---
title: "One Person, 30 Days, Zero Budget: A 10-Page AI Citation Playbook"
description: "A no-spend plan to test 10 buyer questions, make your site readable, publish 10 answer pages, earn three outside mentions, and measure what ChatGPT and Perplexity cite."
url: "https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/611a5e99-be50-4924-8584-a29462b66401.png"
published: "2026-10-26T05:48:12.231Z"
modified: "2026-10-26T05:48:12.288Z"
---

October 26, 2026 · 10 min read

# One Person, 30 Days, Zero Budget: A 10-Page AI Citation Playbook

[Alexander PerelmanHead Of Product @ groas](https://groas.com/author/alexander-perelman)[LinkedIn](https://www.linkedin.com/in/alexander-433793253/)

![A tiny shopkeeper on a popsicle-stick ladder hoists a cardboard footnote marker onto a giant book, earning her small business a spot as a cited source.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/611a5e99-be50-4924-8584-a29462b66401.png)

In this article

1. [1. Week 1: choose 10 buyer questions and save the baseline](#1-week-1-choose-10-buyer-questions-and-save-the-baseline)
2. [2. Week 2: fix access before writing the answers](#2-week-2-fix-access-before-writing-the-answers)
3. [3. Week 3: publish 10 pages, one answer per question](#3-week-3-publish-10-pages-one-answer-per-question)
4. [4. Week 4: earn three mentions on pages the engines already use](#4-week-4-earn-three-mentions-on-pages-the-engines-already-use)
5. [5. Day 30: run the same questions and score citations, not compliments](#5-day-30-run-the-same-questions-and-score-citations-not-compliments)
6. [6. If the month shrinks to two weeks, cut the count](#6-if-the-month-shrinks-to-two-weeks-cut-the-count)
7. [7. When the constraint lifts, fund repetition, not another checklist](#7-when-the-constraint-lifts-fund-repetition-not-another-checklist)

One person, 30 days, zero budget. I would spend none of it on an AEO checklist unless I could trace the work back to a possible citation in ChatGPT or Perplexity.

That leaves four jobs: pick 10 buyer questions and record what the engines cite now; make sure they can fetch your site; publish one answer page per question; and earn three outside mentions on pages they already use. On Day 30, run the questions again. I used to tell small clients to start with a full content audit and a schema project. I was wrong. Both can feel productive while leaving the thing you need, a page an engine can find, quote, and corroborate, untouched.

**The constraint is the editor.** Ten questions, 10 pages, three mentions, one before-and-after sheet. Everything else waits for month two.

## 1. Week 1: choose 10 buyer questions and save the baseline

**Cost: 3 hours and a spreadsheet. Buys: a starting line you can check on Day 30.** Give the sheet columns for prompt, engine, cited domains, cited URLs, and whether your business appears. Write exactly 10 questions a buyer might ask across early research, comparison, and ready-to-buy decisions. Stay close to the services you can actually describe and the places you actually serve. A broad question that returns national publishers may be interesting; it is not automatically the best use of this month.

Run each question logged out in ChatGPT with web search on and in Perplexity. Copy the cited URLs, not just the names in the answer. Use the same wording when you return on Day 30. That same-prompt, Day 0 versus Day 30 loop is the [free baseline method that fits a zero budget](https://searchengineland.com/guide/citation-core). I run the first pass in one sitting rather than relying on memory to reconstruct what an answer looked like later.

Now read the citations as a work list. Which questions pull up service pages? Which pull up directories, forums, or comparisons? Which outside pages appear more than once? A large [analysis of 30 million cited sources found Reddit the most-cited domain across ChatGPT, Perplexity, and Google answers, followed by YouTube, LinkedIn, Wikipedia, and Forbes](https://searchengineland.com/ai-search-engines-cite-reddit-youtube-and-linkedin-most-study-473138). Perplexity leaned harder toward Reddit, LinkedIn, and G2 for B2B; ChatGPT favored Wikipedia, Reddit, and Forbes. That is useful context, not permission to spend the week posting on Reddit. **Your own 10 questions decide where you work.** Circle the outside pages that show up in your results. You will need them in Week 4.

## 2. Week 2: fix access before writing the answers

**Cost: 2 hours. Buys: pages the engines can reach and read.** A quotable answer does nothing if it sits behind a blocked crawler or renders as an empty shell. This is the unglamorous step I would do before touching a headline.

[Perplexity runs a live web search for every question, so a new page can be cited once crawled without waiting for model retraining](https://pages.citepoint.ai/articles/how-perplexity-chooses-its-sources). It reads results and attaches numbered citations to the pages it uses. For ChatGPT with browsing on, [OAI-SearchBot and GPTBot have different jobs](https://getrevised.com/knowledge/ai-and-geo/how-to-get-cited-by-chatgpt): the former concerns search discovery; the latter concerns training. Do not treat a decision about GPTBot as a search-access test. If you want a shot at citations, check the fetchers used for search.

I would make this a short, ordered check, not a sitewide technical audit:

1. Open `robots.txt`. Check access for PerplexityBot, OAI-SearchBot, and ChatGPT-User. [PerplexityBot builds search results; Perplexity-User is a separate on-demand fetcher](https://pages.citepoint.ai/articles/how-perplexity-chooses-its-sources). Note a block, fix it if you control it, and test again.
2. Load your three most important service pages in a text-only fetch. View source or use a free mobile-friendly test. Look for the price, service area, and one proof point as readable text rather than information stranded in an image or JavaScript.
3. If a page fails, fix the block or render and repeat the fetch. Use the fuller check in [_Before You Write for ChatGPT, Check Whether AI Can Read Your Site_](https://groas.com/post/before-you-write-another-page-for-chatgp), then stop.

I am deliberately cutting server-log monitoring and a full crawl. They may tell you more about how bots behave, but neither rescues an answer that cannot be fetched today. **No fetch, no citation opportunity.** Get the essential text reachable and move on to writing.

## 3. Week 3: publish 10 pages, one answer per question

**Cost: about 15 hours at 90 minutes per page. Buys: 10 specific answers an engine can quote.** This is the expensive week, which is why the first two weeks must stay small. Five pages would be easier, but it would leave half the questions in the baseline without an answer page. Under this plan, each of the 10 gets one.

Use the question as the page’s H1. Put a direct, 40-to-60-word answer inside the first 100 words. Then use two natural follow-up questions as H2s. Do not bury the price or service area beneath a long introduction about your commitment to excellence. The structure gives a reader an answer quickly and puts the most useful material where an engine is likely to encounter it: [44.2% of citations in one analysis came from the first 30% of page text](https://keytomic.com/blog/llm-citations-checklist).

My 90-minute page has a tight job list:

1. **Answer the buyer’s question first.** Say what the service is for, the relevant price range if you can state one, and the area you serve. If the answer depends on circumstances, name those circumstances plainly.
2. Add two follow-up answers that help the buyer judge the choice. Say who should use the service and who should skip it. That is more useful than another paragraph insisting the business is trustworthy.
3. Include one dated fact with a live source link beside it, plus a visible Last Updated date. Pages with named statistics and links received [about 2.1 times more citations in one analysis](https://keytomic.com/blog/llm-citations-checklist). A number an engine can place in context is more useful than ‘quality work at fair prices.’
4. Publish, then fetch the page as text. If the answer disappeared between the editor and the live page, Week 2 was not finished.

The price and fact must be ones you can stand behind. Do not invent a neat local price range because a neat sentence looks quotable. The point is to make real business information easy to lift, not to manufacture precision. Equally, do not turn every question into a 2,000-word guide. If two questions need the same supporting detail, keep their direct answers distinct and use that detail where it helps. **One page, one question, answer up front.** That is the publishing rule for this week.

## 4. Week 4: earn three mentions on pages the engines already use

**Cost: 6 hours of writing and outreach. Buys: outside context your own pages cannot provide.** Your site can explain what you do. An independent page can show an engine how you fit among the options it already sees. I cap the work at three mentions because ten thin listings are not a substitute for three relevant placements.

Go back to the URLs in your Week 1 sheet. Use [_11 Off-Site Pages ChatGPT and Perplexity Cite, and How to Get on Them_](https://groas.com/post/11-places-chatgpt-and-perplexity-find-so) to identify the kind of page you are looking at, but let your recorded results pick the targets. I would spend the six hours on these three moves, in order:

1. **Fix one profile** on a directory the engines cited for your questions. Make its hours, prices, and service area agree with your answer pages. An accurate profile is a better use of the hour than another listing the sheet never showed you.
2. **Answer one real thread** on Reddit or a forum that appears in your prompt results. Disclose that you run the business. Give the answer, including a relevant number you can support, in the first two sentences. Do not paste a sales pitch into a question that did not ask for one.
3. **Pitch one comparison correction or addition** on a page that cited competitors. Supply a dated fact and a link to the answer page that supports it. A specific correction gives the page owner something to evaluate; ‘please include us’ gives them work.

Reddit deserves a place here only if your prompts justify it. One panel found Reddit in [68.1% of Perplexity answers but only 4.45% of all Perplexity citations, and in 2.9% of ChatGPT answers in that capture](https://kollavo.com/blog/does-reddit-get-cited-by-ai). Those figures are a good warning against calling any platform a universal shortcut. **Target pages your buyer questions already surface.** Cut the extra directory profile, generic forum posting, and broad outreach list; none has earned a place in this 30-day budget.

## 5. Day 30: run the same questions and score citations, not compliments

**Cost: 2 hours. Buys: an honest result.** Run the identical 10 questions, logged out, with the same wording in both engines. Record brand mentions, cited domains, and cited URLs beside Day 0. A win is a citation with your URL attached. A friendly mention without a link is worth noting, but it is not the result this playbook set out to produce. Neither is a Google ranking change.

![Spreadsheet comparing AI citations for the same questions on Day 0 and Day 30](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/b865d616-ac8a-43e4-a643-a55a5758681d.png)

Group the results into questions where you went from absent to cited, questions where a competitor still holds the citation, and questions where nothing changed. Perplexity may move first because a new page can be used once crawled; ChatGPT depends on its search index catching up. One to three questions moving in 30 days would be a real result for one person with no spend. Zero is information too, but it does not prove that the writing format failed or that the outside mentions alone were too thin.

If a page wins, keep it current: change its Last Updated date when a price or fact changes, add the next follow-up question buyers ask, and repeat the outside placement that proved useful. If nothing moves, do not rewrite all 10 pages on instinct. Recheck the fetch. Then check whether the outside pages you targeted still appear for those questions. **Let the sheet choose the next repair**, not your opinion of which page sounds best.

## 6. If the month shrinks to two weeks, cut the count

**Cost: about 14 hours instead of 28. Buys: a smaller test you can actually finish.** Keep the order; reduce the scope to **five questions, three pages, one fetch fix, and one outside mention**. Choose the five questions within one service and one town or one use case. Run the Day 0 check. Fix access and rendering for the pages that matter to that small set. Publish three answers at 90 minutes each, then work on the single outside page your results make most relevant. Re-run the same five questions at the end.

What gets cut? Two planned answer pages, two outside mentions, the extra profile, schema cleanup, and rank tracking beyond the sheet. I would rather know what happened to three live pages than finish two drafts and call the sprint a content strategy. Do not cut the baseline or the fetch check: without the first, you cannot tell whether anything moved; without the second, the pages may never get a fair shot. **Cut volume before you cut sequence.**

## 7. When the constraint lifts, fund repetition, not another checklist

**Cost under this plan: no extra hours or spend. Buys: a clear decision about whether to run another month.** Count the work before you start: three hours for the baseline, two for access, 15 for 10 pages, six for three mentions, and two for the re-run. That is 28 hours in a month, roughly seven a week. If you do not have them, use the two-week version rather than stalling in Week 3.

Skip the DIY page sprint if you cannot publish to the site, if your questions return only national press and government pages, or if every answer needs a licensed professional. In that position, I would put the available time into one accurate profile and three reviews with specific service details, not 10 pages I cannot publish or responsibly answer. That still respects the one-person, no-spend constraint; it does not pretend the page plan fits every business.

When more time or budget becomes available, do not buy a bigger version of the same manual checklist. Re-run the questions weekly, refresh dates and prices when they change, and keep earning relevant outside mentions. That repetition is the work groas was built to run continuously, with models handling execution and a named strategist holding the direction. **Use the first 30 days to find what moves.** Then give month two to doing more of that work without asking one person to live in the spreadsheet.

## Frequently Asked Questions

### What is the 30-day plan to get my site cited by ChatGPT and Perplexity with no budget?

The plan has four jobs: pick 10 buyer questions and record what the engines cite now, make sure the engines can fetch your site, publish one answer page per question, and earn three outside mentions on pages they already use. On Day 30, run the same questions again and compare.

### How do I set up a baseline for AI citations before making changes?

Write exactly 10 questions a buyer might ask across early research, comparison, and ready-to-buy decisions. Run each one logged out in ChatGPT with web search on and in Perplexity, and copy the cited URLs into a spreadsheet. Use the same wording again on Day 30 to see what changed.

### How do I check whether AI engines can actually read my site?

Open robots.txt and check access for PerplexityBot, OAI-SearchBot, and ChatGPT-User, then fix any block you control. Next, fetch your three most important service pages as text-only and confirm the price, service area, and one proof point appear as readable text rather than inside an image or JavaScript.

### How should I structure an answer page so ChatGPT or Perplexity can quote it?

Use the buyer's question as the page H1 and put a direct 40-to-60-word answer inside the first 100 words. Add two natural follow-up questions as H2s, one dated fact with a live source link, and a visible Last Updated date. One page should answer one question.

### What outside mentions should I pursue to get cited by AI engines?

Target the outside pages your own recorded prompt results already show, capped at three mentions. Fix one directory profile so its hours, prices, and service area match your answer pages, answer one real thread on Reddit or a forum while disclosing you run the business, and pitch one specific comparison correction with a dated fact and link.

### How do I measure whether my AI citation efforts actually worked after 30 days?

Run the identical 10 questions, logged out, with the same wording in both engines and record brand mentions, cited domains, and cited URLs beside the Day 0 results. A win is a citation with your URL attached; one to three questions moving in 30 days would be a real result for one person with no spend.

### What should I cut if I only have two weeks instead of 30 days?

Reduce the scope to five questions, three answer pages, one fetch fix, and one outside mention, keeping the same order of work. Do not cut the baseline or the fetch check: without the first you cannot tell whether anything moved, and without the second the pages may never get a fair shot.

## Paid Ads & SEO, Run For You

Businesses pay one flat monthly fee per service, agencies resell it under their own brand, and groas does the work either way.

[See Pricing](https://groas.com/pricing)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","description":"groas is a done-for-you marketing service, not software: elite people working with proprietary technology systems run search ads, Meta ads, SEO and AI search for businesses, and white label for agencies, with a dedicated senior strategist owning every account.","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#webpage","url":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t","name":"One Person, 30 Days, Zero Budget: A 10-Page AI Citation Playbook","description":"A no-spend plan to test 10 buyer questions, make your site readable, publish 10 answer pages, earn three outside mentions, and measure what ChatGPT and Perplexity cite.","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/611a5e99-be50-4924-8584-a29462b66401.png"},"breadcrumb":{"@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#breadcrumb"},"dateModified":"2026-10-26T05:48:12.288Z","mainEntity":{"@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"One Person, 30 Days, Zero Budget: A 10-Page AI Citation Playbook","item":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#article","url":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t","mainEntityOfPage":{"@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"One Person, 30 Days, Zero Budget: A 10-Page AI Citation Playbook","description":"A no-spend plan to test 10 buyer questions, make your site readable, publish 10 answer pages, earn three outside mentions, and measure what ChatGPT and Perplexity cite.","datePublished":"2026-10-26T05:48:12.231Z","dateModified":"2026-10-26T05:48:12.288Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/611a5e99-be50-4924-8584-a29462b66401.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"pages, questions, answer, page, three, outside, use, chatgpt","wordCount":2097,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/alexander-perelman#person","name":"Alexander Perelman","jobTitle":"Head Of Product @ groas","description":"My background includes Goldman Sachs and Stanford Computer Science.","image":"https://groas.com/media/blog/4f17cac3a81acc224cc1b2cfab8f94bbc295a3089d13e8421125771ef8d4064f.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"},"sameAs":["https://www.linkedin.com/in/alexander-433793253/"]},{"@type":"FAQPage","@id":"https://groas.com/post/one-person-30-days-zero-budget-how-i-d-t#faq","mainEntity":[{"@type":"Question","name":"What is the 30-day plan to get my site cited by ChatGPT and Perplexity with no budget?","acceptedAnswer":{"@type":"Answer","text":"The plan has four jobs: pick 10 buyer questions and record what the engines cite now, make sure the engines can fetch your site, publish one answer page per question, and earn three outside mentions on pages they already use. On Day 30, run the same questions again and compare."}},{"@type":"Question","name":"How do I set up a baseline for AI citations before making changes?","acceptedAnswer":{"@type":"Answer","text":"Write exactly 10 questions a buyer might ask across early research, comparison, and ready-to-buy decisions. Run each one logged out in ChatGPT with web search on and in Perplexity, and copy the cited URLs into a spreadsheet. Use the same wording again on Day 30 to see what changed."}},{"@type":"Question","name":"How do I check whether AI engines can actually read my site?","acceptedAnswer":{"@type":"Answer","text":"Open robots.txt and check access for PerplexityBot, OAI-SearchBot, and ChatGPT-User, then fix any block you control. Next, fetch your three most important service pages as text-only and confirm the price, service area, and one proof point appear as readable text rather than inside an image or JavaScript."}},{"@type":"Question","name":"How should I structure an answer page so ChatGPT or Perplexity can quote it?","acceptedAnswer":{"@type":"Answer","text":"Use the buyer's question as the page H1 and put a direct 40-to-60-word answer inside the first 100 words. Add two natural follow-up questions as H2s, one dated fact with a live source link, and a visible Last Updated date. One page should answer one question."}},{"@type":"Question","name":"What outside mentions should I pursue to get cited by AI engines?","acceptedAnswer":{"@type":"Answer","text":"Target the outside pages your own recorded prompt results already show, capped at three mentions. Fix one directory profile so its hours, prices, and service area match your answer pages, answer one real thread on Reddit or a forum while disclosing you run the business, and pitch one specific comparison correction with a dated fact and link."}},{"@type":"Question","name":"How do I measure whether my AI citation efforts actually worked after 30 days?","acceptedAnswer":{"@type":"Answer","text":"Run the identical 10 questions, logged out, with the same wording in both engines and record brand mentions, cited domains, and cited URLs beside the Day 0 results. A win is a citation with your URL attached; one to three questions moving in 30 days would be a real result for one person with no spend."}},{"@type":"Question","name":"What should I cut if I only have two weeks instead of 30 days?","acceptedAnswer":{"@type":"Answer","text":"Reduce the scope to five questions, three answer pages, one fetch fix, and one outside mention, keeping the same order of work. Do not cut the baseline or the fetch check: without the first you cannot tell whether anything moved, and without the second the pages may never get a fair shot."}}]}]}
```