The 90-Day Visibility Playbook: SEO, GEO, and AEO in Parallel

If you read the visibility stack post and accepted the three-layer model, the next question is the one I get in DMs every time. What do I actually do.

This is that post. The 90-day playbook. Three parallel tracks. Specific weekly outputs. Built for an operator with an existing site, an existing SEO program, and zero current GEO or AEO work in flight.

Not a content strategy in the abstract. A schedule.

The Setup Before Day One

Before you start the 90 days, three things have to be in place.

One. Baseline measurement. Open ChatGPT, Claude, Gemini, and Perplexity. Ask each one to describe your business. Record what each one says. Date the file. That is your baseline. You will compare every monthly check-in against it.

Two. Your existing SEO program is documented. What ranks. What does not. What pages drive traffic. What pages do not. You cannot maintain SEO while adding GEO and AEO if you do not know what your current SEO actually does.

Three. An hour a week of protected time. This is the cost of the playbook. Not five hours a day. One hour a week, on the calendar, recurring. The work compounds because the discipline is steady.

If you cannot commit one hour a week, do not start the playbook. The compounding requires the discipline.

Days 1 to 30: The Audit Month

Month one is audit. You are not building anything new yet. You are measuring what you have and identifying the gaps.

Week one. The AI engine audit. Run the baseline check across the four major engines. Document what each one says about your business. Categorize the errors. Wrong facts. Missing facts. Outdated facts. Confused identity (engines mixing you up with another business). Each category becomes a target for the work that follows.

Week two. The first-party content audit. Open your three most important pages. Homepage. About. Services or products. For each page, answer four questions. What is the H1. Does the first paragraph directly answer what the page is about. Is the structured data on the page valid Schema.org. Does the prose match what you want AI engines to say about you.

Most operators find their homepage fails at least two of the four. That is normal. The pages were written for human visitors, not for AI engine extraction.

Week three. The canonical question audit. List the five questions a prospect is most likely to ask an AI engine about your business. What do you do. What does it cost. Who founded it. What industries do you serve. What is your engagement model.

For each question, write the canonical answer the way you would say it on a sales call. Three sentences or less. This is the raw material for your CORD-AEO envelopes.

Week four. The competitive citation audit. Ask the four engines about your top three competitors. Note which ones get described accurately, which get described with errors, and which barely get mentioned. The pattern tells you where the citation graph favors you and where it does not.

Month one ends with four artifacts. Baseline AI engine descriptions. Page-level first-party content audit. Canonical answers for five questions. Competitive citation map.

Days 31 to 60: The Build Month

Month two is build. You take the audit artifacts and start producing the new content the playbook requires.

Week five. Fix the homepage H1 and opening paragraph. Rewrite both for direct answer extraction. The H1 should name what you do in five words or fewer. The opening paragraph should answer the question "what is this business" in one sentence, then expand. AI engines extract the first sentence preferentially.

Week six. Publish your first CORD-AEO envelope. Use the canonical answer from week three for the most important question. Build the JSON file. Upload to /.well-known/cord-aeo/. Validate at cordspec.org/aeo/validate. The implementation guide walks through the structure step by step.

Week seven. Publish your second and third envelopes. The first envelope teaches you the workflow. The second and third are faster because the patterns are familiar. By the end of week seven you should have three envelopes live and validated.

Week eight. The first GEO content piece. Write a long-form article that comprehensively answers one of your canonical questions. The article exists to build topical authority that AI engines can cite. Different format than the CORD-AEO envelope. The envelope is the canonical answer. The article is the depth that makes the canonical answer credible.

Month two ends with the homepage rewritten, three CORD-AEO envelopes live, and one GEO content piece published. The visibility stack is now active across all three layers.

Days 61 to 90: The Measure Month

Month three is measurement. You started with a baseline. Now you check whether the work is moving the numbers.

Week nine. Re-run the AI engine audit. Same four engines. Same canonical questions. Compare each response to the baseline from week one. Note three things. What improved. What stayed the same. What got worse, if anything.

Most operators see partial improvement after thirty days of envelope publishing. Required claims start appearing more reliably. Some errors disappear. Some engines update faster than others. None of this is uniform. Track it.

Week ten. Schema.org bridge verification. Use Google's Rich Results Test on every page that has Schema.org markup. Fix anything broken. The bridge inside your CORD-AEO envelopes only works if the page-level Schema.org is also clean.

Week eleven. The fourth and fifth envelopes. By this point the workflow is muscle memory. Publish the remaining envelopes from the canonical question list. Five envelopes total covers the most common AI engine queries about your business.

Week twelve. The 90-day debrief. Sit down with the baseline and the current measurement. Three questions to answer.

Did AI engine accuracy improve. If yes, by how much, and for which engines.

Did SEO traffic hold steady or improve. If it dropped, what changed.

Which canonical questions still need work. Add those to the next 90-day cycle.

The 90-day playbook is not a one-time exercise. It is the first iteration of an ongoing operating rhythm. The compounding is in the iteration.

What the Playbook Does Not Include

Three things I deliberately left out.

One. Backlink building. Authority signals matter for SEO and GEO both, but the playbook is for operators with limited time. Building inbound links is slower and harder to schedule. If you have an existing backlink strategy, keep it running. If you do not, the 90-day playbook will not start one.

Two. Paid content distribution. Promoted posts, sponsored placements, paid review sites. None of this is in the playbook because none of it scales with operator time the way the playbook does. If you have a budget and an agency, use them in parallel. The playbook works regardless.

Three. Social media output. LinkedIn, Twitter, YouTube. These build distribution and reach. They do not directly improve AI engine citations. The playbook is focused on the visibility stack. Social distribution is a different track, valuable but separate.

The Weekly Cadence That Actually Works

One hour a week. Same day. Same time. Calendar-blocked.

Inside the hour, the structure is consistent.

Ten minutes. Review last week's output. What got published. What got measured. Any anomalies.

Forty minutes. Execute this week's specific output from the playbook. Audit, build, or measure depending on the month.

Ten minutes. Update the running log. What you did. What you noticed. What you want to remember for next week.

The log is the asset. After 90 days you have twelve weekly entries showing what you built, what you measured, and what changed. That log is the answer to the question every operator's leadership asks. "What are we doing about AI search."

Two Versions of How This Goes Wrong

The playbook fails for two predictable reasons.

One. Skipping the audit month. Operators who jump straight to publishing CORD-AEO envelopes without measuring baseline AI engine accuracy never know whether the work moved the numbers. The envelope is live. The engine response did not change. There is no diagnostic for why. Audit first.

Two. Treating the hour as flexible. Operators who move the weekly hour to "whenever I have time this week" stop having time within four weeks. The discipline is the asset. Protect the hour.

The playbook is not technically hard. The hard part is the discipline of doing it every week without exception for ninety consecutive days.

What Compounds After Day 90

The work pays off in three ways the operator notices.

AI engine descriptions of your business align with what you want them to say. The errors decrease. The required claims survive paraphrase. The right buyers find you through AI engines because the engines describe you correctly.

SEO holds its ground. Because the playbook does not require killing SEO, the ranking and traffic you already have stays in place. The new work compounds on the existing foundation.

Your operating rhythm matures. The weekly hour becomes ingrained. By month four you stop noticing the discipline because it is just part of how you work. That is when the compounding accelerates.

One Honest Note About Scope

This playbook is for operators running solo or with a small team. Larger organizations need a different shape because the work crosses too many functions. Marketing owns SEO. Content owns GEO. Nobody owns AEO. Larger orgs need a coordination layer the playbook does not provide.

If you are at a company with multiple departments touching visibility, the playbook is still useful as the activity catalog. The coordination work is yours to design.

The operator-class playbook works because one person can hold all three tracks in their head at once. That is its strength and its limit.

If you are the right person to run it, the next ninety days are how you find out.

Structure before tools. Strategy before speed. Systems before scale.

The AI Operator Framework is the methodology for designing, governing, and scaling AI operations inside a business, including the data and citation fidelity layers most companies skip. Explore the full framework at JamilAshkar.com/framework. Start with the free course at JamilAshkar.com/course.