Skip to main content
AccessForce
Text size: 100 percent

Step 7 — Scale

Projected Impact

These numbers are hypothetical projections for this prototype, not measured results.

What is measured, and what is projected

Measured in this prototype: 214 fields reduced to 8 questions, Grade 13.2 wording rewritten to Grade 6–8, accommodations moved from page 14 to the first message, and every screen usable by keyboard with adjustable text and contrast. Everything below is a projection with the assumption stated, so you can judge the number instead of trusting it.

  • 60%

    reduction in form abandonment

    Projected impact

    Assumption: Assumes progressive disclosure, no timeout and save-and-resume remove the most common reasons people stop.

  • 50%

    faster completion

    Projected impact

    Assumption: Assumes 214 fields collapse to 8 questions, with repeats removed and answers reused.

  • 70%

    fewer accessibility-related support requests

    Projected impact

    Assumption: Assumes accommodations offered up front prevent the calls that ask where the help is.

  • 100+

    languages possible

    Projected impact

    Assumption: Based on the languages large language models already handle; each needs human review before launch.

  • 24/7

    Agentforce assistance

    Projected impact

    Assumption: Availability of the assistant itself; a human caseworker still keeps office hours.

From compliance to inclusion

Compliance asks: did we pass?

A checklist is measured once, against a static page, and it can be satisfied by a document nobody can actually finish. Passing an audit is not the same as being able to apply for help.

Inclusion asks: could this person finish?

AI lets an experience adapt in the moment — reading level, language, pacing, input method, and support — so one experience can fit many different people instead of averaging them all.

See how it works across existing systems