Hire an AI CRO specialist

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• Findings ranked by how much they hurt, quick wins first, with a page score out of 100
• Every issue sorted into fix now, investigate or test, with its proof
• A second, separate check of the fixed page: 30 yes-or-no results with quotes
• Five fixes still worth making and a second score

How it does the job well

• Find where visitors drop off, with numbers, before proposing fixes.
• Read real customer words (reviews, surveys, support tickets) to learn the doubts the page must answer.
• Rank ideas by expected impact, confidence and effort.
• Fix clear bugs and broken flows directly. Test changes whose effect is uncertain.
• Set the sample size and run time before a test starts, and run whole weeks.
• Write for one named audience and one action. Each asset has a single job.
• Use the customer's own words from reviews, calls or surveys before inventing language.

Mistakes it avoids

• Stopping a test as soon as it looks significant.
• Copying generic best practice without evidence from this page.
• Testing many changes at once so the cause is unclear.
• Generic copy that could fit any brand.
• Comparing periods or campaigns that are not comparable, such as different dates, budgets or tracking.
• Calling a test winner before there is enough data.

What a good result looks like

An example, not your task: Your page gets visits but few sales and you don't know where to start.

• Step 1 audit report. What the page is for, findings rated 0 to 4, quick wins, test ideas and a page score out of 100.
• Step 2 sorted issues. Every issue with its proof, sorted into tracking faults, fix now, investigate and test, in ranked order.
• Step 3 second check. Thirty yes-or-no checks on the fixed page, a quote for every no, five fixes left and a second score.
• HAND-BACK SUMMARY. Each step's file, any step not run and why, and every MISSING line in one list.

How the work is checked

• Each issue cites its evidence: analytics, recordings or customer words.
• Each test has a hypothesis, a primary metric, a sample size and a stop rule.
• The updated page is checked again against the same checklist.
• Each step passed its own checks, or the run stopped at the step that failed and says why.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way Peep Laja and Jakob Nielsen would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Audit the page. Names what the page is for, then reads it as a glance, a skim and a doubt. It rates each finding from 0 to 4, splits quick wins from tests and scores the page.
• Sort what to fix now. Numbers every issue with its proof. Tracking faults come first, then clear fixes to make now, then causes to look into, then ideas to test, ranked by gain and effort.
• Make the fixes. A pause. You make the fix-now changes and save the updated page text and screenshots.
• Check the fixed page. Runs 30 yes-or-no checks at the five stops a buyer makes. It quotes the page for every no, tags it a deal-breaker or a drag, and lists the first five fixes still left.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id cro-specialist only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/cro-specialist/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• Your landing page as a link, text and screenshots
• The ad that sends traffic to it (optional)
• Your brand facts (optional)
• After the pause, the page again with the fix-now changes made

Tools and access

• Landing page and customer evidence
• Analytics or experiment data when available
• A rendered browser preview for page checks
• Any good chat AI: Any current chat AI that reads screenshots can run the three reads. Give it phone screenshots, top to bottom.
• PageSpeed Insights: Shows your page's real-user speed against Google's 'good' marks, so the phone check has numbers.
• Evan Miller's sample size calculator: Tells you how many visitors a test needs, so you know if it can finish in about a month.
• Any good chat AI: Planning an order of research is text work any strong chat AI does with this skill.
• Google Analytics: Free. Shows pages, steps and where people leave. Export visits per step before you start.
• A session replay tool: Optional and often paid, and it needs an account and a snippet on your site. Lets you watch real visits to a page.
• Google Analytics 4: Source of the traffic acquisition, landing page and funnel reports.
• A heatmap and recording tool (optional, often paid): Shows what shoppers do on the page the skill flags. You can read the report without it.
• Your email tool or ad platform: Most have an even split option that does the halves for you.
• A spreadsheet: Holds the keys, counts and read dates for every test in one place.
• An AI with a browser or screenshot upload: It can see the real page and its layout, instead of guessing from pasted text.
• Your analytics tool: Compare the screens where the visitor loses interest with the screens where real people leave.
• Any good chat AI: Sorting and ranking a list with reasons is text work any strong chat AI does with this skill.
• A spreadsheet or task board: Holds the ranked list and tracks what is done.
• Any good chat AI that reads images: It can judge the first-screen and phone checks from your screenshots instead of guessing from text.
• PageSpeed Insights: Free from Google. It gives the phone speed numbers check 30 needs from just the page's address.

Functions and deliverables

• Landing page repair reviewer: Rank landing page problems, separate fixes from test ideas and review the updated page against a 30-point checklist.
  • Findings ranked by how much they hurt, quick wins first, with a page score out of 100
  • Every issue sorted into fix now, investigate or test, with its proof
  • A second, separate check of the fixed page: 30 yes-or-no results with quotes
  • Five fixes still worth making and a second score
• Landing page conversion reviewer: Share your page and screenshots. Get conversion problems ranked by impact, with recommended fixes and a score.
  • The page's job, a one-line verdict, and whether the page matches its ad
  • What a buyer takes from a five-second glance and from a skim
  • A ranked findings table, each with a severity from 0 to 4 and where it sits
  • Up to five quick wins, up to three tests, and what the page can't tell you
  • A page score out of 100 worked out from the severities, now and after the quick wins
• Conversion research planner: Plan your conversion research before you test. Paste your page and analytics. You get an order that finds where people drop and why.
  • A research order with the question each step answers
  • A tracking check to run first
  • A plan for replays, user tests and surveys
  • A template for writing findings as issues
• Store page conversion analyst: Read your GA4 store reports to see which channels and product pages turn visits into purchases, where shoppers drop out, and how reliable the read is.
  • Channels ranked by purchase conversion and engaged share
  • Product pages above and below your store average, each with a first cause to test
  • The funnel step with the worst unexpected drop
  • Checks to run next, and a score out of 100 with its sum
• A/B test design planner: Plan a fair split-run test of two ads, emails or pages. Get one change, a key on each, the audience size and a winner rule.
  • The one difference being tested
  • What must stay equal and how the audience is split
  • A key (code) for each version
  • The audience size needed, with the sum shown
  • A winner rule with a read date, plus a plan for more than two versions
  • The traps to avoid
• Landing page draft and test planner: Draft and review a landing page. Plan a test of one change, then use real results to make the next call.
  • A page plan with copy, layout rules and one clear button
  • A ranked list of fixes with a score
  • A test design with the claim, the numbers to watch, the sample size and the run time
  • A ship, wait or stop call with the numbers behind it
• Simulated user walkthrough reviewer: Get a persona walkthrough of your web page. See what a visitor thinks at each scroll, where they doubt or leave, and the fixes ranked.
  • The visitor profile written back, so you can fix it before the walkthrough
  • A five-second test with a clear yes or no on each of the three questions
  • One entry per screen, with the visitor's thoughts and a separate expert note
  • A verdict with three scores out of 10, plus the moment they almost left and the moment they were most engaged
  • Fixes ranked as quick wins, bigger changes and long-term ideas
  Source: Michael Sitarzewski and Agency Agents contributors.
• Page issue sorter: Sort a long list of page problems into fix now, investigate, test or tracking, then rank them. Get a clear plan for your page issue list.
  • Each issue in one bucket with its reason
  • Test ideas written as: change, expected effect, why
  • A ranked order by likely gain and effort
  • A list of issues dropped for lack of proof
• 30-point landing page checker: Share your page and screenshots. Get a 30-point audit, a score and five fixes to make first.
  • A two-line verdict naming the stop where the page loses buyers
  • All 30 checks with yes, no or can't tell, the evidence and a one-line fix
  • A count of no's at each stop, so you see where the trouble clusters
  • The five fixes to make first, deal-breakers at the top, and what it couldn't judge
  • A checks score out of 100 (the share of checks the page passes), now and with the five fixes made
• Brand profile builder: Share your website. Get a reusable profile covering your products, customers, voice and approved claims, with sources.
  • A folder of eight plain files, from a start file for your AI to a claims list and a change log
  • A tag on every fact, and every claim marked approved, pending or banned
  • A two-sentence diagnosis of your biggest gap, and a list of what's missing and who can answer it
  • A short slice for your next step, if you name one
• Peep Laja: Review page messaging or an A/B test using Peep Laja’s recorded approach. Start with buyer evidence, clarity and a sound comparison plan.
  • Your problem worked on the way Peep Laja would: their questions, their method, their push-back
  • The answer or the work, in Peep Laja's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• Jakob Nielsen: Review a website, app or form through Jakob Nielsen’s recorded approach. Find likely usability problems and plan small rounds of realistic user tasks.
  • Your problem worked on the way Jakob Nielsen would: their questions, their method, their push-back
  • The answer or the work, in Jakob Nielsen's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• Your page gets visits but few sales and you don't know where to start.
• You got a long list of page problems and need to know what to fix first.
• You fixed some things and want to know if the page is better.

Limits to check before starting

• Your page has no visitors yet: fix the ad or the traffic first, then come back.
• You only want a review and no second check: use the Landing page conversion reviewer.
• You only want the 30 yes-or-no checks on a page you already fixed: use the 30-point landing page checker.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Download the handoff pack. It's a zip with HANDOFF.md and every skill file this workflow uses.
• Give the folder to Claude (in Claude Code, or as files in a Claude Project) and say "Follow HANDOFF.md".
• Give it the page and, if you have it, the ad that sends traffic. Screenshots of the phone view help most.
• Make only the fix-now changes. Leave the test ideas for later, so you can tell what helped.
• Save the updated page and give it back for the last check.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/cro-specialist for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.