{"id":22738,"date":"2026-10-07T08:20:48","date_gmt":"2026-10-07T12:20:48","guid":{"rendered":"https:\/\/www.data-mania.com\/blog\/?p=22738"},"modified":"2026-10-07T22:32:55","modified_gmt":"2026-10-08T02:32:55","slug":"thrivestack-citedby-review","status":"publish","type":"post","link":"https:\/\/www.data-mania.com\/blog\/thrivestack-citedby-review\/","title":{"rendered":"How I Used ThriveStack CitedBy to Boost Search Impressions 4x in 6 Days"},"content":{"rendered":"<p>If you pay for an AI visibility tool, or you&#8217;re deciding whether to buy one, this review is for you. I spent almost a month running CitedBy, the AI visibility tool from <a href=\"https:\/\/linkly.link\/2vOlu\" target=\"_blank\" rel=\"noopener\">ThriveStack<\/a>, on my own site and brand. In this article you&#8217;ll see:<\/p>\n<ul>\n<li>The AI search results that ThriveStack drove<\/li>\n<li>Where ThriveStack CitedBy outperforms the other tools in my stack<\/li>\n<li>The off-site problem that no tool in this category has solved yet &#8211; specifically how third-party sites borrow your authority without recommending you<\/li>\n<li>Six cautions to know before you buy an AI visibility tool<\/li>\n<\/ul>\n<p>By the end, you&#8217;ll know how to use a tool like ThriveStack CitedBy to pick the content targets that are most likely to move the needle, how to tell a real result from random variation, and which AI search numbers you can safely report to a client or a board.<\/p>\n<p>While I\u2019ll go into more detail later in this post, the top-line results I\u2019m reporting are:<\/p>\n<ul>\n<li><strong>Search impressions<\/strong> on one of my blog posts went from 13 to 59 in 6 days.<\/li>\n<li><strong>The average search position<\/strong> of a second post moved from 46.8 to 18.4 over 6 days, which on average took it from roughly page 5 of Google to page 2.<\/li>\n<\/ul>\n<p>These results were measured based on Google Search Console and what Google actually served.<\/p>\n<p>How <a href=\"https:\/\/linkly.link\/2vOm3\" target=\"_blank\" rel=\"noopener\">ThriveStack CitedBy<\/a> helped is that it pointed me to both of those posts. The tool produced 72 ranked, evidence-backed content gaps for my site. Each gap is a question that buyers ask AI engines where my site was missing, and each one named the competitors that the engines cited instead.<\/p>\n<p>And just so you know, before I credited ThriveStack CitedBy with anything, I asked whether the result could be random movement. I cut every result that could be (if you want to learn more about why AI visibility scores are unreliable in general, my article on the <a href=\"https:\/\/www.data-mania.com\/blog\/ai-visibility-score-noise-proxy-stack\/\">AI citation measurement crisis<\/a> walks you through the full data).<\/p>\n<p>One disclosure before you go further: ThriveStack has paid me to give them product feedback on ThriveStack CitedBy. You can weigh everything below with both of those facts in mind.<\/p>\n<p>The following is a diary of how I got results with CitedBy and what I learned along the way.<\/p>\n<h2>The emergence of authority-borrowing, and how it can help or hurt your brand<\/h2>\n<p>By working with CitedBy long enough, I exposed a new phenomenon that I call \u201cauthority-borrowing\u201d. I\u2019m sure my own website demonstrates many instances of this practice, although it was unintentional as I was completely unaware of this as a practice until I worked with CitedBy.<\/p>\n<p>Authority-borrowing is a tactic in AI search where a third-party website leverages your expertise, content, website authority, and citations to establish its own credibility and rank in AI-generated answers, in some cases directing the actual brand recommendation and traffic to themselves or someone else.<\/p>\n<p>Here is how the entire process works:<\/p>\n<ol>\n<li><strong>You Publish Original Expertise<\/strong>: You create high-quality, authoritative content on your own site answering specific buyer questions.<\/li>\n<li><strong>Third-Party Sites Extract and Cite<\/strong>: Other websites use AI or manual curation to pull content or reference your articles as a foundational source (giving you a citation to look credible and authoritative to search crawlers).<\/li>\n<li><strong>The AI Engine Pulls the Third-Party Page<\/strong>: When an AI engine answers a user&#8217;s prompt, it retrieves that third-party list page or article because it cites you and appears comprehensive.<\/li>\n<li><strong>The Third-Party or Another Competitor Gets the Recommendation<\/strong>: Even though your brand&#8217;s research or article was used as the supporting source to give the page authority, the third-party page can ultimately recommend a competitor in its final ranking list.<\/li>\n<li><strong>You May Remain Invisible<\/strong>: The AI engine follows the recommendation of the list page, meaning your authority was &#8220;borrowed&#8221; to validate the page, but the third-party or other competitor collects the recommendation, traffic, and prospective business while you stay invisible on your own topic.<\/li>\n<\/ol>\n<h3>2 examples of authority-borrowing in the wild<\/h3>\n<p>When it comes to comparison pieces specifically (e.g.; \u201c10 on its Best Fractional CMO Companies\u201d), there are two types of authority-borrowing, beneficial and malicious authority-borrowing.<\/p>\n<ul>\n<li><strong>Beneficial authority-borrowing<\/strong> is when a third-party site cites your brand\/content to establish credibility and actively ranks or recommends you within their top suggestions.<\/li>\n<li><strong>Malicious authority-borrowing<\/strong> is when a third-party site uses your content or authority to validate its page with search engines, but directs the actual recommendation and referral traffic to themselves or another competitor.<\/li>\n<\/ul>\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" data-src=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/thrivestack-citedby-review-02.jpg\" alt=\"Infographic titled How authority-borrowing helps or hurts your brand, showing five steps (you publish, they cite, AI retrieves, it recommends, value flows) that split into a beneficial outcome where your authority helps you and a malicious outcome where it helps a competitor.\" src=\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\" class=\"lazyload\" style=\"--smush-placeholder-width: 1402px; --smush-placeholder-aspect-ratio: 1402\/1122;\" \/><figcaption class=\"wp-element-caption\">How authority-borrowing works and splits into beneficial or malicious outcomes for brand visibility.<\/figcaption><\/figure>\n<p><a href=\"https:\/\/linkly.link\/2vOm4\" target=\"_blank\" rel=\"noopener\">CitedBy<\/a> makes it easy to expose cases in each category. The table below shows the clearest confirmed examples I have.<\/p>\n<figure class=\"wp-block-table\">\n<table>\n<thead>\n<tr>\n<th>Type of authority-borrowing<\/th>\n<th>Domain<\/th>\n<th>Cites data-mania.com as a source<\/th>\n<th>Lists Data-Mania or Lillian Pierson among its recommendations<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Beneficial<\/td>\n<td>nuoptima.com<\/td>\n<td>Mentions Data-Mania by name in its profile of Lillian Pierson<\/td>\n<td>Yes, ranked first<\/td>\n<\/tr>\n<tr>\n<td>Malicious<\/td>\n<td>geisheker.com<\/td>\n<td>Yes. Referenced 10 on its Best Fractional CMO Companies page is a data-mania.com article<\/td>\n<td>No<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n<h4>The good<\/h4>\n<p><a href=\"https:\/\/nuoptima.com\/best-fractional-cmo-for-ai\" target=\"_blank\" rel=\"noopener\">NUOPTIMA&#8217;s list of the best fractional CMOs for AI companies<\/a>, which ranks me first. It proves that you can earn the recommendation along with the mention, and in my experience so far it&#8217;s the exception.<\/p>\n<h4>The bad<\/h4>\n<p>On 23 September, Geisheker&#8217;s rankings page cited my article as reference 10 and profiled 26 firms, none of them mine &#8211; a clear instance of malicious authority-borrowing (although I simply assumed it was done by AI and that the author was not aware)<\/p>\n<h4>The ugly<\/h4>\n<p>Next, I followed CitedBy\u2019s recommendation and emailed Geisheker about it to see if he would consider mentioning me along with the citation\u2026 <strong>I did not get any reply<\/strong> from him directly, but by 7 October the page had been rewritten to 19 firms, <strong>I was still absent, and the citation to me was gone.<\/strong><\/p>\n<h4>Conclusion<\/h4>\n<p>So, great, the issue is solved, but fact of the matter remains that any AI engine that was drawing on that page gets my article as a source and a list of other firms as the recommended answer. That\u2019s not cool for me, to say the least &#8211; and it happens all the time.<\/p>\n<p><strong>Key Takeaway:<\/strong> You can do the work and earn the citation, and your competitor can still end up with the recommendation.<\/p>\n<h2>3 steps to spotting authority-borrowing with CitedBy<\/h2>\n<p>To identify authority-borrowing you need two views, read in order. Here\u2019s how to get it done.<\/p>\n<h3><span style=\"font-weight: 400;\">Step 1: Look at the Domains view to see specific websites that are competing<\/span><\/h3>\n<p><span style=\"font-weight: 400;\">Open <\/span><b>AI Visibility &gt; Domains<\/b><span style=\"font-weight: 400;\"> and sort by <\/span><b>Retrievals<\/b><span style=\"font-weight: 400;\">. The view defaults to the last 14 days.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Read <\/span><b>Source Class<\/b><span style=\"font-weight: 400;\"> against <\/span><b>Data-Mania Mentioned<\/b><span style=\"font-weight: 400;\">. Yes in that column means the site is directly competing with me, because my brand was mentioned in the same answers where that source was cited. No means the source was cited into answers on my topics and I did not come up at all.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Sorted that way, my own domain sits third. linkedin.com leads at 54 retrievals, <\/span><b>geisheker.com follows at 29<\/b><span style=\"font-weight: 400;\">, and data-mania.com comes next at 20. A site that keeps appearing beside me is getting pulled into my topics more often than I am.<\/span><\/p>\n<figure id=\"attachment_22742\" aria-describedby=\"caption-attachment-22742\" style=\"width: 800px\" class=\"wp-caption aligncenter\"><img fetchpriority=\"high\" decoding=\"async\" class=\"size-large wp-image-22742\" src=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain-1024x445.jpg\" alt=\"\" width=\"800\" height=\"348\" srcset=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain-1024x445.jpg 1024w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain-300x130.jpg 300w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain-768x334.jpg 768w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain-90x39.jpg 90w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain-1536x668.jpg 1536w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Red-Annotation-Highlights-Competing-Domain.jpg 1902w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/><figcaption id=\"caption-attachment-22742\" class=\"wp-caption-text\">AI Visibility &gt; Domains<\/figcaption><\/figure>\n<h3><span style=\"font-weight: 400;\">Step 2: Look at the URLs view to see specific pages that are competing<\/span><\/h3>\n<p><span style=\"font-weight: 400;\">Open <\/span><b>AI Visibility &gt; URLs<\/b><span style=\"font-weight: 400;\"> and set the brand filter to <\/span><b>Min 1 Brand<\/b><span style=\"font-weight: 400;\">. That narrows the table to pages where a tracked brand came up in the answer, which is the list you want.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Sort by <\/span><b>Retrievals<\/b><span style=\"font-weight: 400;\"> and look at which domains repeat. Mine returns geisheker.com three times and gofractional.com twice inside the top six, every row marked Earned with a citation rate of 1.00. Geisheker&#8217;s strongest page is <\/span><b>How to Hire a Fractional CMO in 2026, at 6 retrievals<\/b><span style=\"font-weight: 400;\">.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Citation rate reads 1.00 across these rows, so it does not separate them. Retrievals is the column doing the work.<\/span><\/p>\n<figure id=\"attachment_22744\" aria-describedby=\"caption-attachment-22744\" style=\"width: 800px\" class=\"wp-caption aligncenter\"><img decoding=\"async\" class=\"size-large wp-image-22744 lazyload\" data-src=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted-1024x445.jpg\" alt=\"\" width=\"800\" height=\"348\" data-srcset=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted-1024x445.jpg 1024w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted-300x130.jpg 300w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted-768x334.jpg 768w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted-90x39.jpg 90w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted-1536x668.jpg 1536w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/Citedby-Dashboard_-Direct-Competition-Highlighted.jpg 1902w\" data-sizes=\"(max-width: 800px) 100vw, 800px\" src=\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\" style=\"--smush-placeholder-width: 800px; --smush-placeholder-aspect-ratio: 800\/348;\" \/><figcaption id=\"caption-attachment-22744\" class=\"wp-caption-text\">AI Visibility &gt; URLs with Min 1 Brand applied<\/figcaption><\/figure>\n<h3><span style=\"font-weight: 400;\">Step 3: Run one prompt to identify whether that page cites you and recommends someone else<\/span><\/h3>\n<p><span style=\"font-weight: 400;\">The first two steps tell you a page competes with you. They do not tell you whether it is built on your work, but you can settle that question with a single prompt.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Give this to any AI assistant with web access:<\/span><\/p>\n<hr \/>\n<p><span style=\"color: #339966;\"><span style=\"font-weight: 400;\">Read [PAGE URL].<br \/>\n<\/span>1. Does this page mention, cite, or link to [YOUR BRAND] or [YOUR NAME] anywhere, including body text, references, sources, or footnotes? Quote the exact sentence or reference line verbatim, with its reference number if the page numbers its sources.<\/span><br \/>\n<span style=\"color: #339966;\">2. Separately, does this page recommend, rank, or list any [YOUR CATEGORY] as options to hire? Name every one of them.<\/span><br \/>\n<span style=\"color: #339966;\">3. Is [YOUR BRAND] or [YOUR NAME] among those recommended?<\/span><\/p>\n<hr \/>\n<p>&nbsp;<\/p>\n<p><span style=\"font-weight: 400;\">Keep it as three questions. A page can answer yes to the first and no to the third, and that gap is what you\u2019re looking for. <\/span><\/p>\n<p>If the page is a comparison page and you&#8217;re being both cited and recommended, you\u2019re seeing beneficial authority-borrowing in action. If, however, you&#8217;re being cited and omitted from recommendations, you\u2019ve got an instance of malicious authority-borrowing going on.<\/p>\n<h3>Actions I am taking to counteract malicious authority-borrowing<\/h3>\n<p>So what am I doing about malicious authority-borrowing? As a fractional CMO and GTM engineer, I focus on system-level facts rather than prompt hacks.<\/p>\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" data-src=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/thrivestack-citedby-review-05.jpg\" alt=\"Infographic titled 3 ways to counter malicious authority-borrowing: don't rely on direct AI instructions, build first-party owned content with precise credentials, named frameworks and defined service categories, and earn offsite recommendations.\" src=\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\" class=\"lazyload\" style=\"--smush-placeholder-width: 1402px; --smush-placeholder-aspect-ratio: 1402\/1122;\" \/><figcaption class=\"wp-element-caption\">Three methods to counteract malicious authority-borrowing, focusing on first-party and offsite authority over direct AI instructions.<\/figcaption><\/figure>\n<p>I can\u2019t change someone else\u2019s website, and if they don\u2019t want to collaborate, I am left with only a few methods:<\/p>\n<ul>\n<li><strong>Direct AI instructions or command directives on web pages<\/strong> &#8211; Don\u2019t do this. It will eventually fail because modern AI models treat explicit instructions as potential prompt injection and aggressively discount them.<\/li>\n<li><strong>First-party owned content publishing<\/strong> &#8211; A real lever is publishing canonical, quotable, factual self-descriptions. By providing structured, verifiable facts (such as precise credentials, named frameworks, and defined service categories) that AI engines can naturally retrieve, cite, and recommend your brand during answer synthesis.<\/li>\n<li><strong>Offsite earned content published<\/strong> &#8211; Working on off-site recommendations is sure to help supersede cases of malicious authority-borrowing. CitedBy has tons of great recommendations for this as well.<\/li>\n<li><strong>ThriveStack CitedBy&#8217;s real strength is choosing what you work on.<\/strong> Two of its recommendations produced gains that Google Search Console recorded within 6 days.<\/li>\n<li><strong>AI visibility scores move a lot on their own.<\/strong> I trust a result only when Google counted it or when I could confirm it by reading the page.<\/li>\n<li><strong>Reading your own top pages turns up wins that automated tools may not flag.<\/strong> One of my three counted gains came from updating a page that wasn&#8217;t highlighted in automated recommendations.<\/li>\n<\/ul>\n<p>Beyond identifying and counteracting authority-borrowing, measuring the precise impact of these updates requires a clear framework for analyzing performance metrics.<\/p>\n<h2>ThriveStack CitedBy&#8217;s picks moved two of my posts up in Google within 6 days<\/h2>\n<p>As mentioned previously, before I credited ThriveStack CitedBy with anything, I sorted every result into one of three groups, based on how much you can trust it:<\/p>\n<ol>\n<li><strong>Counted results<\/strong> come from Google Search Console, which records every time Google shows your page in search. You can pull the same number twice and get the same answer twice, so these are the numbers I trust most.<\/li>\n<li><strong>Facts<\/strong> are things I confirmed by reading a page, so they need no statistics at all.<\/li>\n<li><strong>Sampled results<\/strong> come from ThriveStack CitedBy asking AI engines a set of prompts and scoring the answers. Those answers change every time you ask, so I treat a single run as a lead.<\/li>\n<\/ol>\n<p><strong>What survived the variance check<\/strong><\/p>\n<p>The table below shows how I split the two kinds of AI search metrics, and how to use them:<strong>.<\/strong><\/p>\n<figure class=\"wp-block-table\">\n<table>\n<thead>\n<tr>\n<th>Metric type<\/th>\n<th>Example<\/th>\n<th>How the metric behaves across runs<\/th>\n<th>What to use the metric for<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Sampled<\/td>\n<td>An AI visibility score<\/td>\n<td>Standard deviation of 10.3 points across 40 runs of the same 25 prompts on data-mania.com, with a trend r-squared of 0.00, meaning the runs showed no trend at all<\/td>\n<td>Direction and target selection<\/td>\n<\/tr>\n<tr>\n<td>Counted<\/td>\n<td>AI Overview saturation (the share of your ranking keywords whose results page carries an AI Overview) and search impressions<\/td>\n<td>Pull the metric twice and you get the same answer twice<\/td>\n<td>Reporting a result to a board or a client<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n<p>Source: 40 consecutive runs of the same 25 prompts against data-mania.com, April 3 to September 6, 2026. The full analysis is in <a href=\"https:\/\/www.data-mania.com\/blog\/ai-visibility-score-noise-proxy-stack\/\">AI citation measurement is broken<\/a>.<\/p>\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" data-src=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/thrivestack-citedby-review-06.jpg\" alt=\"Infographic titled Know which AI search numbers you can report, comparing counted metrics such as search impressions with sampled metrics such as AI visibility scores, and showing a 10.3-point standard deviation across 40 runs and two tools scoring the same domain 22 and 4.\" src=\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\" class=\"lazyload\" style=\"--smush-placeholder-width: 1122px; --smush-placeholder-aspect-ratio: 1122\/1402;\" \/><figcaption class=\"wp-element-caption\">Understanding the difference between counted and sampled AI search metrics for accurate client and board reporting.<\/figcaption><\/figure>\n<figure class=\"wp-block-table\">\n<table>\n<thead>\n<tr>\n<th>Finding<\/th>\n<th>Evidence class<\/th>\n<th>Source<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Search impressions on the data-mania.com post about how an AI-native growth system works rose from 13 to 59 in 6 days<\/td>\n<td>Counted. A census of what Google served.<\/td>\n<td>Google Search Console<\/td>\n<\/tr>\n<tr>\n<td>Average search position on the data-mania.com post about product-led versus sales-led growth moved from 46.8 to 18.4 in 6 days<\/td>\n<td>Counted, and averaged across every impression<\/td>\n<td>Google Search Console<\/td>\n<\/tr>\n<tr>\n<td>Three new section-level search results, meaning results that point to one section of a page, began indexing at position 4 for a data-mania.com article<\/td>\n<td>Counted<\/td>\n<td>Google Search Console<\/td>\n<\/tr>\n<tr>\n<td>The summary tables in the most-cited data-mania.com article featured data-mania.com in only one of four revenue bands<\/td>\n<td>Fact, verified by reading, with no measurement involved<\/td>\n<td>A close reading of the published page<\/td>\n<\/tr>\n<tr>\n<td>One Gemini answer drew 7 of its 10 citations from data-mania.com<\/td>\n<td>Sampled. A candidate that still needs repeat runs.<\/td>\n<td>ThriveStack CitedBy, single run<\/td>\n<\/tr>\n<tr>\n<td>Citation-flip gains and topic-percentage gains on data-mania.com<\/td>\n<td>Sampled, and excluded because the gains fell below the significance threshold<\/td>\n<td>ThriveStack CitedBy<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n<h3>2 fixes that ThriveStack CitedBy picked produced gains that Google recorded<\/h3>\n<p>Impressions count how many times Google showed your page in search results, and average position tells you how high the page appeared, where a lower number is better.<\/p>\n<p><strong>For the first recommendation, ThriveStack CitedBy showed me that my site held zero of the 10 citations for a prompt about AI-native growth systems<\/strong>, and that cxl.com held the top citation spot. I already had a post on the topic, so I added one section to my article on <a href=\"https:\/\/www.data-mania.com\/blog\/ai-native-growth-system-how-it-works-b2b\/\">how an AI-native growth system actually works<\/a>. Within 6 days, that post&#8217;s impressions more than quadrupled, from 13 to 59.<\/p>\n<p><strong>The second recommendation led me to <\/strong><a href=\"https:\/\/www.data-mania.com\/blog\/product-led-growth-vs-sales-led-growth-which-fits-your-saas\/\"><strong>the post I rebuilt on product-led versus sales-led growth<\/strong><\/a><strong>, which turned into a much bigger job than closing a content gap<\/strong>. I refreshed the sources, rebuilt the body, updated the author credit, and expanded the FAQ section. Within 6 days, its average position moved from 46.8 to 18.4.<\/p>\n<p>The third fix came from no recommendation at all, and you&#8217;ll find it in the next section.<\/p>\n<p><strong>Two ThriveStack CitedBy recommendations and one fix I found myself, the change I made, and the counted result are described in the table below.<\/strong><\/p>\n<figure class=\"wp-block-table\">\n<table>\n<thead>\n<tr>\n<th>The recommendation<\/th>\n<th>What I changed<\/th>\n<th>Search metric <strong>before<\/strong><\/th>\n<th>Search metric <strong>after<\/strong><\/th>\n<th>Window and source<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>ThriveStack CitedBy prompt &#8220;How does an AI-native growth system actually work for B2B companies?&#8221; ThriveStack CitedBy gap score of 9 on its 10-point scale, zero of 10 citations to data-mania.com, while cxl.com held rank 1<\/td>\n<td>Added one section to a data-mania.com post that was already published<\/td>\n<td>13 search impressions<\/td>\n<td><strong>59<\/strong> search impressions<\/td>\n<td>6 days, Google Search Console<\/td>\n<\/tr>\n<tr>\n<td>ThriveStack CitedBy prompt &#8220;How does an AI-native growth system actually work for B2B companies?&#8221; measured by average search position<\/td>\n<td>Added one section to a data-mania.com post that was already published<\/td>\n<td>Average position 8.8<\/td>\n<td>Average position <strong>7.2<\/strong><\/td>\n<td>6 days, Google Search Console<\/td>\n<\/tr>\n<tr>\n<td>ThriveStack CitedBy prompt &#8220;Product-led growth versus sales-led growth for B2B SaaS.&#8221; ThriveStack CitedBy gap score of 8, zero of 5 citations to data-mania.com<\/td>\n<td>Refreshed the sources, rebuilt the body, updated the author credit, and expanded the FAQ from 2 questions to 6<\/td>\n<td>Average position 46.8<\/td>\n<td>Average position 18.4<\/td>\n<td>6 days, Google Search Console<\/td>\n<\/tr>\n<tr>\n<td>No ThriveStack CitedBy recommendation. I found this fix myself by reading the data-mania.com growth advisors roundup<\/td>\n<td>Rewrote both summary tables so that data-mania.com appears in three of four revenue bands instead of one<\/td>\n<td>No section-level search results<\/td>\n<td>Three section-level search results at position 4<\/td>\n<td>Google Search Console<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n<p><strong>Key Takeaway:<\/strong> All three of my counted gains came from improving pages that I had already published. Before you write something new for a recommendation, check whether one of your existing pages can do the job.<\/p>\n<p>To make things for more concrete, let me show some examples of recommendations I got from CitedBy and how I would distribute those responsibilities across marketing roles:<\/p>\n<ul>\n<li><strong>Technical Recommendations (for Technical SEO):<\/strong> Resolve the duplicate @id collisions causing schema validation failures across all three audited page templates; backfill descriptive alt text site-wide to improve AI crawler comprehension.<\/li>\n<li><strong>Existing Content Recommendations (for SEO &amp; Content Writers):<\/strong> Refresh the 232-citation fractional CMO roundup to a 2026 edition with self-contained comparison tables, retaining the existing slug; add a standalone 40-to-60-word direct answer and a visible credentialed byline to the Fractional CMO Services page.<\/li>\n<li><strong>New Content Recommendations (for Content Writers &amp; Agencies):<\/strong> Create two buyer guide articles against uncaptured competitor prompts identified in the content gap analysis: the AI-native growth system versus manual outbound decision (zero of 46 citations) and the timeline to measurable revenue after marketing leadership (zero of 7 citations).<\/li>\n<li><strong>Entity Signal Recommendations (for Brand Marketing):<\/strong> Consolidate conflicting Organization and Person JSON-LD into a single canonical block emitted identically site-wide, with a standardized sameAs union that includes an official company LinkedIn page, currently absent, so LLMs correctly map the brand across profiles.<\/li>\n<li><strong>Off-Site Recommendations (for PR &amp; Link Building Teams):<\/strong> Claim the self-serve profile listing on rankedcmo.com, which names 52 fractional CMOs and carries an AI SaaS section; pitch geisheker.com for an AI and data category on its Best Fractional CMO Companies rankings, where Data-Mania already appears as a cited source but not as a listed firm.<\/li>\n<\/ul>\n<h3>Reading my own most-cited page turned up a third counted win<\/h3>\n<p>The third counted win started with a fact that I confirmed by reading an existing page\u2026 in other words, no amount of run-to-run movement can expose this.<\/p>\n<p><a href=\"https:\/\/www.data-mania.com\/blog\/top-7-growth-advisors-for-tech-startups-in-2025-who-to-hire-based-on-your-growth-stage\/\">My growth advisors roundup<\/a> is my most-cited article, and it sorts its recommendations into revenue bands. When I read it closely, I saw that its summary tables featured my own services in only one of the four bands. I updated both tables so that I appear in three bands instead of one, and Search Console then started to show three new results that point to specific sections of that page, all at position 4.<\/p>\n<p><strong>Key Takeaway:<\/strong> An AI visibility tool can guide you on which page to prioritize, but a direct review of your top pages will reveal what items on them are worth updating. For me, that manual review produced a counted win that automated tools wouldn&#8217;t typically flag.<\/p>\n<h3>AI engines lift my self-contained tables into their answers whole<\/h3>\n<p>The most durable gain from this month sits outside the scores&#8230; When I looked through my own citation records, I saw that AI engines regularly lift a table out of my page and quote it as a standalone answer. One of those recorded citations begins with a table header.<\/p>\n<p>If an engine is going to quote a fragment of your page, you want that fragment to make complete sense on its own, with every column header and every cell naming what it measures. I built the tables in this article to that standard, and recommend that you rebuild the tables on your top pages the same way.<\/p>\n<h2>ThriveStack CitedBy picks better targets than the rest of my stack<\/h2>\n<p><a href=\"https:\/\/linkly.link\/2vOm5\" target=\"_blank\" rel=\"noopener\">ThriveStack CitedBy<\/a> tells you which buyer questions you&#8217;re missing from in AI answers, and which competitor is showing up in your place. I run more than one AI visibility tool. Below you can see the five places where ThriveStack CitedBy outperforms them for me, with one example of each:<\/p>\n<ol>\n<li><strong>ThriveStack CitedBy ranks every gap by its evidence.<\/strong> Each recommendation carries a gap score out of 10, where a higher score means a bigger gap, and a citation count, so you know where you stand before you write a word. My AI-native growth system prompt scored 9 out of 10, with zero of the 10 citations going to my site.<\/li>\n<li><strong>ThriveStack CitedBy names the domain that you&#8217;re losing to.<\/strong> For the AI-native growth system prompt, ThriveStack CitedBy showed me cxl.com in the top citation spot, so I knew exactly which domain my new section had to outrank.<\/li>\n<li><strong>ThriveStack CitedBy reads several engines.<\/strong> A score from one engine tells you what one model produced on one day. ThriveStack CitedBy reports across engines, and cross-engine reporting is the design choice that my own audit of <a href=\"https:\/\/www.data-mania.com\/blog\/ai-visibility-score-noise-proxy-stack\/\">AI citation measurement<\/a> pointed me toward.<\/li>\n<li><strong>ThriveStack CitedBy keeps on-page work and off-site work in separate lists.<\/strong> You get one list of edits for your own pages and a second list of third-party sites to pursue. The three off-site targets that I pitched this month came out of its off-site recommendations.<\/li>\n<li><strong>ThriveStack CitedBy routes every recommendation to a named role<\/strong>, so a 127-item list becomes five team queues. The rest of my stack hands me findings while CitedBy one hands me prioritized assignments I can validate and delegate quickly. If you have ever tried to split an audit between a developer, a writer and a PR contact, you already know that translation is most of the work.<\/li>\n<\/ol>\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" class=\"aligncenter size-large wp-image-22745 lazyload\" data-src=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View-1024x445.jpg\" alt=\"\" width=\"800\" height=\"348\" data-srcset=\"https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View-1024x445.jpg 1024w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View-300x130.jpg 300w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View-768x334.jpg 768w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View-90x39.jpg 90w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View-1536x667.jpg 1536w, https:\/\/www.data-mania.com\/blog\/wp-content\/uploads\/2026\/10\/CitedBy-Agentic-Actions-View.jpg 1910w\" data-sizes=\"(max-width: 800px) 100vw, 800px\" src=\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\" style=\"--smush-placeholder-width: 800px; --smush-placeholder-aspect-ratio: 800\/348;\" \/><figcaption class=\"wp-element-caption\">CitedBy&#8217;s All Actions view, assigning each of 127 recommendations to a role and an effort level across six stages.<\/figcaption><\/figure>\n<p>All five of those strengths come down to choosing targets. ThriveStack CitedBy&#8217;s scores are sampled and serve as a very good fit for deciding where you spend your next week of work.<\/p>\n<p>Beyond target selection and task delegation, ThriveStack has expanded CitedBy over the past four weeks with several advanced features and integrations that enhance its workflow:<\/p>\n<ul>\n<li><strong>Prompt Research: <\/strong>Fine-tune the prompts and answers you want to win using Prompt Research, which includes prompt imports from providers like Otterly ai and Peec.ai.<\/li>\n<li><strong>SEO and AEO Content Audits: <\/strong>Perform integrated SEO and AEO content audits directly within the product to evaluate key client pages (includes 5 free audited pages, with additional capacity available via paid usage-based pricing).<\/li>\n<li><strong>Analytics &amp; Search Console Integrations: <\/strong>Native support for Google Search Console and GA4 integrations to verify organic and AI search metrics alongside site traffic.<\/li>\n<li><strong>Model Context Protocol (MCP): <\/strong>Enhanced MCP integration with stability fixes allowing AI assistants to directly interact with your CitedBy visibility data.<\/li>\n<\/ul>\n<h2>6 analytical caveats to keep your AI visibility results honest<\/h2>\n<p>When you analyze AI visibility data, keeping key general analytical considerations in mind ensures that your interpretations stay objective. The table below outlines common analytical pitfalls, explains their potential impacts, and provides simple habits that will help you navigate them effectively.<\/p>\n<figure class=\"wp-block-table\">\n<table>\n<thead>\n<tr>\n<th>Caution<\/th>\n<th>Risk<\/th>\n<th>What to do instead<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Recommendations may reflect historical snapshot data rather than real-time re-scores.<\/td>\n<td>You may spend time working to close a gap that you have already resolved.<\/td>\n<td>You should verify findings against the latest prompt runs before starting work.<\/td>\n<\/tr>\n<tr>\n<td>Automated gap analyses focus on missing content rather than existing published inventory.<\/td>\n<td>You might write an entirely new post when editing an existing post would achieve the goal.<\/td>\n<td>You should search your own site first whenever you evaluate a recommendation.<\/td>\n<\/tr>\n<tr>\n<td>AI visibility metrics measure citations rather than content accuracy or endorsement nuances.<\/td>\n<td>Problems on your best-performing pages can remain live without anyone noticing them.<\/td>\n<td>You should personally fact-check and read through your top pages.<\/td>\n<\/tr>\n<tr>\n<td>On-page and off-site recommendation metrics often utilize different scoring scales.<\/td>\n<td>You may misjudge which list of recommendations you should prioritize working on.<\/td>\n<td>You should rank recommendations inside each list independently and never compare across the two lists.<\/td>\n<\/tr>\n<tr>\n<td>Generative AI scoring metrics inherently rely on sampled prompt outputs.<\/td>\n<td>You risk reporting random noise as if it were a meaningful result.<\/td>\n<td>You should publish your prompt count along with your date window, and lead with counted metrics.<\/td>\n<\/tr>\n<tr>\n<td>AI visibility tools track citations and brand presence rather than direct revenue.<\/td>\n<td>A gain in visibility may be mistaken as a substitute for an actual pipeline metric.<\/td>\n<td>You should keep Google Search Console and your analytics tools open beside ThriveStack CitedBy.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n<p>These analytical considerations directly connect to my own results. Every counted win in this review came from an update to a page that I had already published, which illustrates why auditing existing inventory is essential alongside automated recommendations. One of those wins came from a page that automated analysis didn&#8217;t flag. That underscores the importance of combining tool insights with human review.<\/p>\n<h2>ThriveStack CitedBy earns its place as the tool that picks your targets<\/h2>\n<p>ThriveStack CitedBy is the best tool in my stack for this job. It picked the targets behind two of my three counted wins, and it named the competing domains on every gap it gave me.<\/p>\n<p><strong>Actionable Takeaways:<\/strong><\/p>\n<ul>\n<li><strong>Audit before creating:<\/strong> Check existing published pages to close content gaps before drafting new articles.<\/li>\n<li><strong>Optimize for extraction:<\/strong> Use self-contained comparison tables and canonical entity descriptions that models can lift directly.<\/li>\n<li><strong>Lead with counted metrics:<\/strong> Prioritize verified Google Search Console data over sampled AI visibility scores when measuring success.<\/li>\n<\/ul>\n<p>You can start with <a href=\"https:\/\/linkly.link\/2vOlv\" target=\"_blank\" rel=\"noopener\">ThriveStack CitedBy<\/a> on your own domain. If you&#8217;d like help deciding which of its recommendations to act on first, book <a href=\"https:\/\/linkly.link\/2vOlx\" target=\"_blank\" rel=\"noopener\">a 30-minute conversation<\/a> with me. Let&#8217;s map it out together.<\/p>\n<h2>Quick answers to six common AI visibility questions<\/h2>\n<h3>Is an AI visibility score reliable?<\/h3>\n<p>On its own, no. When I ran the same 25 prompts 40 times on my site, the score typically landed about 10 points above or below its average (a standard deviation of 10.3) and showed no trend at all (an r-squared of 0.00). A move needs to clear roughly 20 points before you can tell it apart from random variation. Use the score for direction, and use counted metrics for reporting.<\/p>\n<h3>How many prompts should sit behind an AI visibility score?<\/h3>\n<p>Ask the vendor, and treat the answer as your first gate. My score is computed on 25 prompts, which means a single prompt that flips moves my overall score by 2 points. A topic that rests on five prompts moves 6.7 points when a single engine flips. The prompt count tells you how much noise you are looking at.<\/p>\n<h3>Why do two AI visibility tools report different scores for the same website?<\/h3>\n<p>Two tools disagree because each score describes the tool&#8217;s prompt set as much as it describes your brand. On September 6, one of my tools reported 22 and the other reported 4, for the same domain in the same week. Both were accurate on their own terms. Pick one tool, stay on it, and prioritize a tool that reads several engines.<\/p>\n<h3>Which AI search metrics can I actually report to a client or a board?<\/h3>\n<p>Report the counted ones. Search impressions record every time Google showed your page, and AI Overview saturation measures the share of your ranking keywords whose results page carries an AI Overview. Both are counts, so you can pull them twice and get the same answer twice. Citation and visibility scores are samples of a random process. Lead your reporting with the counted numbers, and place the sampled numbers underneath them.<\/p>\n<h3>Can I make AI engines recommend my brand when they cite my website?<\/h3>\n<p>You cannot force a recommendation via direct web page instructions, as LLMs ignore directive commands to prevent prompt injection. Instead, publish clear, canonical self-descriptions and structured tables. When information is structured as definitive, self-contained entity data, models naturally quote and recommend your brand.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>My ThriveStack CitedBy review: how two of its picks grew Google impressions 4x in 6 days, and the authority-borrowing problem no AI visibility tool has solved.<\/p>\n","protected":false},"author":1,"featured_media":22731,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_wp_convertkit_post_meta":{"form":"-1","landing_page":"0","tag":"0","restrict_content":"0"},"footnotes":"","_links_to":"","_links_to_target":""},"categories":[846,851],"tags":[],"class_list":["post-22738","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-search-visibility","category-sponsored-content"],"_links":{"self":[{"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/posts\/22738","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/comments?post=22738"}],"version-history":[{"count":8,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/posts\/22738\/revisions"}],"predecessor-version":[{"id":22750,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/posts\/22738\/revisions\/22750"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/media\/22731"}],"wp:attachment":[{"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/media?parent=22738"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/categories?post=22738"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.data-mania.com\/blog\/wp-json\/wp\/v2\/tags?post=22738"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}