What is your AI chatbot really telling participants?
Participants are already asking your chatbot about enrollment, contribution limits, withdrawals, and rollovers, often before they ever speak to a person. Chat InSIGHT is DALBAR's independent evaluation of that AI experience.
We test the chatbot as a real participant would, assessing both Answer Quality and the Participant Journey, then score it against the same rigorous standards used in our Mobile InSIGHT and UXploration Lab programs.
AI Answer Quality + Participant Journey
Every chatbot is scored on eight subcategories across two dimensions.
Answer Score
Journey Score
All 17 providers, ranked by Overall Score
Average Answer Score and Average Journey Score show points earned out of the maximum possible (8 points per question, averaged across all questions evaluated). Overall Score expresses the combined result as a percentage of the maximum possible score (16 points). Testing period: June–July 2026.
Tested like a real participant
17 providers
Selected the participant-facing chatbots of 17 major retirement plan providers.
20 questions
Asked the same 20 standardized questions covering enrollment through retirement planning.
8 categories
Scored each response on Answer Quality and Participant Journey.
1 overall score
Combined results into a single comparable rating per provider.
Read the full methodology
We evaluated the AI chatbots on the participant-facing websites of 17 retirement plan providers, including Fidelity, Vanguard, TIAA, Empower, Principal, and others. Each chatbot was tested with the same set of 20 standardized questions, covering common participant needs such as enrollment, contributions, employer match, investments, loans and withdrawals, rollovers, retirement planning, taxes, and account management.
All questions were asked from the perspective of a plan participant browsing the site before logging in, reflecting the experience of a first-time or prospective user.
Each response was scored using a consistent rubric that measured both the quality of the answer (accuracy, completeness, avoidance of unsupported claims, and clarity) and the overall user experience of reaching and using the chatbot (how easy it was to find, how much friction stood in the way, how quickly it responded, and whether it resolved the question directly or redirected the user elsewhere). These scores were combined into an overall rating for each chatbot, allowing for direct comparison across providers and topic areas.
Full scorecard
Two firms, the same question
A side-by-side look at how differently two real retirement plan chatbots handled the same enrollment question during our 2026 review. Firm names are withheld.
From blind spot to benchmark
Benchmark AI performance against peers
See exactly where your AI chatbot stands relative to other retirement plan providers, based on the questions participants ask most.
Identify AI experience gaps
Understand where your chatbot provides strong answers and where gaps in accuracy, completeness, trustworthiness, clarity, or participant journey create friction.
Build a path forward for AI improvement
Receive a documented, defensible AI evaluation to improve chatbot capabilities, enhance participant experience, and support AI initiatives.
Curious how your AI chatbot would score?
Tell us a bit about your organization and we'll follow up to scope a Chat InSIGHT evaluation for your participant-facing chatbot.
Request your evaluation
A DALBAR team member will reach out within two business days.