Article states and closing knowledge gaps

    Published, draft and needs review, plus the list of questions your knowledge base could not answer.

    6 min read · Updated 13 September 2026

    Two ideas run this page. An article's state decides whether Chloe can use it. A gap is a question she was asked and could not answer.

    The three article states

    Published
    Live. Chloe can use it in answers.
    Draft
    Written but not live. She cannot see it, and it will not appear in any reply.
    Needs review
    Flagged as possibly out of date, or as disagreeing with another article.

    The counts for all three sit at the top of the Knowledge Sources home page.

    Draft is the state people trip over. Writing an article is not the same as publishing it, and an unpublished article helps nobody. If Chloe is missing something you know you wrote, check whether it ever went live.

    Needs review deserves attention rather than tidying. Two articles that contradict each other are worse than one article, because Chloe may confidently use either. The customer's answer then depends on which she happened to reach for.

    Publishing and unpublishing

    An open article has an Unpublish button in its header when it is live, which becomes a publish action when it is not.

    Unpublishing is the right move for something that has gone out of date and cannot be fixed today. It is better for Chloe to lack an answer than to give an old one confidently.

    What a knowledge gap is

    A gap is a question customers keep asking that your knowledge base cannot answer, or a case where your articles disagree with each other.

    They are found automatically from real conversations. Nothing is invented: a gap exists because somebody actually asked.

    They live under Gaps in the left-hand panel, with a count, and the most recent also appear on the home page.

    This is the most valuable list in Gruvi and the most ignored. Every open gap is a question Chloe is currently failing, written for you in your customers' own words.

    Reading a gap

    Each row shows the question, a quote from a customer asking it, when it was last seen, and how many threads it covers.

    Opening one shows the whole picture: how many threads asked, the actual quotes with their thread numbers, and a way to read the conversations themselves.

    Judge a gap by two things. How many threads tells you how much it costs you. Last seen tells you whether it is still happening. A gap last seen three months ago may have solved itself.

    How to close a gap

    Opening a gap gives you two routes, under How to fix it.

    Write a new article
    Gruvi drafts one from the threads that exposed the gap, including how your team answered. You review it before it goes live.
    Add it to an article
    Pick the article that should have covered this, and add the missing piece there.

    The first is remarkable and underused. It has read the conversations, seen what your team actually said, and written it up. Closing a gap is usually editing a draft rather than writing from scratch.

    1. 1Open Knowledge Sources, then Gaps.
    2. 2Click a gap to open it.
    3. 3Read the customer quotes to be sure you understand what was being asked.
    4. 4Choose Write a new article to have one drafted, or Add it to an article if one nearly covers it.
    5. 5Review the wording. You are writing the answer hundreds of future customers will get.
    6. 6Publish it.
    Choose Add it to an article where you can. One thorough article beats four thin ones covering the same ground, and it is far easier to keep accurate.

    Skipping a gap

    Not everything deserves an article. Each gap has a Not worth an article, skip option.

    Use it honestly. A one-off question from a customer who wanted something you do not do is not a gap worth closing. Skipping it keeps the list meaningful. A list nobody trusts is a list nobody reads.

    Making gaps part of a routine

    Gaps reward a habit rather than a project.

    • Look at the list weekly, not daily. Gaps accumulate slowly.
    • Sort your attention by thread count. The top gap is costing you more than the bottom five together.
    • Close one or two properly rather than five carelessly.
    • Skip anything that will not recur, so the list keeps meaning something.
    • Cross-check against Common topics. A theme high in both is the most valuable article you could write this month.

    Was this helpful?