equ is a leading digital agency that designs, builds, and optimizes high-performing digital experiences. As a trusted partner of Progress, equ works closely with enterprise organizations to modernize their digital ecosystems from content management and customer experience to AI-powered innovation.
As digital leaders, equ believes in "practicing what we preach." When the opportunity arose to transform their own website experience, they chose to implement AI search powered by Progress Agentic RAG. Their goal was simple but ambitious: move beyond traditional keyword search and deliver instant, trustworthy, conversational answers grounded entirely in their own content.
In this interview, Stephen Dodge, co-founder of equ, shares why traditional site search was no longer meeting user expectations, why governance and brand safety made a RAG-based approach essential, and how they implemented AI search without re-platforming their CMS. They also discuss what differentiates Progress Agentic RAG from general-purpose AI tools, how they measure success, and what early results they've seen since going live.
What follows is a candid conversation about shifting from "browse and find" to "ask and know" and how AI search can become both a better user experience and a strategic insight engine for modern digital agencies.
Introduction
Q: Before implementing AI search, what challenges were you seeing with your website experience?A: Like many professional services firms, our website is rich in information-case studies, service capabilities, and technical insights. While the information is valuable, requiring a user to read through multiple detailed pages to find a specific answer creates friction, regardless of how intuitive the navigation is. We realised that users didn't want to "browse"; they wanted specific answers. The challenge was that traditional search would just provide a list of links, forcing the user to do the hard work of reading and synthesising the information themselves.
Q: What signals told you that traditional site search wasn't enough anymore?
A: As a digital agency, we know that traditional site search often falls short of user expectations. The technologies and budgets generally aren't there to provide a "Google-esque" search experience on a standard corporate website. We also saw a growing gap between "finding" and "knowing." Traditional keyword search fails when users ask natural language questions or complex queries that require reasoning across multiple pages. We realised that for complex enquiries, users were looking for a "concierge" experience, not a librarian simply handing them a stack of documents.
Q: What business or customer-experience goals were you trying to improve with AI search?
A: Our primary goal was to reduce friction during the consideration phase for prospective customers. We wanted to move users from "curiosity to confidence" instantly. For example, if a user asks about our specific capabilities in a certain sector, we want them to get a synthesised answer immediately, rather than having to piece it together from three different case studies. As digital leaders, we also wanted to prove to our clients that AI is attainable with real use cases. It was critical for us to "practise what we preach" and demonstrate this capability live.
Why Progress Agentic RAG?
Q: What initially attracted you to a RAG-based approach for AI search?A: The absolute requirement for governance and brand safety. We needed a "closed ecosystem" where the AI reasons only across our approved content. We could not afford an AI that hallucinates services we don't offer or speculates on facts. Progress Agentic RAG allows us to ground every response in our own data, ensuring accuracy while still delivering a conversational experience.
Q: What concerns did you have about deploying AI search on a public-facing website?
A: The biggest concern was misinformation risk-the AI "making things up". That is why the Progress Agentic RAG approach was non-negotiable; it is designed not to hallucinate. If the answer isn't in our source documents, the system is configured to say "I don't know" rather than guess.
Q: How important was trust, accuracy, and source traceability in your decision?
A: Critical. In a B2B agency context, our credibility is our product. The ability for the system to provide direct citations and links back to the source content was a major factor. It transforms the AI from a "black box" into an auditable tool that drives traffic back to our source of truth.
Implementation
Q: How did you approach integrating Progress Agentic RAG on your website?A: We took a "layer above" approach. Rather than re-platforming our CMS or migrating data, we used the solution to passively index our existing site structure and documents. This allowed us to activate our full content estate without a massive integration project.
Q: What types of content did you start with?
A: We ingested a mix of structured web pages (HTML) and unstructured documents like PDFs and service capability statements. The platform's ability to handle over 60 common file formats meant we didn't have to rewrite content for the AI; we just pointed it at what we already had.
Q: What surprised you the most about the implementation process?
A: The "multilingual ROI". We discovered that by indexing our English content, the Progress Agentic RAG system could automatically answer questions in other languages (like Hindi or Spanish) without us needing to translate the underlying source material. It effectively killed the "translation tax" while expanding our accessibility.
The Progress Difference
Q: From your perspective, what truly differentiates Progress Agentic RAG from other AI tools?A: It is the "Agentic" layer-the reasoning capability. It doesn't just retrieve text; it plans a path to the answer. For example, it can compare two different service offerings or synthesise a timeline from multiple documents. Standard chatbots can't do that; they just follow scripts or keyword matches.
Q: Why does grounding responses in your own content matter for public-facing use cases?
A: It protects brand integrity. By restricting the AI to approved sources only, we eliminate the risk of the AI referencing unverified external information or competitors. It turns the AI into a "digital concierge" that represents us and our specific tone of voice, rather than a generic assistant trained on the open internet.
Q: Why is Progress Agentic RAG better suited for websites than a general-purpose AI chat?
A: General-purpose AI is designed to be "creative" and answer anything, which is often a liability for a public corporate website where accuracy is paramount. Progress Agentic RAG is designed for "accuracy." It keeps users within our ecosystem by citing our specific pages, whereas a general chatbot might answer the question without ever directing the user to our services or contact forms. It effectively acts as a conversion tool, not just a conversation tool.
Measuring Success
Q: How do you define success for AI search on your website?A: Success is defined by the quality of the interaction, not just the volume. We look at whether users are getting "one-shot" answers-getting the right information in the first query without needing to refine it.
Q: What KPIs or signals are you tracking today?
A: We track the questions users are actually asking. This provides a "real-time signal of user intent" that is far richer than traditional keyword analytics. It tells us exactly where our content gaps are-if people keep asking a question the AI can't answer, we know we need to write that content.
Q: What early results have you already seen?
A: We've seen a shift in how users engage. They are moving away from browsing menus to asking direct, complex questions. The platform has become a "strategic insight engine" for us, revealing emerging themes in what our audience cares about.
Learn More: Experience Progress Agentic RAG in Action
equ's journey demonstrates what happens when AI search moves from concept to production. By implementing Progress Agentic RAG, they transformed their website from a traditional browsing experience into an intelligent digital concierge that delivers accurate, cited answers in real time.If you'd like to learn more about equ, and test website search using Progress Agentic RAG. Go to: equ.com.au
If you're exploring how to modernize search across your website, customer portal, or internal knowledge base, you can see the same capabilities firsthand.
Start a free trial today.
Jay Sanderson
Jay Sanderson is a seasoned digital strategist and practitioner, with a passion for helping businesses achieve growth by exploiting the benefits of marketing technology.