Information Architecture

Internal teams reliably reproduce the org chart because they cannot unlearn it. This engagement's value is external evidence: card sorts and tree tests with real users, scored against targets set before the work starts. If your team's structure passes the tree test, you have cheap confirmation; if it fails, you found out before the rebuild, not after.

The deliverable is a structure with test scores attached: task-level success and directness rates from real users, a page-by-page migration map, and a CMS-ready content model. Every recommendation traces to crawl data, analytics, search logs, or research results, fixed price, named deliverables, acceptance criteria in the SOW.

That's the case, and Avalon says so before anyone has to ask. The risk controls offered instead: firm fixed price, milestone acceptance, a methodology exhibit specific enough to evaluate on its merits, and a scope small enough that the agency's exposure is weeks, not years.

A migration moves content; it does not reorganize it. Replatforming without IA is how agencies pay twice, once to move the confusion and once, later, to fix it. Doing the IA before the migration contract is scoped is the cheapest sequencing available.

The PRA governs standardized collections from ten or more members of the public; it does not prohibit user research. Avalon designs the research to fit your existing generic or fast-track clearance for customer-experience work, or structures it within the nine-respondent threshold.

The engagement is deliberately sized for a senior-led two-to-three-person team, named in the proposal with a key-personnel commitment. Avalon takes one IA engagement at a time at this stage, which is exactly why the people on the proposal are the people doing the work.