Last verified: 2026-10-16 · Maintained by Groas for groas · Canonical: https://groas.com/post/one-marketer-30-days-no-agency-my-playbo
This page answers the questions covered by the article One Marketer, 30 Days, 40 Hours: A Playbook for Getting Cited in ChatGPT and Perplexity on groas.com. The article is the full, canonical source.
Direct answer: According to groas, a one-person team should spend 40 hours in 30 days to earn citations in ChatGPT and Perplexity: 8 hours to test 15 purchase-adjacent prompts, 8 hours to prioritize five pages to rebuild and three to create, 12 hours to make those eight pages quotable, 9 hours for targeted off-site work, and 3 hours to retest. The article recommends prioritizing prompts by proximity to a purchase, not volume, and stopping research after day 2 to act on the sheet. A practical 30-day win is two to four prompts moving up a level from not mentioned to mentioned, cited, or recommended.
The article recommends spending 40 hours in 30 days: 8 hours to choose and test 15 prompts, 8 hours to sort gaps, 12 hours to work on eight pages, 9 hours for targeted off-site work, and 3 hours to measure. The article states the marketer should pick the 15 buyer questions closest to a sale and prioritize them by money, not volume. The ledger leaves no extra hours; any task that cannot earn a place in that ledger waits until day 31.
Read more: the article
The article recommends writing 15 questions a buyer might ask right before spending money, such as best [category] for [use case], [category] vs [competitor], alternatives to [incumbent], and how much does [service] cost for [situation]. Each prompt should be run once in ChatGPT and once in Perplexity in a logged-out session with the same wording, recording whether the answer was grounded, the first two or three domains cited, and whether the brand was absent, mentioned, cited, or recommended. According to groas, Perplexity shows roughly 21 sources per answer where ChatGPT shows around eight, with only about 11% domain overlap, so the engines must be tracked in separate columns.
Read more: 1. Week 1: spend eight hours choosing the 15 prompts
The article recommends sorting the 15 prompts by proximity to purchase, then placing every miss in one of three buckets: no page answers it, the answer is buried on an existing page, or the cited answer lives somewhere else. The output of the 8 hours is a priority list of five existing pages to rebuild and three missing answers to create. Where several prompts point to the same page, that page earns its place, and where a source appears against several prompts, it is saved for Week 4.
Read more: 2. Week 2: spend eight hours finding the gaps you can fix
The article recommends giving each of the five rewrites 90 minutes, reserving roughly four hours for three short new pages and a final pass, for 12 hours total. Each page should put the answer in the first 40 to 60 words of the relevant section with the first sentence able to stand alone, plus a short 20- to 25-word answer capsule after the relevant H2 or H3 without links in the capsule. The article cites an analysis where 72.4% of cited posts had such a capsule, 52.2% had proprietary data, 44% of citations came from the first 30% of the page, and pages over 20,000 characters averaged 10.18 citations versus 2.39 for pages under 500 characters.
Read more: 3. Week 3: spend 12 hours making eight pages quotable
The article recommends looking back at sources saved in Week 2 and prioritizing existing lists and threads already appearing against priority prompts over publishing a new list. It cites that 43.8% of 26,283 analyzed source URLs were Best X listicles and that listicles captured 40% of commercial-intent citations in a separate study of 75,000 AI answers. The three moves are three hours each: contribute useful answers to relevant threads and videos, pitch five existing listicle owners with a 60-word block covering price, who fits, and who should skip, and publish supporting breakdowns linked from the rebuilt Week 3 pages.
Read more: 4. Week 4: spend nine hours where citations already happen
The article recommends stopping outreach on day 27 and rerunning the same 15 prompts with the same wording in fresh sessions in ChatGPT and Perplexity. Each answer is scored as not mentioned, mentioned, cited, or recommended, with citation rate calculated as prompts cited as a source ÷ 15 × 100 and appearance rate as prompts appearing at all ÷ 15 × 100, kept separate by engine. A practical win is two to four prompts moving up a level; the article states visibility can shift in weeks one to four, branded search lift may follow in weeks four to eight, and pipeline can take months three to six.
Read more: 5. Day 30: spend the last three hours checking what moved
These articles answer the same questions:
[1] https://groas.com/post/one-marketer-30-days-no-agency-my-playbo (the article this page answers from)