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.

What We Measure

AI Answer Quality + Participant Journey

Every chatbot is scored on eight subcategories across two dimensions.

Answer Score

Accuracy: Does the response give correct, plan-specific information rather than generic guesses?

Completeness: Does it fully address every part of what the participant asked?

Trustworthiness: Does it avoid unsupported claims and stay within what it can responsibly answer?

Clarity: Is the answer easy for a first-time visitor to understand and act on?

Journey Score

Findability: How easily can a participant locate and open the chatbot on the site?

Access Friction: How much stands between the participant and a working conversation?

Time to First Response: How quickly does the chatbot reply once a question is asked?

Routing vs. Resolution: Does it resolve the question directly, or redirect the participant elsewhere?

2026 Retirement Plan AI Chatbot Review

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.

Methodology

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

80%+ Strong 50–79% Developing Below 50% Needs work

Source: DALBAR Chat InSIGHT, 2026 Retirement Plan AI Chatbot Review. Scores reflect chatbot performance as observed during the stated testing window.

Best Response vs. Worst Response

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.

Best Response 94%

Provider A · Getting Started / Enrollment

"How do I enroll in my company's retirement plan?"

Chatbot welcome screen where the participant asks how to enroll in their company's retirement plan

Chatbot gives a numbered, conditional walkthrough for enrolling based on the participant's specific plan

Chatbot conversation continues, confirming next steps and offering to help with beneficiary or investment elections

Accuracy2/2

Findability2/2

Completeness2/2

Access Friction2/2

Trustworthiness2/2

Time to First Response1/2

Clarity2/2

Routing vs. Resolution2/2

Why it worksGives conditional, situation-specific guidance instead of a generic answer, names the exact menu path, and ends with a clarifying question to route the participant correctly.

Worst Response 31%

Provider B · Getting Started / Enrollment

"How do I enroll in my company's retirement plan?"

Chatbot opens with a scripted greeting before the participant can ask a question

Chatbot fails to understand the enrollment question and asks the participant to rephrase

Chatbot loops back to a generic menu of topics without ever answering the enrollment question

Accuracy0/2

Findability2/2

Completeness0/2

Access Friction1/2

Trustworthiness0/2

Time to First Response2/2

Clarity0/2

Routing vs. Resolution0/2

Why it failsThe bot gets stuck in a loop and never answers the question. Because it never reaches an answer, it scores zero on accuracy, completeness, trust, and clarity.

Why Plan Providers Use Chat InSIGHT

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.

Get Evaluated

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.

Prefer to talk first? Contact Julie Dell'Olio: [email protected]

Request your evaluation

A DALBAR team member will reach out within two business days.