--- title: "Introducing OptimizeCamp 2.0: The Closed-Loop AI Visibility Platform" url: https://blog.optimizecamp.com/introducing-optimizecamp/ date: 2026-03-09 modified: 2026-08-09 author: "Imtiaz Rayhan" description: "On January 27, 2026, roughly four in ten domains cited by Google AI Overviews stopped being cited. Nobody published a bad article that day. Nobody removed a schema tag. Google..." categories: - "AI Visibility Tracking" image: https://blog.optimizecamp.com/wp-content/uploads/2026/03/OptimizeCamp-2.0-1024x576.png word_count: 2928 --- # Introducing OptimizeCamp 2.0: The Closed-Loop AI Visibility Platform On January 27, 2026, roughly four in ten domains cited by Google AI Overviews stopped being cited. Nobody published a bad article that day. Nobody removed a schema tag. Google made Gemini 3 the default model behind AI Overviews, and the citation pool reshuffled overnight. SE Ranking measured the damage across 100,000 keywords in 20 niches. That single event explains why OptimizeCamp is not the product we launched in March. Version 1.0 was an AI-citation readiness auditor. You pasted an article, we scored it on accuracy, authority and citability, and you accepted inline fixes. It worked. It also answered the wrong question. It told you whether a page *looked* citable. It never told you whether ChatGPT actually cited it on Tuesday. **An audit is a photograph. AI visibility is a film.** 0% of domains previously cited in Google AI Overviews were replaced after Gemini 3 became the default model on January 27, 2026. Source: SE Ranking, 100,000-keyword study across 20 niches, February 2026 ## What Is OptimizeCamp 2.0? OptimizeCamp 2.0 is a closed-loop AI visibility platform. It runs your buyers' real questions against ChatGPT, Google AI Overviews and Perplexity on a schedule, measures how you and every competitor appear in the answers, fixes the pages costing you citations, generates the content you are missing, then re-measures to prove the work moved the number. Four verbs: measure, fix, create, prove. Version 1.0 covered part of the second one. OptimizeCamp 1.0 Paste an article. Get a composite score out of 100. Accept inline fixes. The score was our opinion of your page. Nothing measured what AI actually said about you, and nothing proved a fix worked. OptimizeCamp 2.0 Track real answers on a schedule. Score is computed from those answers, identically for you and every rival. Fixes, briefs and drafts are built from that evidence. Every publish is re-measured. ## Why One Audit Stopped Being Enough Three things broke the audit-only model in the space of a year. AI answers now absorb the click, citation stopped tracking rank, and the citation pool changes whenever a model does. A score you computed in March cannot know any of that. Start with the click. The [Pew Research Center](https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/) tracked 68,879 real Google searches from 900 US adults. When an AI summary appeared, users clicked a result 8% of the time. Without one, they clicked 15% of the time. Only 1% clicked a link inside the summary itself. Then the rank link snapped. Ahrefs analysed 863,000 keywords and 4 million AI Overview URLs. Only 38% of cited pages also ranked in the top 10. Seven months earlier that figure was 76%. Ranking and getting cited are now close to independent events. 8% click rate when an AI summary appears, versus 15% without one Pew Research Center, 68,879 searches 38% of AI Overview citations also rank top 10, down from 76% Ahrefs, 863,000 keywords 84% of AI citations come from earned, third-party media Muck Rack, 25M+ links, May 2026 Google disputes the traffic read. In its own [public response](https://www.emarketer.com/content/google-denies-that-ai-overviews-harm-referral-traffic--stats-say-otherwise), Google says total organic click volume has held steady year over year, and that click quality has improved. We report the conflict rather than resolve it. Either way, the argument is about clicks. The visibility question sits upstream of it. ## The Loop: Measure, Fix, Create, Prove OptimizeCamp 2.0 is built as a loop, not a dashboard. Measurement produces evidence. Evidence produces work. Work gets published. The next scheduled run measures whether the number moved, and the result feeds the following cycle. Nothing enters the queue without evidence behind it. THE OPTIMIZECAMP LOOP 01 Measure Real answers from three platforms, on a schedule 02 Fix Audit and repair the pages losing citations 03 Create Brief and draft the pages you do not have 04 Prove Re-measure and issue an Impact Card, or admit no gain Every cycle compounds. The loop never stops at step one. The screenshots below come from a live project on our own account, tracking the WordPress popup plugin ConvertForce. Real numbers, including the unflattering ones. ## Measure: Four Signals, One Visibility Score For every prompt on every platform, OptimizeCamp parses the actual answer the way a reader would. It records four signals: whether your brand is **mentioned**, whether one of your pages is **cited** in the source list, what **position** your first mention takes among all brands named, and the **sentiment** of the sentence around it. Those four roll into a Visibility Score from 0 to 100, plus a competitive Share of Voice. The rules are identical for you and every tracked competitor. The underlying answer is always one click away, with your brand and each rival highlighted inline. On the ConvertForce project, that reads: 726 answers analysed across three platforms, Visibility Score 45.2, mention rate 46.7%, citation rate 33.3%, sentiment 82.1. The per-platform split matters more than the headline. ConvertForce scores 61.3 on ChatGPT, 46 on Perplexity, and 45.2 on Google AI Overviews. One brand, three very different realities. VISIBILITY SCORE, ONE PROJECT, 15 DAYS ConvertForce (now 45.2) OptinMonster, top competitor Source: OptimizeCamp, ConvertForce project, 19 July to 2 August 2026. Scores are 7-day rolling averages, indicative shape read from the dashboard. Scores are trended on 7-day rolling averages, so daily answer variance never reads as real movement. Edit a prompt or an alias and the chart gets a break marker. A method change should never be mistaken for a result. ## Prompts: The Questions You Are Actually Judged On Your visibility is only ever measured against the questions you choose to track. OptimizeCamp proposes them from your brand profile, grouped into branded, comparison, unbranded and informational buckets, and tells you why each one was suggested. You check, uncheck, rewrite inline, or add your own. The prompt table is where the pattern usually shows up. In the ConvertForce project, every branded prompt returns "mentioned and cited" on all three platforms. Almost every unbranded prompt returns "absent" on Google AI Overviews and Perplexity. **That is the whole problem in one screen: people who already know the brand find it, and people shopping the category never see it.** We wrote a longer piece on this exact decision in [Query Fanout Explained](https://blog.optimizecamp.com/query-fanout-explained/). ## Citations And Sources: Two Halves Of One Answer Every AI answer has a source list, and most of it is not your website. OptimizeCamp splits that into two views. **Citations** covers URLs on your own domain: which pages get cited, on which prompts, how often, and whether they are still in the latest answers or have quietly dropped out. **Sources** covers everyone else's. The split is not cosmetic. According to Muck Rack's May 2026 edition of [What Is AI Reading?](https://muckrack.com/blog/what-is-ai-reading-may-2026), analysing more than 25 million links across ChatGPT, Claude and Gemini, earned third-party media accounts for 84% of all AI citations. Paid and advertorial content accounts for 0.3%. WHERE AI CITATIONS ACTUALLY COME FROM 25M links analysed 84%Earned media, journalism 27% of the total 15.7%Everything else, including your own site 0.3%Paid and advertorial content Source: Muck Rack, What Is AI Reading?, May 2026 edition, 25M+ links across ChatGPT, Claude and Gemini The ConvertForce project has 5,181 earned citations spread across 865 third-party domains. It appears on 270 of the sources cited in the latest run, and is absent from 130 of them. Those 130 are not a report. Each one is a queued action, which is what Plays exist for. ## Competitors: Same Prompts, Same Runs, Same Rules Competitive numbers are worthless unless everyone is measured on the same data. OptimizeCamp benchmarks every brand from the same platform-complete snapshots, and tells you exactly which snapshots those were. The default scope is neutral discovery: active unbranded prompts that name neither you nor a tracked competitor. That scope is deliberately unkind. Branded prompts inflate everybody. Strip them out and the ConvertForce project drops to third, behind Popup Maker and OptinMonster. VISIBILITY SCORE ON NEUTRAL DISCOVERY PROMPTS Scale: 0 to 100. Higher is better. Popup Maker51.0 OptinMonster48.2 ConvertForce (tracked brand)39.3 Elementor21.8 Icegram15.8 Source: OptimizeCamp, ConvertForce project, 8 neutral discovery prompts across 24 platform-complete cells, 31 July to 2 August 2026 Below the table sits a head-to-head grid. Pick one rival and see every prompt, every platform, and who takes the better position. Against Popup Maker on this project: you lead 3, they lead 5, and the movement line reads 2 of 8 yesterday to 3 of 8 now. There is also a list of untracked brands appearing in your answers, with a one-click button to start tracking them. Elementor Pro showed up in 98 of 709 answers before anyone thought to add it. ## Alerts And The Action Center: One Queue, Three Lanes Alerts fire after every tracking run and answer one question: what changed. Lost a citation you had in three of the last three runs. Vanished from a prompt you were in yesterday. Won back a platform after six runs absent. Each alert names the prompt, the platform, and the run history behind the call. The Action Center then turns all of that into a single queue, split into three lanes so unlike things never get ranked against each other. Lane 1 · Ready To Finish Work already in flight. A generated draft waiting for review. A brief approved but not written. Workflow state, no invented priority score. Lane 2 · Evidence-Backed Opportunities Page priority, content gaps and earned-source gaps. Each carries its own measure and the evidence that produced it, such as "competitors cited on 6 matched prompts". Lane 3 · Measurement And Results Verdicts on work you already shipped. Including the honest ones: "no verified gain after 4 verification runs" and "no impact claim, evidence inconclusive". ## Optimize: The 1.0 Auditor, Now Aimed At Real Answers The original audit engine did not get thrown away. It got a target. Optimize audits pages across four engines: accuracy, citability, authority and AI coverage. The difference from 1.0 is what decides which page you audit next. In 1.0, you picked. In 2.0, a priority queue ranks pages by the citations they are losing in tracked answers. A page might surface because competitors are cited on 6 of its matched prompts while it has never been cited on its own 8. Accept fixes, publish back to WordPress over the REST API, and the page enters a verification state. If your pages are still built for rankings rather than answers, [How To Structure Competitor Comparison Pages For AI Search](https://blog.optimizecamp.com/comparison-pages-ai-search/) covers the format that gets extracted. ## Generate: Briefs Built From Gaps, Not Guesses Generate answers one question: what should we write next, and how do we know. It groups the prompts you are losing and the competitor pages AI keeps citing into named content gaps, each scored by opportunity and confirmed across independent batches. A gap on the ConvertForce project reads like this. Topic "alternative convert", opportunity 84, three tracked prompts lost on Google AI Overviews, absent on every answering platform, competitors cited 12 times, confirmed in 2 of 2 batches. It lists the exact pages cited instead, then recommends updating an existing post at 67% topic match rather than writing a new one. **A brief you can argue with is worth more than a brief you have to trust.** Drafts are fact-checked, every claim linked to a source, and they only ever reach WordPress as drafts. A human always publishes. ## Plays: The Citations You Cannot Publish Yourself Plays is the module that follows the 84% number to its conclusion. If most citations live on domains you do not own, then most of the work is off your site. Plays turns each of those domains into a specific task. Every third-party source cited on your prompts where you are absent becomes a play, with an opportunity score, an effort estimate, a status, and a record of who is being cited instead of you. The ConvertForce project has 238 open plays, mostly roundup pitches to sites already shaping its category's answers. This is the AI-era version of what PR teams have always done, ranked by measured impact instead of instinct. ## Prove It Worked: Impact Cards And Receipts Every published fix gets verified or it gets nothing. Once the next two tracking runs land, within fourteen days, an Impact Card reports what changed on the prompts that page targets: citation gained, mention rate moved, Visibility Score delta. If the evidence is thin, the card says so and makes no claim. Here is the last 30 days on the ConvertForce project, unedited. Two prompts won back. Eight new citations confirmed. Five content gaps closed. Share of voice up 6.3 percentage points. Visibility Score down 0.5 points. That last number is the point. A tool that only ever reports wins is a marketing asset, not an instrument. One card on this project currently reads "no verified citation or presence gain across 4 verification runs". We shipped that state deliberately. ## What Changed From 1.0 To 2.0 Everything in 1.0 still exists. It stopped being the whole product and became one module inside a loop. Here is the direct mapping. | Dimension | OptimizeCamp 1.0 (March 2026) | OptimizeCamp 2.0 (August 2026) | | --------- | ----------------------------- | ------------------------------ | | Core question | Does this page look citable? | Are you in the answer, and who is instead? | | Input | An article you paste in | A URL, then your real buyer questions | | Measurement | Composite score from our heuristics | Mention, citation, position and sentiment in real answers | | Platforms | None tracked | ChatGPT, Google AI Overviews, Perplexity | | Competitors | SERP pages for one keyword | Up to 10 brands, scored on identical snapshots | | Off-site work | Not covered | Sources and Plays, 865 domains on one project | | Content creation | Gap-filler snippets | Evidence-backed gaps, briefs and fact-checked drafts | | Proof | Score went up in the editor | Impact Cards after two verification runs | | Cadence | Whenever you remembered | Weekly to daily, depending on plan | ## Who OptimizeCamp 2.0 Is For 2.0 is for anyone whose category is being recommended by an AI answer before a human ever reaches their site. In practice that means four groups, and the fit is sharpest for teams who can actually publish. SaaS and product teams Your category has "best tool for X" queries. Someone is being named in them. Track whether it is you, and which comparison page is doing it. Agencies and consultants Clients now ask why they are missing from ChatGPT. Ten projects, white-label PDFs and read-only client share links answer it with evidence. Content and SEO teams You already have a publishing engine. What you lack is a queue ordered by measured citation loss rather than search volume. Founders doing their own marketing No team, no time to guess. The Action Center hands you the next task and tells you why it is next. It is a weaker fit if you have no way to ship content. OptimizeCamp will tell you exactly what to fix. Somebody still has to fix it. If you are not sure your site is even the right lever, read [Do You Need A Website For AI Visibility?](https://blog.optimizecamp.com/do-you-need-a-website-for-ai-visibility/) first. ## What Is Coming Next Gemini and Copilot are the next two answer engines, and they slot into the same pipeline as the current three. Beyond that, the roadmap follows the loop rather than the feature list. More evidence types feeding the Action Center, more play formats beyond roundup pitches, and tighter verification windows on Impact Cards. Pricing is early access and starts at $79 a month on Starter. Every plan runs the full loop and pays only for volume. Every first charge carries a 14-day money-back guarantee. ## Frequently Asked Questions ### Is OptimizeCamp 1.0 Going Away? No. The 1.0 auditor is now the Optimize module inside 2.0, with a fourth engine added for AI coverage. You can still import any page by URL or paste it straight into the editor. What changed is that the priority queue now decides which page deserves the audit, based on measured citation loss. ### How Long Does Setup Take? About 15 minutes, and three decisions. You paste a URL, we read your sitemap and build a brand profile, then you confirm that profile, approve the competitor slate, and pick your prompts. The baseline run streams results in live. No tags to install and no data to upload. ### Which AI Platforms Does OptimizeCamp Track? ChatGPT, Google AI Overviews and AI Mode, and Perplexity. Gemini and Copilot are on the roadmap. Run cadence depends on your plan: weekly on Starter, daily Google AI plus three times a week for ChatGPT and Perplexity on Pro and Agency. ### How Is The Visibility Score Calculated? It blends four signals from real answers: mention rate, citation rate, position of your first mention, and sentiment of the surrounding sentence. The result is a single 0 to 100 number, computed on identical rules for you and every tracked competitor. Trends use 7-day rolling averages so daily variance does not read as movement. ### Do I Need WordPress To Use OptimizeCamp? No. WordPress gets one-click sync and publish-back over the REST API, which is the smoothest path. On any other CMS you import pages by URL or paste them in, then copy the optimized version back. Tracking, competitors, sources and plays work the same regardless of your stack. ### How Do I Know A Fix Actually Worked? An Impact Card, issued after the next two tracking runs land within fourteen days. It compares the prompts that page targets before and after, and reports citation gained, mention rate change and Visibility Score delta. When the evidence is inconclusive, the card says so and makes no claim. Start The Loop Find out whether AI is recommending you or your competitor. Paste your URL. Fifteen minutes later you have a baseline, your three weakest prompts flagged, and a direct path into fixing them. [Get started free](https://optimizecamp.com/signup) Or run the [free AI Visibility Checker](https://optimizecamp.com/tools/ai-visibility-checker), no signup required. Published August 2, 2026 · Last updated August 2, 2026 · OptimizeCamp, a product of DotCamp LLC