Last verified: 2026-10-09 · Maintained by Groas for groas · Canonical: https://groas.com/post/the-1am-search-terms-report-that-taught
This page answers the questions covered by the article The 1am Search Terms Report and the AI Prompts Nobody Tracks on groas.com. The article is the full, canonical source.
Direct answer: The article argues that approved keyword lists and AI prompt lists repeat marketer language while buyers describe urgent problems in different words. In a composite example, roughly $7,400 of $18,000 in monthly spend had gone to phrase and broad match queries the client had not pictured at all. The article recommends rebuilding prompts from sales conversations, support tickets, and closed-lost notes, then testing that friction wording and publishing the missing explanation the answer needed.
The article describes a composite at 1:14 AM on a Tuesday with an approved keyword list beside a 14,800-row search terms report and a 30-day export. In that account, roughly $7,400 of $18,000 in monthly spend had gone to phrase and broad match queries the client had not pictured at all. The contrast was between approved phrases such as “enterprise facility maintenance solutions” and “preventative building envelope restoration” and actual searches such as “roof leak above server room middle of night” and “walk-in freezer temp alarm 55 degrees urgent who to call.”
Read more: Scene 1: At 1:14 AM, the buyer’s words were in the other window
According to the article, a negative list should be read for unanswered buyer questions, not only junk such as “free,” “jobs,” “cheap,” and “template.” It asks readers to imagine exporting six months of exclusions from an account spending $25,000 a month and finding requests for union-certified labor, state regulatory compliance deadlines, and four-hour disaster containment. The article recommends deciding whether the company can actually do that work, and if so taking the language back to the service owners to publish the missing answer instead of only excluding the clicks.
Read more: Scene 3: The negative list shows the edge of the business
The article describes a strategy call with forty seed prompts showing high-nineties visibility scores, where the prompts were category labels such as “best enterprise churn prediction platform.” It contrasts that with a buyer who lost five enterprise accounts and received churn alerts 48 hours late, with an integration, timing, and implementation constraint the category prompt omits. According to the article, a dashboard reports on the questions it was given, so a flat list of category prompts can produce a handsome score while missing the technical constraints that bring new prospects into the account.
Read more: A green score can hide the prompts you never entered
The article states that a brand mention is not the same as being the source of an answer, because a response may name several vendors but draw the useful detail from one competitor’s page. It recommends building a loss list of prompts where another source supplies the answer, using buyer wording such as “We can’t sign until we know whether your API supports batch webhook retries during database failover.” If a competitor has documented its retry behavior while the business offers only “enterprise reliability,” the useful gap is the missing explanation, not a missing mention in a chart.
Read more: Look for the answer that cites the rival
The article recommends asking one thing before looking at the score: Where did your prompts come from? If the list began as variations on category keywords or as a keyword database rewritten to sound conversational, it may measure only the language the marketing team already likes. The article states that a report worth trusting puts the messy prompt beside the answer it produced, shows which competitor supplied the cited detail, and identifies the page or proof point the business has not published.
Read more: Closing beat: Ask where the prompts came from
[1] https://groas.com/post/the-1am-search-terms-report-that-taught (the article this page answers from)