The 90-day plan · a proven method

The 90-Day AEO Success Model

A staged, measured path from a baseline near 4% Share of Answer to a defensible outcome — what happens each phase, who owns it, and what ships.

The agency playbook, not a promise

Most agencies approach Answer Engine Optimization the way they approached SEO in 2012: reactively, inconsistently, without a repeatable system. They run an audit, publish a few pages, and wait. Nothing happens.

The agencies winning AEO mandates in 2026 have a model — a structured, milestone-driven 90-day engagement that moves a client from zero visibility in AI answers to measurable, defensible citation share. They can show a prospect exactly what happens in week one, what ships in week six, and what the dashboard looks like on day ninety.

This is that model — a week-by-week execution guide, built around Ardent, a fictional mid-market B2B software brand, as the working example throughout.

The core premise: AEO is not a one-time optimization. It is a compounding system. The agencies that win long-term retentions show clients a repeatable engine, not a one-off project.

How the model is structured

Three phases, each building on the last. Skipping or compressing a phase is the single most common reason AEO engagements underdeliver.

Each phase runs in weekly sprints. Each sprint has a defined focus, a deliverable, and a role assignment so Account Managers, strategists, and client stakeholders know exactly what they own.

The Ardent context

Ardent is a mid-market B2B SaaS company selling project-management software to operations teams at 200–2,000-employee companies. Before engaging, Ardent had:

• ●Zero tracked citations in AI answers across ChatGPT, Perplexity, and Google AI Overviews.
• ●Strong domain authority (DR 58) but content written for keyword ranking, not AI extraction.
• ●A 14-page content library — blog posts, product pages, one comparison guide — none structured for answer-engine consumption.
• ●A clear competitive threat — two direct competitors already appeared in AI answers for Ardent's core queries.

A representative starting point: established presence, decent content volume, zero intentional AEO architecture.

Phase 1 · Foundation (Days 1–30)

Objective: establish a verified baseline of AI citation visibility, identify the highest-value opportunities, and align on strategy — before a single word of new content is written. The most expensive mistake in AEO is producing content before you understand the citation landscape.

Week 1 — Kickoff & baseline audit

Account Manager: run the kickoff (confirm stakeholders, approval workflow, publishing access); set up the tracking dashboard monitoring citation frequency across ChatGPT, Perplexity, and Google AI Overviews for the 20–30 highest-priority queries; document the existing content inventory with URLs, dates, and organic performance.

Strategist: run the AI citation audit — manually query each priority term across all three engines and record which domains are cited, how often, in what context; identify the citation gap (which competitors appear, for which queries).

Deliverable: the Week 1 Baseline Report — current citation share (almost always zero for new clients), competitor citation frequency, and the query set that anchors the engagement.

Ardent: the team ran 28 queries across Ardent's core clusters. Competitors A and B appeared in 11 of 28. Ardent appeared in zero. That gap became the engagement's north star.

Leadership note: the baseline audit is also a sales tool. If a prospect asks "what does month one look like?", this is your answer — concrete, auditable, and it sizes the opportunity immediately.

Week 2 — Query mapping & topic prioritization

Not all queries are equal. Some are high-frequency but dominated by entrenched third-party sources. Others are mid-tail queries where the client's authority gives a realistic path to citation in 60–90 days. Week two separates them.

Account Manager: facilitate the query prioritization workshop; confirm content approval timelines (AEO content often moves faster than traditional editorial calendars).

Strategist: build the Query Priority Matrix — score each query on citation difficulty, business value, and content feasibility; identify "quick win" queries where no authoritative source dominates and a well-structured piece could earn a citation within 30 days.

Deliverable: the Query Priority Matrix — a scored, ranked list segmented into Quick Wins (Phase 2), Core Targets (by day 60), and Stretch Goals (Phase 3+).

Ardent: of 28 queries, 7 were Quick Wins. The most valuable: "how to track team tasks without a spreadsheet" and "project management software for operations teams" — neither had a dominant citation source; both mapped to Ardent's primary buyer.

Week 3 — Content architecture & briefs

AEO content is architecturally different from SEO content. The goal is not to rank on page one — it is to be the source an AI engine extracts and cites. Structural principles applied this week:

• ●The 40–60 word direct answer rule — every target page opens with a concise, self-contained answer to its primary query. This is the paragraph the engine extracts; if it needs context from the rest of the article, it will not be cited.
• ●H2/H3 as extraction architecture — subheadings mirror the exact phrasing of follow-on queries ("What does it cost?" beats "Pricing overview").
• ●Evidence density — specific data points, named examples, and cited sources get extracted; generic claims do not.

Account Manager: deliver content briefs (target query, required opening answer, H2/H3 structure, evidence requirements); confirm the production workflow and SLAs.

Deliverable: content brief package for the first wave of Quick Win pages (5–7 briefs).

Ardent: the brief for "how to track team tasks without a spreadsheet" specified a 52-word opening answer, four H2 sections mirroring follow-on queries, a comparison table, and at least two cited external studies. It left no room for interpretation.

Week 4 — Phase 1 review & production launch

Account Manager: present the Phase 1 summary (baseline findings, Query Priority Matrix, architecture principles, production plan); secure sign-off on the briefs.

Leadership: QA the summary before the client presentation; verify the prioritization is defensible and the briefs meet the agency's AEO standards. Use this as a template for prospect pitches.

Deliverable: Phase 1 summary deck (client-facing) + signed-off content production plan.

Ardent: the VP of Marketing attended. Zero citations against competitors' 11-of-28 created immediate urgency; the matrix showed which investments move the needle first. She approved the full Phase 2 plan in the same meeting.

Phase 2 · Execution (Days 31–60)

Objective: produce, publish, and technically optimize the content from Phase 1; begin tracking citation movement and adjust early on what the data shows.

Week 5 — First wave content production

Targets the 5–7 Quick Win queries — the pieces most likely to earn citations inside the 90-day window.

Account Manager: manage the workflow; flag any brief deviations early (a piece that drifts from its structure won't perform).

Strategist / Writer: produce first drafts following the brief — self-contained opening answer, query-mirroring H2/H3, evidence density, internal linking. Each piece includes at minimum one comparison table or structured list, one callout with a specific data point, and a concluding section answering a natural follow-on query.

Deliverable: first drafts of all Quick Win pieces, ready for internal AEO review.

Ardent: five pieces in week five. The top piece came in at 1,840 words with a 54-word opening answer, three comparison tables, and two cited studies. Review flagged one section whose answer needed the previous paragraph to make sense; it was rewritten to be self-contained before client submission.

Week 6 — Client review, revision & publication

Account Manager: submit drafts with a structured review guide (what to check — accuracy, brand voice, missing context — and what NOT to change: structural elements that exist for AEO, not editorial preference); manage revisions on a 48-hour SLA; coordinate publication with the client's web team.

Key instruction: when a client wants to "clean up" the opening answer (shorten it, make it more marketing-friendly, remove the direct-answer framing), push back. That paragraph is the extraction target. Frame it in outcomes: "This structure is what gets us cited. Changing it changes the result."

Deliverable: all Quick Win pieces published and live; URLs submitted for indexing.

Ardent: two change requests. One was a legitimate factual correction (accepted immediately). The other asked to replace the opening answer with a brand narrative — the AM walked the client through how engines extract content, using a competitor's cited page as a live demo. The original structure was preserved.

Week 7 — Structured data, schema & technical layer

Well-structured content is necessary but not sufficient. The technical layer tells engines what a page is and how to categorize it.

Strategist / Technical: audit all Phase 2 pages for schema; verify canonicals and that nothing is blocked from indexing; check page speed (engines favor fast, accessible pages).

Account Manager: coordinate implementation with the client's web team; confirm all pages indexed within 72 hours.

Deliverable: technical audit report — schema status, indexing confirmation, page-speed scores.

Ardent: three of five pages had no schema. The team added FAQPage to two and Article to all five. Page speed (1.8s avg) needed no intervention.

Week 8 — Core Target production & first citation check

By week eight the Quick Win content has been live ~two weeks — enough for a meaningful first check.

Account Manager: run the first post-publication citation audit (all 28 queries × three engines); begin the Core Target brief package (tier two); schedule the mid-engagement check-in and present early data, even if modest — momentum matters.

What to expect: most engagements see 1–3 citations in the first check. If zero appear after two weeks, escalate for a content audit.

Deliverable: first citation movement report + Core Target briefs (5–8 pieces).

Ardent: the week-eight check showed the first two citations — "how to track team tasks without a spreadsheet" in Perplexity, and "project management software for operations teams" in Google AI Overviews. The VP called them "the most tangible results we've seen from any content investment in two years."

Phase 3 · Optimization (Days 61–90)

Objective: analyze citation data at scale, optimize underperformers, close competitive gaps, and build the retention case for month four and beyond. The engagement shifts from production to intelligence.

Week 9 — Full citation audit & gap analysis

By week nine there is enough data to draw conclusions. The audit answers four questions: which pages are earning citations, and where; which target queries still show none despite published content; where competitors gained or lost share; and which queries generate citations for sources you did not anticipate.

Account Manager: run the full audit (all 28+ queries × three engines); build the Gap Analysis Report (expected vs. actual per piece); flag underperformers (live 3+ weeks with zero citations).

Strategist: diagnose underperformers — opening answer not self-contained, insufficient evidence density, topic-authority mismatch, or technical indexing issues — and recommend specific remediation.

Deliverable: full citation audit + gap analysis report.

Ardent: citations on 4 of 5 Quick Win pages and 2 of 6 Core Target pages. One underperformer, "best workflow tools for operations managers," had zero citations after three weeks. Diagnosis: the opening answer was too broad and referenced other sections. Remediation: rewrite it as a self-contained 48-word direct answer.

Week 10 — Content remediation & second-wave optimization

Remediation is not failure — it is the normal process of AEO. Even well-structured content sometimes needs adjustment after you see how engines respond. Actions, in priority order:

• 1.Rewrite the opening answer to be fully self-contained (most common fix).
• 2.Add or restructure the evidence layer (replace vague claims with specific, cited data).
• 3.Expand the H2/H3 structure to capture follow-on queries the current structure misses.
• 4.Add or correct schema if the technical audit reveals gaps.
• 5.Build internal links from high-authority pages to the underperformer.

Account Manager: submit remediation requests with specific instructions; frame it to the client as optimization, not error-correction.

Deliverable: remediated pages, re-published and re-submitted for indexing.

Ardent: the "best workflow tools" page was remediated in 48 hours with a self-contained opening answer naming specific tool categories. Within ten days it earned a citation in ChatGPT.

Week 11 — Stretch goals & competitive displacement

By week eleven the Quick Win and Core Target content is performing or remediated. Now the engagement pursues queries where a named competitor is currently cited.

Displacement is a longer game — typically 60–90 days of consistent, higher-quality content on the same topic. Beginning in week eleven lays the groundwork for the month-four retention conversation.

Strategist: identify the 3–5 highest-value queries where a competitor is cited and the client has a credible authority claim; build displacement briefs designed to be demonstrably more comprehensive, evidence-dense, and structurally optimized than the cited page; begin production.

Account Manager: present the displacement strategy (show the competitor pages being cited and why yours will outperform); set realistic timelines — week-eleven content may not earn citations until month four or five.

Deliverable: competitive displacement brief package (3–5 pieces) + competitor citation-share analysis.

Ardent: Competitor A was cited for "project management software for remote teams" across all three engines — a 900-word overview, no table, no schema, generic opening. Ardent's displacement brief called for a 2,200-word guide with a 55-word opening answer, a six-column feature table, HowTo schema, and three cited studies — intentionally the best available answer to that query on the internet.

Week 12 — 90-day report & retention roadmap

The 90-day report is both a performance document and a sales asset. The retention conversation should feel inevitable, not like a pitch.

Account Manager: build the report from the accumulated tracking data; prepare the retention recommendation with scope options — maintenance retainer (sustain share), growth retainer (expand coverage), or full-service retainer (AEO plus broader content strategy).

Leadership: review before delivery; use the anonymized Ardent result (0 → 8 citations across 28 queries in 90 days) as a case-study template for pitches.

Deliverable: 90-day performance report (client-facing) + internal case-study template.

Ardent: by day 90, 8 citations across 28 queries — 4 Quick Win, 3 Core Target, 1 remediated. Competitor A had lost 2 (pages not updated in 14 months). The retention recommendation proposed a growth retainer targeting 12 new queries per quarter, projecting 40% citation share by month nine. Ardent signed a six-month extension before the meeting ended.

The complete deliverable map

Every deliverable across all twelve weeks, in one reference.

Leadership note: prospects who ask for a detailed SOW before signing are looking for evidence you've done this before. This map is that evidence. Customize it with their brand name and query clusters and it becomes a near-bespoke proposal built on a proven system.

What makes it work — the three non-negotiables

1 · Content architecture before content volume

AI engines do not reward volume; they reward extractability. One page with a self-contained opening answer, proper schema, and evidence-dense content outperforms ten structurally generic pages. Phase 1 is not overhead — it is the foundation that decides whether Phase 2 produces citations or just content.

2 · Citation tracking as a core deliverable

Organic traffic and keyword rankings are not AEO metrics. Citation frequency across AI platforms is. Track it weekly and you can adjust in real time; track it quarterly and you're flying blind. The minimum: 20–30 queries audited weekly across the three engines; a citation log (date, platform, query, URL) for every citation earned; and competitor tracking on the same set, so you know if you're gaining share or just adding citations while competitors grow faster.

3 · Client education as an ongoing process

The most common source of friction is a client who doesn't understand why AEO content looks different — the opening answer feels blunt, the subheadings mechanical, the evidence excessive. The agencies that retain past 90 days invest in education from week one. Every touchpoint reinforces why the structure matters and what it produces.

The compounding effect: the first 30 days produce the baseline, the next 30 the first citations, the final 30 the data that makes the retention conversation easy. Set expectations in the kickoff, reinforce at every check-in, and the 90-day model closes itself.

Quick-reference: phase-gate checklists

Phase 1 exit (Day 30): baseline audit complete (all queries, three engines) · Query Priority Matrix finalized (3-tier) · brief package approved (min. 5) · production workflow confirmed · summary deck presented and plan signed off.

Phase 2 exit (Day 60): all Quick Win pieces published and indexed · schema implemented on all pages · first citation report delivered · Core Target briefs approved and in production · mid-engagement check-in complete.

Phase 3 exit (Day 90): full citation audit complete (vs. day-0 baseline) · underperformers remediated and re-published · displacement briefs delivered · 90-day report presented · retention recommendation delivered with scope options.

Red flags — escalate to leadership when: zero citations at week eight (live two weeks, no movement) · a client requesting structural changes to opening answers, H2/H3, or schema (education call before approving) · competitor citation share growing faster than the client's · content production delays exceeding five business days.

Final note

The 90-day model is not a guarantee — results depend on domain authority, content quality, query competition, and approval speed. What it guarantees is a repeatable, auditable process that gives every engagement the best possible chance of measurable citation growth in a defined timeframe. The agencies that win long-term AEO retentions are not the ones with the best writers or the deepest technical knowledge — they are the ones with the most credible process.

90-day schedule

The 90 days, mapped

Four workstreams across twelve weeks, three inspection gates. Read it left to right — the plan a client can track.