AGENTIC SEARCH OPTIMIZATION

Overview

● v1.0 · Research PreviewNorthstar AI
FROM VISIBILITY TO SELECTION

Measure whether AI agents can actually choose you.

ASO evaluates the full decision path—not just whether a brand is mentioned, but whether an agent can discover, understand, verify, qualify, select and act on it with sufficient confidence.

Northstar AIShortlist enterprise AI visibility vendorsEvidence-backed scoring
DISCOVERUNDERSTANDVERIFYQUALIFYSELECT / ACT
ASO SCORE
74.6selection readiness
Selection Readinessweighted
74.6
Selection-ready with material gaps
Baseline51.8
Readiness Delta+22.8
SELECTION FUNNEL

Decision-stage progression

Current normalized score by stage.

live from assessment
PRIORITY GAPS

What blocks selection

Lowest-scoring criteria first.

auto-ranked
EVALUATION SIGNALS

Latest probe set

Derived from Evaluation Lab runs.

demo run
ASO FRAMEWORK v1.0 · DEVELOPED BY SHEN XU · 2026

Six stages between being visible and being selected.

The framework treats agentic search as a decision system. An entity must move through six distinct gates. Failure at any gate can reduce recommendation probability even when visibility is high.

24scored criteria
ASSESSMENT CONTEXT

Define the entity and decision task

This context anchors every score and evidence item.

autosaved locally
EVALUATION ENGINE

Run decision-task probes

Use demo mode or connect an OpenAI-compatible endpoint.

DEMO
Ready0 / 0
RUN OUTPUT

Agent decision signals

Probe results are experimental signals and remain separate from the 24-criterion framework score.

Separate experimental layer
EVIDENCE COVERAGE
0%traceable proof
PROVENANCE PROFILE

What supports the score

Coverage rewards traceable URLs and stronger evidence types.

EVIDENCE MATRIX

Criterion-level provenance

24 criteria
StageCriterionCurrentSource typeEvidenceURL
BENCHMARK SET

Compare selection readiness

Use audited or externally sourced scores.

RELATIVE ASO SCORE

Competitive readiness

0–100