation API
- RAG / Vector Search
Solution
How Cloudelligent Accelerated GoodBed’s Objectives
We built a conversational AI layer on Amazon Bedrock that sits on top of GoodBed’s existing recommendation engine, making it more engaging and easier to explore.
1
A Conversational Layer Powered by Amazon Bedrock
Bedrock is the foundation for the chatbot, turning static quiz results into a back-and-forth conversation. Customers can now ask why a mattress was recommended, what the trade-offs are, or how two options compare.
2
One Experience, Many Data Sources
GoodBed’s product details, reviews, and editorial content live across Postgres, APIs, CMS systems, and S3. We connected these sources through retrieval-augmented generation and vector search, so the chatbot pulls the right context for each question instead of leaving customers to dig for it themselves.
3
Answers Grounded in GoodBed’s Own Logic
Every response stays tied to GoodBed’s existing recommendation engine. The chatbot calls GoodBed’s Recommendation API directly for quiz results, then wraps that data in conversational flows, prompts, and guardrails built to keep answers accurate, including cases where a customer’s quiz responses are incomplete.
4
A Testing Ground Before Launch
We stood up a standalone interface so GoodBed’s team can test and refine the chatbot before it goes live on the main site. That setup gives GoodBed room to add new content sources and features later without starting over.
Results & Benefits
The Value of Our Cloud-Native Solutions
The chatbot turns GoodBed’s quiz results into a starting point for conversation rather than an endpoint, giving customers a new way to understand and trust their mattress match. Key outcomes include:
A More Engaging Recommendation Experience
Customers can now talk through their results instead of reading them once and moving on, turning a static quiz into an ongoing conversation about their mattress match.
Easier Access to GoodBed's Product Knowledge
Reviews, ratings, and editorial content that used to sit behind separate pages are now available directly in the chat, surfaced based on what each customer is asking.
Deeper Product Exploration
Shoppers can ask follow-up questions, compare products side by side, and get pointed to GoodBed’s existing comparison tools and detailed resources, all without leaving the conversation.
A Foundation Built to Grow
This first version gives GoodBed a base to build on, with room to add new content sources, expand integrations, and layer in new conversational features over time.
Client Testimonial
“The Cloudelligent team was a pleasure to work with from the outset. They took the time to really understand our unique needs and concerns, diligently incorporating them into the product’s design and architecture. They were extremely responsive throughout the process, working quickly to address any issues that arose and communicating equally well with both our technical and non-technical team members. Overall, it was a great working relationship, and we remain very pleased with the end result.”
Michael Magnuson
Founder and CEO, GoodBed
Why Amazon Web Services?
Why Cloudelligent?
Cloudelligent is an AWS Premier Tier Services Partner that helps businesses architect, build, migrate, and manage their workloads to accelerate their journey to the agile cloud. We support organizations to achieve the desired outcomes and reach new emerging global markets with a diverse team of AWS-Certified Solution Architects and Engineers.
Ready to Embark on an Epic Cloud-Native Journey?
ascend as an industry pioneer with Cloudelligent
right by your side.