<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by WIRE.AD on Medium]]></title>
        <description><![CDATA[Stories by WIRE.AD on Medium]]></description>
        <link>https://medium.com/@wire.ad?source=rss-5472296675bd------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*XOn4DcBz1PLzqFPcEtHHIA.png</url>
            <title>Stories by WIRE.AD on Medium</title>
            <link>https://medium.com/@wire.ad?source=rss-5472296675bd------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 10:53:37 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@wire.ad/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Launching during Google’s November core update? Read your search data a week later]]></title>
            <link>https://medium.com/@wire.ad/launching-during-googles-november-core-update-read-your-search-data-a-week-later-a25dfa4dbbef?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/a25dfa4dbbef</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[seo]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[press-release]]></category>
            <category><![CDATA[public-relations]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Mon, 16 Nov 2026 00:00:08 GMT</pubDate>
            <atom:updated>2026-11-16T06:45:17.085Z</atom:updated>
            <content:encoded><![CDATA[<p>When to compare Search Console numbers, where Google Search traffic lands in analytics, CDN blocks, AI impersonation ads, the FCA’s stablecoin promotion rules and Bing’s “-ai” switch. Signal Weekly, November 16, 2026.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wv4j3dddYeWUbNKJ.png" /></figure><p>If your launch goes out before December, the November 2026 core update will still be moving your search numbers while you try to read them. Google released it on November 10 and says the rollout may take up to two weeks. Two days later it published a page on finding Google Search traffic in your own analytics. Also this week: a Google Ads rule on AI impersonation of public figures, the FCA consultation that closes Wednesday, and an off switch for AI answers in Bing.</p><h3>This week in 30 seconds</h3><ul><li>Google released the November 2026 core update on November 10. The rollout “may take up to 2 weeks to complete,” and Google advises waiting at least a full week after it ends before comparing data in Search Console.</li><li>A new Google Search Central page, added November 12, says Google Search uses the origin referrer policy, so the Referer header contains only the scheme and host and doesn’t include the search query.</li><li>Google warned on November 11 that CDN or firewall challenges make its crawlers back off, and pages “will begin to drop out of Search” if the block stays.</li><li>Google Ads clarified on November 11 that ads impersonating public figures using AI are not allowed “even if they disclose that they were generated or edited using AI.” The FTC’s docket on platform impersonation ads showed 156 posted comments on November 16.</li><li>The FCA’s CP26/36 closes on November 18. Its draft guidance says the new technical services exclusion “has not been replicated in the financial promotions regime.”</li><li>Microsoft described two ways to search Bing without AI answers: a browser extension, in preview, and adding “-ai” to a search. New York’s crypto advertising guidance of November 12 has <a href="https://wire.ad/articles/nydfs-crypto-advertising-guidance">its own article</a>, and so does <a href="https://wire.ad/articles/claim-your-business-in-chatgpt">Businesses in ChatGPT</a>.</li></ul><p>The Google Search Status Dashboard <a href="https://status.search.google.com/incidents/WLxcJyC6jMWMfryxsKMc?utm_source=wire.ad&amp;utm_medium=referral">logged the update</a> at 10:07 Pacific time on November 10: “Released the November 2026 core update, which applies globally and to all languages. The rollout may take up to 2 weeks to complete.” On the morning of November 16 the incident still carried that single entry, with no completion notice.</p><p>Google’s <a href="https://developers.google.com/search/docs/appearance/core-updates?utm_source=wire.ad&amp;utm_medium=referral">core updates page</a> sets out how to read the data afterwards. Confirm the rollout has finished. Wait “at least a full week after a core update completes before analyzing your site in Search Console,” then compare that week with a week before the rollout started. For a small drop in position, Google says there is “no need to take drastic action,” and it calls deleting content “a last resort.”</p><p>On November 12 Google added a page on how to <a href="https://developers.google.com/search/docs/monitor-debug/identify-google-search-traffic?utm_source=wire.ad&amp;utm_medium=referral">identify Google Search traffic</a> in web analytics or server logs. Google Search uses the origin referrer policy, so the Referer header “contains only the scheme and host of the referring page, without the URL path or query parameters.” It “doesn’t include the search query.” Browsers can still drop the header, for example in private browsing. Among the reasons totals differ from Search Console: client-side redirects may replace the Referer with your own URL, redirects that drop the query string lose parameters, and Search Console counts only organic clicks.</p><p>Our reading: if your launch falls inside the rollout window, launch-week organic numbers tell you little. Add a <a href="https://developers.google.com/search/blog/2025/11/custom-chart-annotations?utm_source=wire.ad&amp;utm_medium=referral">custom annotation</a> in Search Console on the launch date (up to 120 characters; everyone with access to the property sees it). Hold the search comparison until a week after Google marks the rollout complete, and compare against a week before November 10. Before the release goes out, check that your analytics counts referrals from google.com, local Google domains such as google.co.uk, and the Google app on Android as organic, that no custom channel rule or referral exclusion reclassifies them, and that the short links and tracking redirects in the release keep the query string. Our <a href="https://wire.ad/articles/press-release-utm-tracking">UTM tracking guide</a> covers the link side. Report Google Ads clicks separately.</p><h3>Exempt verified crawlers before launch day</h3><p>A <a href="https://developers.google.com/search/blog/2026/11/crawling-cdn?utm_source=wire.ad&amp;utm_medium=referral">Search Central post</a> on November 11 described what happens when a site moves behind a content delivery network. Faster responses lead Google’s crawlers to raise their crawl rate, while a cold cache sends every first request back to the origin server. When a CDN or web application firewall meets crawlers with a CAPTCHA, “Google’s crawlers fail the test,” and if that stays in place, “our crawlers will back off,” and pages “will begin to drop out of Search.”</p><p>Google’s advice is to allow verified bots in CDN and firewall settings, and to “programmatically fetch your highest-traffic, upcoming peak-traffic pages (such as promotional or seasonal launches)” so edge caches are ready. If the origin needs a break, return 429 or 503 “for a short window (a few hours, or 1–2 days at most).”</p><p>Our reading: launch week is a common time to tighten bot rules or move a newsroom to a new host. Ask for the verified-bot exemption in writing before launch day, then warm the cache for the release page, the newsroom post and the product page. AI crawlers need the same check. OpenAI’s <a href="https://developers.openai.com/api/docs/bots?utm_source=wire.ad&amp;utm_medium=referral">crawler documentation</a> says “Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links,” and recommends allowing OAI-SearchBot in robots.txt and allowing requests from OpenAI’s published IP ranges. Add those ranges to the same firewall exemption.</p><blockquote><strong><em>Does AI name you when buyers ask?</em></strong><em> Run a free AI Shortlist Check to see whether AI answers name you.</em></blockquote><blockquote><a href="https://wire.ad/ai-shortlist-check"><em>Get a free AI Shortlist Check</em></a></blockquote><h3>Google Ads rules on AI impersonation of public figures</h3><p>On November 11 Google <a href="https://support.google.com/adspolicy/answer/18891933?hl=en&amp;utm_source=wire.ad&amp;utm_medium=referral">updated its Unacceptable Business Practices policy</a> “to clarify that ads that impersonate public figures using AI are not allowed, even if they disclose that they were generated or edited using AI.” Google says the clarification “does not change the enforcement of the policy.”</p><p>The FTC question is the platform side of the same problem. Its <a href="https://www.federalregister.gov/documents/2026/10/01/2026-20143/rule-on-impersonation-of-government-and-businesses">request for comment</a>, published October 1, asks whether search engine, social media and digital marketplace platforms should face rules to prevent impersonation scams, and comments close November 30. On November 16 the <a href="https://www.regulations.gov/docketBrowser?rpp=50&amp;so=DESC&amp;sb=postedDate&amp;po=0&amp;dct=PS&amp;D=FTC-2026-1552">docket</a> showed 156 posted comments, most from individuals. None of the posted titles names a search engine or social media company. A company whose brand or founder has been impersonated in ads can file one until November 30.</p><p>Our reading: launches attract fake ads, and if a scam ad uses AI to impersonate your founder or another public figure, an AI disclosure line does not make it acceptable under Google’s policy. Before launch day, publish one page that lists your official channels, ad accounts and spokespeople, and link it from the release. When a fake ad appears, report it to Google under this policy. The release and that page give you something concrete to point to.</p><h3>The FCA’s CP26/36 closes Wednesday</h3><p>Last week we said we would read the PERG 8 financial promotion changes in the FCA’s <a href="https://www.fca.org.uk/publications/consultation-papers/cp26-36-cryptoasset-perimeter-guidance?utm_source=wire.ad&amp;utm_medium=referral">CP26/36</a> before it closes on November 18. The <a href="https://www.fca.org.uk/publication/consultation/cp26-36.pdf?utm_source=wire.ad&amp;utm_medium=referral">consultation paper</a> explains that the Cryptoassets Amendment Regulations make issuing qualifying stablecoins a controlled activity and qualifying stablecoins “a controlled investment in their own right.” A new article 73ZAA exemption covers communications about the transfer or exchange of a UK qualifying stablecoin, but that limb does not apply to “other cryptoassets, including other qualifying stablecoins.” The FCA says the effect is that the exemption “applies to many payment-related activities involving UK qualifying stablecoins.”</p><p>A second point sits in the draft PERG 18 guidance on the technical services exclusion (Question 18.8.34). The exclusion applies to arranging activities, and the draft guidance says it “does not determine whether communications made through or in connection with the technical service are subject to the financial promotion restriction.” It adds that the exclusion “has not been replicated in the financial promotions regime,” so providers “should therefore consider separately whether they communicate invitations or inducements to engage in investment activity.”</p><p>The FCA aims to publish final guidance in January 2027. The regime starts on October 25, 2027.</p><p>Our reading: a wallet, interface or DeFi front end that expects to rely on the technical services exclusion should still consider separately whether its UK launch copy is a financial promotion. Before the announcement is written, ask counsel which exemption applies or which authorized firm approves the copy. For a stablecoin payment feature, name the coin and its issuer exactly, because the exemption turns on whether it is a UK qualifying stablecoin.</p><blockquote><strong><em>Announcing to UK readers next quarter?</em></strong><em> Get the copy approved first, then send us the brief. Every WIRE.AD release carries the risk wording of each market it runs in, and the facts are confirmed with you before anything goes live.</em></blockquote><blockquote><a href="https://wire.ad/#brief"><em>Send a brief</em></a></blockquote><h3>Bing offers an off switch for AI answers</h3><p>On November 11 Microsoft <a href="https://blogs.bing.com/search/2026/11/your-choice-turn-ai-answers-on-or-off-in-bing/?utm_source=wire.ad&amp;utm_medium=referral">described two ways</a> to search Bing without AI answers. The Microsoft Bing AI Search Choice extension, in preview for Chrome and Edge, adds a toggle that applies across all Bing searches and also sets Bing as the default search engine. The other way: add “-ai” to any query.</p><p>Our reading: “-ai” gives you a quick side-by-side check. On launch day, search your company and product name in Bing twice, once as usual and once with “-ai”, and save both pages. If the AI answer describes the company from an old page and leaves the launch out, fix that page before the campaign spends. Our <a href="https://wire.ad/ai-visibility">AI visibility</a> add-on records what AI assistants say about a company before and after a campaign.</p><h3>Number of the week: 156</h3><p>That is how many posted comments the regulations.gov docket for the FTC’s impersonation rule (FTC-2026–1552) showed on November 16, two weeks before comments close. The organizations among them include scam victims’ groups and a security company.</p><h3>What we’re watching next week</h3><p>We’ll check the Search Status Dashboard for the end of the core update rollout, and the FTC docket for filings from the platforms themselves. After CP26/36 closes, the FCA aims to publish final guidance in January 2027. OpenAI has yet to give a date for the shared training signal. Its crawler documentation, checked November 16, still names GPTBot as the training control. The SFC’s news releases through November 16 show no misleading-names bill yet.</p><p><em>Originally published at </em><a href="https://wire.ad/articles/signal-weekly-2026-11-16"><em>https://wire.ad</em></a><em> on November 16, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a25dfa4dbbef" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[351 press releases, 30 days: where the coverage went]]></title>
            <link>https://medium.com/@wire.ad/351-press-releases-30-days-where-the-coverage-went-cc326d0bbb1e?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc326d0bbb1e</guid>
            <category><![CDATA[press-release]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[public-relations]]></category>
            <category><![CDATA[startup-marketing]]></category>
            <category><![CDATA[crypto-marketing]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Thu, 05 Nov 2026 00:00:36 GMT</pubDate>
            <atom:updated>2026-11-05T11:34:10.654Z</atom:updated>
            <content:encoded><![CDATA[<p>A random sample of crypto, fintech and AI announcements, which kinds earned a news article, and which pages AI search cited instead.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*T3fqdqnP92qCSe4t.png" /></figure><p>Press release pickup is far thinner than most launch plans assume. WIRE.AD drew a random sample of 351 press releases that crypto, fintech and AI companies sent through major newswires in August and September 2026, then followed each one for 30 days. 65% got no news article written about them. 39% left no trace in news search beyond the newswire page itself.</p><p>The second half of the study looks at AI search, the other place a buyer may go to check a company. We asked an OpenAI model with web search about every one of those announcements, twice: once by company name, once by describing the news without the name. The results say a lot about which pages AI answers actually read.</p><h3>The release pickup study in numbers</h3><p>Releases with no news article in 30 days: 65.0% (228 of 351)</p><p>Releases with no trace beyond the newswire page: 38.7% (136 of 351)</p><p>Releases with only reprints, aggregator copies or listings: 26.2% (92 of 351)</p><p>Releases with at least one news article: 35.0% (123 of 351)</p><p>Releases with three or more news articles: 11.1% (39 of 351)</p><p>Share of all news articles taken by the top 10% of releases: 65%</p><p>AI search described the news when asked by company name: 68.9%</p><p>AI search named the company when asked about the news without the name: 59.3%</p><p>A news article here means a story on a news publication under its own headline. A reprint is the release text republished under the original headline. Aggregators are finance and crypto portals that repost wire feeds automatically.</p><h3>Most releases stop at the wire</h3><p>The typical release in the sample earned zero articles, and the median was zero in every vertical. Crypto releases got at least one article 34.5% of the time, fintech 33.1%, AI 37.7%. The 95% confidence range for the headline figure is 60% to 70%, so the gap is not a quirk of the sample.</p><p>Reprints made the picture look busier than it was. A quarter of releases (26.2%) did show up on other sites, but only as copies on finance portals, market-data pages and exchange news feeds. Those pages repeat the headline word for word. They prove the release was distributed.</p><p>They rarely put it in front of a reader who was not already looking for the company, which is the job most launch budgets are paying for.</p><p>Coverage was also heavily concentrated. The top 10% of releases collected 65% of all the news articles in the sample. A handful of announcements, mostly from large companies or with a well-known partner attached, picked up between five and twenty stories each.</p><p>Most of the rest got none.</p><p><strong>How WIRE.AD does it.</strong> Every package pairs the wire with placements on outlets chosen for the client’s vertical, and each one gets a version of the release written for that outlet. The report lists every placement with its URL and its label, so the client sees articles and reprints separately, never as a single count.</p><h3>Funding news and original data earn articles</h3><p>The kind of news mattered more than the vertical. Half of all funding announcements (50%) got at least one article, and releases built on original research or survey data did best at 55%. Executive hires, event and award news, and financial results sat at 25% to 28%.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Wh6mbjYseKqTcB2a.png" /></figure><p>Partnerships landed in the middle at 41.9%, and product launches just below at 36.8%. The ones that drew several stories tended to come from companies reporters already follow, or named a large partner such as a stock exchange or a global bank. Two practical points follow. A release about a sponsorship, a conference booth or an award is distribution, and should be priced and planned as distribution. A release with a number journalists can quote, a funding amount or a dataset, has a real chance of earned coverage and deserves outlet outreach on top of the wire.</p><p><strong>How WIRE.AD does it.</strong> The brief asks what the news is before it asks where it should run, and the proposal matches outlets to the story: funding and data news to outlets that cover deals and numbers, event and award news to labeled placements built for reach.</p><h3>AI search reads the company’s own newsroom</h3><p>When we asked “What did [company] announce in [month]?”, the answer described the right news 68.9% of the time. The links behind those answers were the surprise. 59.3% of all citations pointed to the company’s own website: its newsroom, blog or investor page. Newswire pages made up 18.3%. News outlets made up 6.9%.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*oWeBgHFodattMtaJ.png" /></figure><p>For a branded question, then, the most important page is one the company controls. A release that lives only on the wire and never on the company’s own site gives AI search less to work with. We covered the wire side of this in <a href="https://wire.ad/articles/chatgpt-newswire-citations">our look at newswire citations in ChatGPT</a>.</p><p>Questions without the company name behaved differently, and this is where earned coverage starts to count. When the question described the news instead (for example, “Which crypto infrastructure firm launched an institutional OTC desk in September?”), news outlets rose to 14% of citations, almost level with newswires at 14.9%.</p><h3>Companies with articles were named more often</h3><p>This is the finding that matters most for launches. On the question that described the news without naming the company, AI search named the company 70.7% of the time when the news had at least one article. When it had none, the figure was 53.1%.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8j8EJlhlkbTu_FiU.png" /></figure><p>The difference is statistically clear (p ≈ 0.001), and it held whether the news had one article or several. It is still a correlation. Companies that earn articles may also be better known, and AI answers vary from one run to the next. What the data does show is the direction: an independent article gives AI search a second source that tells the same story in different words, and answers that do not start from your name find you more often.</p><p><strong>How WIRE.AD does it.</strong> Any campaign can add an AI visibility reading before and after, built on the questions your buyers ask, both branded and unbranded. The <a href="https://wire.ad/ai-visibility">AI visibility page</a> explains the reading, and the free <a href="https://wire.ad/ai-shortlist-check">AI Shortlist Check</a> shows where your company stands today.</p><h3>What to do with this before your next launch</h3><ul><li>Publish the full release on your own newsroom the same day it goes out on the wire. It is the page AI search cites most.</li><li>Treat the wire as distribution, then budget separately for coverage on outlets your buyers read.</li><li>Lead with news that has a number in it: a funding amount, a usage figure, a dataset.</li><li>Ask for a report that separates articles, reprints and aggregator copies. One total hides the difference.</li><li>Check AI answers before and after, with questions that do not include your company name.</li></ul><p>To plan coverage around your next announcement, <a href="https://wire.ad/#brief">send a brief</a> and WIRE.AD replies with a fixed-price package. The <a href="https://wire.ad/crypto-press-release-distribution">crypto press release distribution</a> page explains what a package includes.</p><h3>How we ran the study</h3><p>We took every release listed from 1 August to 30 September 2026 in the cryptocurrency, financial technology and artificial intelligence feeds of a large business newswire, plus the full feed of a crypto-focused newswire: 5,866 releases. From these we drew a stratified random sample of 360 (120 per vertical). Each draw had to be in English, come from a company working in that vertical and be a real company announcement. Correction notices, duplicates and two non-English items found on review were removed, leaving 351.</p><p>For each release we searched Google News for the company name from the day before publication to 31 days after, and checked each result against the release. We then asked an OpenAI model with web search two questions about each announcement on 4 November 2026 and recorded every cited link. Finally, we reviewed a random set of 79 of those judgments by hand, and 76 of them held up.</p><p>Three limits apply. Google News misses many small crypto sites, paywalled trade titles and non-English press, so true coverage is somewhat higher than measured. Some articles under their own headline may be paid placements, which search results do not reveal. And one newswire supplied most of the fintech and AI sample, so its industry categories (which include banks and asset managers) shape those groups.</p><p><em>Originally published at </em><a href="https://wire.ad/articles/press-release-pickup-study"><em>https://wire.ad</em></a><em> on November 5, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cc326d0bbb1e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The press release headline Google shows is assembled, so write it for assembly]]></title>
            <link>https://medium.com/@wire.ad/the-press-release-headline-google-shows-is-assembled-so-write-it-for-assembly-183ee8119a8e?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/183ee8119a8e</guid>
            <category><![CDATA[press-release]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[public-relations]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[seo]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Tue, 03 Nov 2026 00:00:55 GMT</pubDate>
            <atom:updated>2026-11-03T06:37:08.854Z</atom:updated>
            <content:encoded><![CDATA[<p>Why search results swap the title you approved, what a 2022 study of 80,959 titles says about length and brackets, and a checklist for launch teams.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*c_ERVCN9qKWtjZbi.png" /></figure><p>A press release headline that holds up in search names the company and the news in the first few words, runs about 51 to 60 characters, and appears the same way in the page’s title tag, its H1 and its og:title on every outlet that runs it. Google builds the title a buyer sees from several parts of the page.</p><p>It changes titles that are vague, boilerplate or out of step with the page, and one vendor study found it rewrote most titles that were very long or very short. A headline written for that process is the one most likely to reach the reader, and the AI answers that read the same pages, in the words you approved.</p><h3>Google assembles the headline buyers see</h3><p>The headline you send is an input. Google calls the clickable line in a result a title link, and its documentation says the process is “completely automated and takes into account both the content of a page and references to it that appear on the web.” The same title link shows on other Google properties, Google News among them.</p><p>Google lists nine sources for it: the &lt;title&gt; element, the main visual title on the page, heading elements such as the H1, the og:title meta tag, other large and prominent text, other text on the page, anchor text on the page, the text of links pointing to the page, and WebSite structured data. The og:title entry was added in August 2024.</p><p>For a release, that list matters because one announcement lives on many pages. Each outlet that runs it has its own template, its own title tag and its own heading styles. One adds its site name in brackets, another puts a category label in bigger type than your headline. Google picks page by page, and its documentation says that when a page has “more than one large, prominent heading,” it “may use the first heading as the text for the title link.”</p><p>Rewrites are common. In Zyppy’s 2022 vendor study of 80,959 title tags on 2,370 sites, Google showed a different title for 61.6% of them in desktop results. Many changes were small, but it is the most detailed public measure we found of which title shapes Google leaves alone.</p><p>There is also a newer layer. In March 2026 Google confirmed to The Verge a “small” and “narrow” experiment that used generative AI to replace headlines in Search results, “not specific to news publications.” Its stated goal was to “identify content on a page that would be a useful and relevant title to a users’ query.” In one example The Verge cited, a headline that ended in a negative verdict on an AI tool was cut to five words that, in the publication’s words, “almost sounds like we’re endorsing” it. Google said that a launch based on the test “would not be using a generative model.” We found no later announcement of a wider rollout, and the title link documentation does not mention generated titles. Before you approve a headline, cut it after the first six words and check that what is left still states the news correctly.</p><p><strong>How WIRE.AD does it.</strong> Facts are confirmed with the client before publication, and each outlet gets its own version. Every version comes with an SEO title and description written from the confirmed facts, so the copy each outlet receives carries the wording you approved.</p><h3>Keep the title tag to 51 to 60 characters</h3><p>Length is the cause of rewrites you control most directly. Google’s own guidance gives no limit and says the title link is “truncated in Google Search results as needed, typically to fit the device width.” The Zyppy data shows where the trouble starts.</p><ul><li>Titles over 70 characters were rewritten 99.9% of the time.</li><li>Titles over 60 characters were rewritten more than 76% of the time.</li><li>Titles of 20 characters or fewer were rewritten more than half the time; very short ones were usually padded with extra words.</li><li>The 51 to 60 character range had the lowest rewrite rate, 39% to 42%.</li></ul><p>So a 90-character headline that reads well in a pitch email will almost never show in search as written. The on-page headline can run longer, as long as its first 50 or so characters carry the company and the event.</p><p>“Acme Pay launches instant payouts for freelancers in Brazil” is 59 characters. It names the company, the product and the market. “Acme Pay Unveils Revolutionary Next-Generation Payment Experience” says less and is longer.</p><p><strong>How WIRE.AD does it.</strong> Every package includes an SEO title and description for each outlet version. The longer headline stays on the page, where readers have room for it.</p><h3>Use one headline across title tag, H1 and og:title</h3><p>The second cause is a mismatch between the parts of the page Google reads. If the title tag says one thing and the H1 says another, Google has a reason to choose, and it may use the visible heading instead.</p><p>Zyppy measured this with numbers in titles. When the title tag held a number and the H1 did not, Google showed a title without the number 25.8% of the time. When both held the number, Google kept a number in the title 97.3% of the time. For a funding round, a user count or a token supply, that gap is the difference between “Acme Chain raises $12M” and “Acme Chain raises seed round” in the result a buyer reads.</p><p>The og:title is the third piece. The Open Graph protocol defines it as “the title of your object as it should appear within the graph,” and Google now lists it among its title sources. Outlet templates set it in different ways, so ask your distributor which title tag and og:title each outlet will publish.</p><p>A clean release version has:</p><ol><li>A title tag of 51 to 60 characters with the company and the event.</li><li>An H1 that matches it or extends it without changing the facts.</li><li>An og:title with the same words as the title tag.</li><li>One heading styled as the main title, with category labels and outlet slogans in smaller type.</li></ol><p><strong>How WIRE.AD does it.</strong> Every live page is audited for its label, links and risk wording, and the report lists each placement with its label, level and role in the package.</p><h3>Put the news first and keep the brand short</h3><p>Google’s guidance on branding is direct: put the site name “at the beginning or end of each &lt;title&gt; element, separated from the rest of the text with a delimiter such as a hyphen, colon, or pipe.&quot; Outlets add their own name this way, and that name comes out of your character budget.</p><p>The punctuation affects survival too. In the Zyppy sample, titles with square brackets were rewritten 77.6% of the time, and Google dropped the bracketed text entirely in 32.9% of cases. Parentheses fared better, with the text removed 19.7% of the time. Pipes were removed or replaced 41.0% of the time and dashes 19.7%. A headline like “[Press Release] Acme Chain lists ACME token on two exchanges” puts the least useful words where Google is most likely to cut them, and the label belongs on the page anyway, next to the headline.</p><p>Word order is the other half. When a title link is truncated to fit the screen, the end goes first. Put the company and the event at the start, and the qualifier after it: “Acme Chain opens staking for ACME holders in the EU” survives being cut at the word “holders.” A headline that opens with “In a major milestone for the industry” loses the news first.</p><p><strong>How WIRE.AD does it.</strong> Every paid placement is labeled Sponsored or Press release on the page itself. Our guidance for SEO titles is to lead with the company and the event. Labels and risk wording sit in the body, never in the title.</p><h3>Headlines for crypto, fintech and AI announcements</h3><p>Regulated announcements add a second test. The headline is the line most often quoted alone, in a title link, a link preview or the source list under an AI answer, so it has to be accurate without the paragraph that qualifies it.</p><p>A few patterns cause problems in this niche:</p><ul><li>A headline that says “approved,” “licensed” or “registered” needs the exact status and the regulator, because a trimmed title can drop “applies for” and leave the company name next to the regulator’s. Write the status the filing supports, in the first half.</li><li>Yield, return or price language in a token headline turns a news line into a promotion in several markets. Keep the event (a listing, a mainnet date, an audit) and leave performance out of the headline.</li><li>“First,” “largest” and “leading” claims need a source in the body, and in a headline they rarely survive a rewrite with their qualifier attached.</li><li>AI launch headlines often name a model or benchmark. Name the product and what it does first. The benchmark can sit in the subhead.</li></ul><p>Our crypto guide on <a href="https://wire.ad/articles/how-to-write-a-crypto-press-release">how to write a crypto press release</a> covers the full release shape and the risk wording by market. The headline rule here is narrower. Read the first half of your headline alone; if it overstates the filing or the product, rewrite it before it goes out.</p><p><strong>How WIRE.AD does it.</strong> Each version carries risk wording where needed, and the buyer’s claims are attributed to the buyer. Every paid placement is labeled. The same rules apply to <a href="https://wire.ad/crypto-press-release-distribution">crypto press release distribution</a> orders.</p><h3>Check what searchers and AI answers show after launch</h3><p>Google has to recrawl a page to notice a changed title, which “may take a few days to a few weeks.” Checking on launch day tells you what the outlet published. Checking a week later tells you what Google chose to show.</p><p>Search each outlet URL and compare the title link with the headline you approved. When they differ, look for the cause in Google’s list: a mismatched H1, a missing title tag, a second large heading, a site name that took up the space. Send the mismatch to your distributor or the outlet; some can fix the title tag in their CMS, others cannot.</p><p>The same check applies to AI answers. Ask an assistant the question a buyer would ask about your launch and see which page it cites and how it names the event. If the cited page has a trimmed or wrong title, that is the outlet version to fix first. Our <a href="https://wire.ad/ai-visibility">AI visibility</a> service measures this before and after a campaign, and it promises no citations. The free <a href="https://wire.ad/ai-shortlist-check">AI Shortlist Check</a> shows how an AI assistant answers five buyer questions in your category today.</p><p><strong>How WIRE.AD does it.</strong> With the AI visibility service, AI answers are measured before placements go out and again four to eight weeks later, on the same questions. Every package includes 30 days of corrections: if your facts change after launch, the live pages are updated.</p><h3>A press release headline checklist</h3><ul><li>The company name and the event sit in the first 50 characters.</li><li>The title tag runs 51 to 60 characters, including the outlet’s site name.</li><li>The H1 and og:title carry the same facts and numbers as the title tag.</li><li>No brackets around labels or dates; labels sit on the page next to the headline.</li><li>Status words such as “approved” or “licensed” match the actual filing.</li><li>No yield, return or price language in a token headline.</li><li>One main heading per page, styled larger than any category label.</li><li>Every outlet URL is searched a week after launch and compared with the approved headline.</li></ul><p><em>Originally published at </em><a href="https://wire.ad/articles/press-release-headline"><em>https://wire.ad</em></a><em> on November 3, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=183ee8119a8e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Signal Weekly Nov 2: AI training opt-out in robots.txt]]></title>
            <link>https://medium.com/@wire.ad/signal-weekly-nov-2-ai-training-opt-out-in-robots-txt-07506edff84a?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/07506edff84a</guid>
            <category><![CDATA[seo]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[press-release]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Mon, 02 Nov 2026 00:00:49 GMT</pubDate>
            <atom:updated>2026-11-02T08:35:01.542Z</atom:updated>
            <content:encoded><![CDATA[<p>One robots.txt line for AI training, and four regulator moves that matter for announcements Subtitle: Google’s shared AI training opt-out and Gemini grounding, VARA’s token disclosure review, an SEC case on filings shown as approval, plus the FCA and the FTC, October 26 to November 1, 2026.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*YRe4YGQ9DcuuZGtP.png" /></figure><p>If your newsroom adds an AI training opt-out to robots.txt next year, the same line can take your launch out of Gemini’s cited answers. That is the change to settle first this week. Also: what VARA expects Dubai exchanges to publish per token, why “filed with the SEC” must never become “approved”, which channels the FCA counts in a promotion, and the “$0” claim the FTC took to court.</p><h3>This week in 30 seconds</h3><ul><li>Google plans to treat a shared AIPREF opt-out in robots.txt as a Google-Extended block, which also covers grounding in Gemini Apps. Google aims to support it in 2027.</li><li>VARA’s circular of October 28 restates what Dubai broker-dealers and exchanges must already publish for every token under its rulebooks, and shows where its review found gaps.</li><li>The SEC charged two entities it says showed investors false SEC filings as proof of approval.</li><li>The FCA settled its court case over HTX with High Court undertakings that cover the website, apps and social media.</li><li>The FTC and state partners sued Toast over a “$0 per month” plan and a “24/7/365 support” promise.</li></ul><p>On October 30 Google <a href="https://developers.google.com/search/blog/2026/10/ai-training-control?utm_source=wire.ad&amp;utm_medium=referral">announced plans</a> to support a shared AI training control in robots.txt, built on the AIPREF working group at the IETF. A site will set one “policy token” for all participating AI providers instead of writing a rule for each one. When a site opts out of training with that token and says nothing about Google-Extended, Google “plans to interpret that as if the publisher had disallowed Google-Extended.” Company-specific controls keep working alongside it. Google’s <a href="https://developers.google.com/crawling/docs/crawlers-fetchers/google-common-crawlers?utm_source=wire.ad&amp;utm_medium=referral">crawler documentation</a> gives the token as user-agent: __usage_ai-training and the start as early 2027.</p><p>The detail that matters for announcements is what Google-Extended covers. Per the same documentation, it controls Gemini training and also grounding in Gemini Apps, and the blog post says sites that “specifically allow Google-Extended are eligible to appear in responses and receive clicks via grounding in Gemini Apps.” Google-Extended has no effect on inclusion in Google Search or on ranking.</p><p>Our reading: a company that adds the shared opt-out to its newsroom without a Google-Extended line may drop out of Gemini’s grounded answers about its own launch. Google-Extended is one switch for both Gemini training and Gemini grounding, so decide before 2027 which matters more for your newsroom. An explicit Google-Extended allow next to the shared token keeps you eligible for grounding in Gemini Apps, but it also lets Google use the pages for Gemini training. The shared token would then opt you out of the other participating providers only. Ask whoever runs your CMS or CDN before they ship a blanket AI block.</p><p>Our <a href="https://wire.ad/ai-visibility">AI visibility</a> service records what AI assistants say about a company before a campaign and asks the same questions again afterwards. It measures; it does not guarantee citations.</p><h3>VARA circular on public disclosures</h3><p>VARA <a href="https://www.vara.ae/en/news/?utm_source=wire.ad&amp;utm_medium=referral">published a circular on October 28</a> with the results of a thematic review, begun in June 2026, of the public disclosures of Dubai-licensed crypto firms. The review covered a representative population of licensed firms, and token data is one part of it. The <a href="https://media.umbraco.io/dwtc/dczjo4qp/public-disclosure-circular-for-vasps-final.pdf?utm_source=wire.ad&amp;utm_medium=referral">circular</a> creates no new obligation: it restates duties already in VARA’s rulebooks and describes good practice. It found disclosures that were incomplete, vague, hard to find or out of date, and says some licensed firms treated disclosure “primarily as a licensing deliverable, rather than as a continuing regulatory obligation.” For every token offered, broker-dealers and exchanges must publish, among other things, the name and symbol, issuance date, market capitalization and fully diluted value, circulating supply, smart-contract audit information and specified price history. VARA found cases where that data was missing or covered only some tokens, and cases where firms linked out to third-party data sites instead of disclosing the data themselves. Its good-practice example is a dedicated asset-information page for each token. It also flags statements that client assets are “protected”, “secure” or “segregated” with no explanation behind them.</p><p>For a token team listing on a Dubai-licensed exchange, our reading: the exchange already has to publish this data, and the circular shows supervisors are checking it, so expect the exchange to ask you for it in a fixed format. Send a token fact sheet with the listing request that follows VARA’s fields: name and symbol, issuance date, circulating supply, market capitalization and fully diluted value with an as-of date, price history, and the date and author of the last smart-contract audit. Use the same figures and the same as-of date in the listing announcement. The same caution applies to words like “secure” in the release itself. Our <a href="https://wire.ad/articles/crypto-pr-dubai">Dubai crypto PR guide</a> covers VARA’s marketing rules.</p><h3>SEC case over filings shown as proof of approval</h3><p>On October 29 the SEC charged two Colorado-incorporated entities run from abroad, and its <a href="https://www.sec.gov/enforcement-litigation/litigation-releases/lr-26679">litigation release of October 30</a> sets out the allegations. From at least October 2025 through June 2026, the scheme promised risk-free returns from crypto trading “signals” and raised more than $130 million from tens of thousands of retail investors. The complaint says the defendants filed false Form ADV and Form D documents that were “shared with investors as supposed proof that the SEC had licensed or approved them.” Neither entity was registered with the SEC. The filings have been removed from the SEC website.</p><p>Legitimate funding announcements slip into the same vocabulary. A filing on EDGAR shows that a document was filed. Describe a filing exactly, by form name and date. Use “SEC-registered” only if the company is actually registered (a Form D is <a href="https://www.sec.gov/resources-small-businesses/exempt-offerings/filing-form-d-notice">a notice of an exempt offering</a>, not a registration), and never say “SEC-approved”: the SEC does not approve companies or offerings. Check the investor deck and the press kit for the same slip.</p><blockquote><strong><em>Announcing a round or a listing?</em></strong><em> Every WIRE.AD release goes out with the risk wording of each market it runs in, and every placement is labeled editorial or sponsored.</em></blockquote><blockquote><a href="https://wire.ad/#packages"><em>See the packages</em></a></blockquote><h3>FCA settles its case over HTX</h3><p>The FCA <a href="https://www.fca.org.uk/news/press-releases/fca-stops-htxs-crypto-promotions?utm_source=wire.ad&amp;utm_medium=referral">said on October 29</a> that it had settled its court case over HTX, formerly Huobi, and secured High Court undertakings from Huobi Global S.A., the first defendant, that it will not knowingly promote crypto services unlawfully to UK consumers. The undertakings cover “the HTX website, software applications and social media platforms.” HTX has left the UK: deposits are closed and new UK users cannot sign up. The settlement terms are confidential and carry no admission of liability. Proceedings began on October 21, 2025. “Being based overseas does not put them beyond the reach of UK law,” said Steve Smart, the FCA’s executive director of enforcement and market oversight.</p><p>For non-UK projects, the useful part is the list of channels. The undertakings name social media alongside the website and apps. Our reading: a launch plan that excludes the UK should cover social posts and paid targeting as well as the website, and the sign-up flow should block UK users too; check the details with UK counsel. Before launch, list every channel that will carry the promotion and confirm the UK exclusion on each one.</p><h3>FTC lawsuit over a fintech’s “$0” and “24/7” claims</h3><p>The FTC, joined by seven state attorneys general and two California local offices, <a href="https://www.ftc.gov/news-events/news/press-releases/2026/10/ftc-states-take-action-against-toast-inc-over-hidden-unauthorized-fees-deceptive-subscription">sued Toast on October 29</a>. The complaint says Toast advertised a “Pay-as-you-Go” plan with hardware and software for “$0 per month” while charging fees it did not clearly disclose, including $85 a month for payment processing during a temporary closure and an early termination fee of $150 for each remaining month. It also targets the “24/7/365 support” listed in every subscription package, and a $0.99 fee added to some online orders above $10 from September 2021 to late July 2023. The Commission voted 2–0. These are allegations, and the court will decide the case.</p><p>A fintech launch release repeats the same kind of claims: free tiers, zero fees, round-the-clock support. If support runs only on weekdays, the announcement should say so, and the landing page it links to should say the same thing.</p><h3>Number of the week: $150</h3><p>That is the early termination fee the FTC says Toast charged for each remaining month of a contract sold next to a “$0 per month” headline, according to the <a href="https://www.ftc.gov/news-events/news/press-releases/2026/10/ftc-states-take-action-against-toast-inc-over-hidden-unauthorized-fees-deceptive-subscription">October 29 press release</a>. The complaint says the fees were not clearly disclosed. It is an allegation, and the court will decide.</p><p>For a fintech launch, the lesson sits in the gap between the headline and the fee. A “$0” or “free” claim needs its conditions in the same paragraph, in the release and on the page it links to, because a regulator reads them together.</p><blockquote><strong><em>Want to see how your last announcement reads in AI answers?</em></strong><em> Run a free AI Shortlist Check and add your last PR report: we check whether AI names you when buyers ask and review each placement, whether it is live, labeled, indexed and how it links to you.</em></blockquote><blockquote><a href="https://wire.ad/#mini-audit"><em>Get a free AI Shortlist Check</em></a></blockquote><h3>What we’re watching next week</h3><p>The FTC docket on platforms’ role in impersonation scams has four weeks left, and we’ll look at who files before November 30. We’ll also read the FCA’s <a href="https://www.fca.org.uk/publications/consultation-papers/cp26-36-cryptoasset-perimeter-guidance?utm_source=wire.ad&amp;utm_medium=referral">CP26/36</a> for its PERG 8 changes on financial promotions, and track whether other AI providers commit to the shared training token.</p><p><em>Originally published at </em><a href="https://wire.ad/articles/signal-weekly-2026-11-02"><em>https://wire.ad</em></a><em> on November 2, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=07506edff84a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Get Cited by ChatGPT: Win the Page It Links]]></title>
            <link>https://medium.com/@wire.ad/how-to-get-cited-by-chatgpt-win-the-page-it-links-3bdc4347d3f3?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/3bdc4347d3f3</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[ai-search]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[seo]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Fri, 30 Oct 2026 00:00:53 GMT</pubDate>
            <atom:updated>2026-10-30T06:41:58.083Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KFf66znFl8x0Qpf9.png" /></figure><p>To get cited by ChatGPT, get your company onto the pages ChatGPT already links when it answers a buyer’s question, and make sure its search crawler can read your own pages. A <a href="https://derivatex.agency/report/b2b-saas-ai-citation-study/?utm_source=wire.ad&amp;utm_medium=referral">study by DerivateX</a> found that ChatGPT cited the recommended tool’s own website for 11.6% of its software recommendations.</p><p>Of the recommendations that carried a citation, 87.4% linked a third party instead: a niche blog, a competitor’s comparison post, a media article or a Reddit thread. For a crypto, fintech or AI company, how to get cited by ChatGPT is mostly a question of which outside pages carry your name, and how well they carry it.</p><h3>What the study measured</h3><p>DerivateX ran 40 buyer questions, one for each of 40 B2B software categories, through ChatGPT with web search on. Each question went into a fresh temporary chat ten times. The team logged every tool ChatGPT named, whether the tool got a clickable citation, and which domain that citation pointed to. The dataset holds 233 recommendations of 219 distinct tools. The report was published on May 31, 2026.</p><p>Keep three limits in view. The study covers one assistant and one type of question: vendor discovery, the “what are the best tools for X” moment. The categories are B2B software such as CRM, analytics, developer tools and security, so crypto and fintech are adjacent rather than tested. And DerivateX sells AI search services, so it has a commercial view on the result. The numbers are still useful because the method is published and the question type is the one that matters most to a buyer.</p><h3>The citation usually goes to someone else’s page</h3><p>ChatGPT attached a citation to 92.3% of the tools it named. So being named with a link is close to automatic once ChatGPT decides to recommend you. The open question is whose page sits behind the link, and in 87.4% of cited recommendations it belonged to someone other than the vendor.</p><p>Picture the buyer’s side. They read that your product is a strong choice, click the source, and land on a list someone else wrote. That list decides how your product is described, which competitors stand next to it and whether the reader clicks through to you. In 25 of the 40 categories, no recommended vendor was cited through its own website at all.</p><p>This is why a launch plan that stops at “get mentioned” leaves the most valuable part to chance. The cited page is the part a buyer actually opens.</p><p><strong>How WIRE.AD does it.</strong> Each placement in the campaign report is listed with its URL, label and date. That gives you the exact pages to look for when you later check which sources ChatGPT links in answers about your category.</p><h3>Where the cited pages come from</h3><p>The study grouped every citation by source type:</p><p>Review platforms barely register. The large group is specialist content: blogs that cover one category closely, and vendors that publish an honest comparison of their own market. The study names Procurify, a procurement software company whose “best procurement software” article was the cited source for five different brand recommendations inside one answer.</p><p>For crypto, fintech and AI launches, the specialist layer has a clear equivalent. It is the crypto news site that writes category roundups, the fintech trade publication that covers payments infrastructure, and the AI newsletter that compares developer tools. A story on those outlets sits on pages ChatGPT can retrieve for your category’s questions. A placement on a general site with a high authority score may sit nowhere near them.</p><p><strong>How WIRE.AD does it.</strong> Outlets in a package are chosen by vertical and audience first. A crypto exchange gets crypto outlets, a payments company gets fintech outlets, and every placement carries its editorial, sponsored or press release label.</p><h3>Let OAI-SearchBot read your pages</h3><p><a href="https://developers.openai.com/api/docs/bots?utm_source=wire.ad&amp;utm_medium=referral">OpenAI’s crawler documentation</a> is direct on this point. OAI-SearchBot is the agent that surfaces websites in ChatGPT’s search features, and sites that opt out of OAI-SearchBot “will not be shown in ChatGPT search answers”, apart from navigational links. GPTBot is a separate agent for model training, so you can allow OAI-SearchBot and still disallow GPTBot. OpenAI says a robots.txt change takes about 24 hours to reach its search systems.</p><p>Check your own robots.txt, your newsroom and any press page you host. The <a href="https://wire.ad/articles/ai-crawler-controls-press-releases">AI crawler controls guide</a> walks through each agent and what each setting does.</p><h3>Build the page ChatGPT can quote</h3><p>The study retrieved the pages ChatGPT cited and found a consistent shape. Every cited page used a numbered or bulleted list. Most carried the current year in the title, and many included a comparison table and an FAQ. Small sites focused on one category dominated the cited set, and domain authority did not predict which pages were cited.</p><p>Your own site can compete for the citation when it publishes the page buyers would want: a current comparison of your category that names competitors fairly and dates every fact it uses. Product pages rarely win this. For a crypto or fintech company, that comparison also has to carry the same risk and licensing wording your legal team requires on any public claim.</p><p>The category matters too. Own-site citations clustered in technical and developer-facing categories, where vendors publish deep documentation and technical comparisons. The study reported a 50% own-site citation share in single sign-on and identity. If you sell an API, an SDK or infrastructure, your documentation is a citation asset in its own right.</p><h3>Named from memory or cited from search</h3><p>The tools ChatGPT named without any link were mostly long-established category leaders, the names a model can list from its training data. Newer tools in the same answers were the ones pulled from fresh lists and linked. The study describes these as two separate routes into an answer.</p><p>A challenger brand cannot speed up the first route. It can work on the second one now, by appearing in the current pages ChatGPT retrieves and by publishing one of those pages itself. That is also the route a launch can move, because a launch creates news that specialist outlets want to cover.</p><p><strong>How WIRE.AD does it.</strong> An AI visibility reading is taken before the campaign and again after it. WIRE.AD asks the buyer questions for your category in a fresh chat with web search on, records whether your company is named and which page is cited, and compares the two readings. The report names the assistant that gave each answer. The <a href="https://wire.ad/articles/how-to-measure-brand-visibility-in-chatgpt">ChatGPT measurement guide</a> shows how to run a simple version of the same check yourself.</p><h3>A launch checklist</h3><ul><li>Write down the five to ten questions a buyer in your category asks ChatGPT, and save today’s answers with the linked pages.</li><li>Allow OAI-SearchBot in robots.txt on your site, newsroom and documentation.</li><li>Publish one honest category comparison on your own domain, with a table, an FAQ and dated facts.</li><li>Place your launch news on the specialist outlets your category’s answers already link to, labeled Sponsored or Press release where it is paid.</li><li>Ask the same questions four to eight weeks later and compare which pages ChatGPT cites.</li></ul><p>The free <a href="https://wire.ad/ai-shortlist-check">AI Shortlist Check</a> researches five buyer questions in your category with an AI agent and emails the answers with their sources and the method used. To place your launch on outlets chosen for your segment and markets, <a href="https://wire.ad/#brief">send a brief</a> and get a fixed-price package with every placement labeled.</p><p><em>Originally published at </em><a href="https://wire.ad/articles/how-to-get-cited-by-chatgpt"><em>https://wire.ad</em></a><em> on October 30, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3bdc4347d3f3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Press Release Is Not Finished When the PDF Exports]]></title>
            <link>https://medium.com/@wire.ad/a-press-release-is-not-finished-when-the-pdf-exports-107f4b6ca535?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/107f4b6ca535</guid>
            <category><![CDATA[seo]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[press-release]]></category>
            <category><![CDATA[advertising]]></category>
            <category><![CDATA[pr-distribution]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Wed, 21 Oct 2026 11:16:48 GMT</pubDate>
            <atom:updated>2026-10-21T11:16:48.579Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tnBInnlZJbZYX9KXv5k2Fw.png" /></figure><h3>The web page and the download are two public products. Review them that way.</h3><p>A press release can be approved, uploaded and still be wrong in public.</p><p>The reason is simple: the release usually exists in more than one form. There is an HTML page built for reading, navigation and sharing. There may also be a PDF built for downloading, forwarding or archiving. Those versions often begin with the same copy, but they do not stay identical by default.</p><p>A late correction can reach the web page and miss the PDF. A new PDF can be uploaded under the old filename while a cache continues to serve the previous file. A polished document can look correct on screen but contain no usable text. The page can point to the wrong download. The publication date, availability language or supporting links can drift between versions.</p><p>That is why “the copy was approved” is not a complete publication check. The real question is whether every public version carries the approved facts and works for the reader who receives it.</p><h3>Treat each format as a separate output</h3><p>HTML and PDF solve different jobs.</p><p>The web page is usually the best reading surface. It can adapt to smaller screens, connect to related pages and be corrected without asking readers to download another file. The PDF is useful when someone needs a fixed document to save, circulate internally or attach to a record.</p><p>Neither format automatically validates the other. A correct page does not prove that the download is current. A correct file in a project folder does not prove that the live website serves that file.</p><p>The review therefore has to start from the public URLs, not from the working files.</p><h3>First, compare the facts</h3><p>Open the live page and the live PDF side by side. Check the details most likely to change late in production:</p><p>- product stage and availability;<br>- launch or announcement date;<br>- prices, regions and eligibility conditions;<br>- company and product names;<br>- quotations and speaker titles;<br>- supporting links and contact details;<br>- version or revision markers, when used.</p><p>This is not a fresh editorial review of every sentence. It is a comparison of the claims that would create a real contradiction if the two versions disagreed.</p><p>Pay particular attention to cautious language. “Available,” “in beta,” “rolling out” and “planned” are not interchangeable. If legal or product review changed one of those phrases, the same correction must appear everywhere the release is public.</p><h3>Then inspect the PDF as a document</h3><p>A visual scan is necessary, but it is not enough.</p><p>Select and copy a paragraph. Search for a phrase that is visible on the page. Check the reading order if the document uses columns, sidebars or decorative layouts. Open every important link. Review the document on a smaller screen instead of assuming that a desktop export is comfortable to read.</p><p>A PDF made from page images may look finished while offering no dependable text to search, select or read with assistive technology. W3C’s PDF7 technique describes using optical character recognition to add actual text to scanned PDF content.</p><p>This is also the point to check the unglamorous production details: cropped marks, missing characters, broken line wraps, accidental blank pages and metadata left over from an earlier draft.</p><h3>Verify what the website actually serves</h3><p>Do not stop after opening a local export with the expected filename.</p><p>Follow the download link from the published page. Record the final web URL and the final file URL. If the file was replaced, open it again through the public link and confirm that the replacement is the version readers receive.</p><p>For a high-stakes announcement, the handover record can stay very small. It needs four things:</p><p>1. the page URL;<br>2. the download URL;<br>3. the approved revision or filename;<br>4. the date and time of the public check.</p><p>That record makes later questions easier to answer. It also prevents a familiar failure: everyone remembers approving the right file, but nobody can show which file was live at publication time.</p><h3>Keep search claims separate from publication checks</h3><p>Google lists both HTML and PDF among supported indexable file types. That means the formats can be processed by Google; it does not promise that a particular release will be indexed, rank well or appear in an AI-generated answer.</p><p>Publication QA should establish that the release is accurate, accessible and available at the intended URLs. Search performance is a later measurement problem. Mixing those two checks produces confident reports without evidence.</p><h3>A practical acceptance rule</h3><p>A release is ready when a reviewer can open every public version, confirm the same material facts, use the document as intended and identify the exact revision being served.</p><p>The export is only the start of that check. The public page and the public download are the evidence.</p><p>For a compact version of this workflow, see the WIRE.AD checklist: <a href="https://wire.ad/articles/html-press-release-pdf">https://wire.ad/articles/html-press-release-pdf</a></p><h4>References</h4><p>- Google Search Central, “File types indexable by Google”: <a href="https://developers.google.com/search/docs/crawling-indexing/indexable-file-types">https://developers.google.com/search/docs/crawling-indexing/indexable-file-types</a><br>- W3C, PDF7, “Performing OCR on a scanned PDF document to provide actual text”: <a href="https://www.w3.org/WAI/WCAG22/Techniques/pdf/PDF7">https://www.w3.org/WAI/WCAG22/Techniques/pdf/PDF7</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=107f4b6ca535" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to write a crypto launch announcement people can actually use]]></title>
            <link>https://medium.com/@wire.ad/how-to-write-a-crypto-launch-announcement-people-can-actually-use-27855fddc280?source=rss-5472296675bd------2</link>
            <guid isPermaLink="false">https://medium.com/p/27855fddc280</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[press-release]]></category>
            <category><![CDATA[crypto-marketing]]></category>
            <category><![CDATA[content-marketing]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <dc:creator><![CDATA[WIRE.AD]]></dc:creator>
            <pubDate>Wed, 07 Oct 2026 09:05:42 GMT</pubDate>
            <atom:updated>2026-10-07T09:05:42.931Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*86lb0_4GyTNJ6oazpLYD7A.png" /><figcaption>How to write a crypto launch announcement people can actually use by WIRE.AD</figcaption></figure><p>A practical guide to availability, evidence and preparing your news for distribution.</p><p>A developer reads your launch announcement and opens the documentation. A potential user looks for access. A journalist tries to identify what changed.</p><p>Your release should give each of them enough information to take the next step.</p><p>That starts with the first paragraph. What launched? Who is it for? What can they use today?</p><p>“An innovative platform for the next generation of Web3” leaves those questions open. A description of the product, its availability and its intended user gives readers something they can evaluate.</p><p>Here is how to build an announcement around those details.</p><h3>Start with the product and its current stage</h3><p>A launch can mean several different events: a public testnet, a private beta, an open waitlist or a product becoming generally available.</p><p>State the stage explicitly. Readers should be able to distinguish between access that exists now and work scheduled for later.</p><p>For a wallet, useful details could include supported networks, available countries and whether access requires an invitation.</p><p>For a developer tool, explain what developers can build with it, where the documentation lives and which capabilities are ready to use.</p><p>Consider this hypothetical opening:</p><p>“Acme Chain has opened a public testnet for developers building payment applications. Documentation and test tokens are available today. Mainnet is planned for December.”</p><p>The paragraph identifies the product, the audience and the stage. It also gives the reader a clear boundary between the announcement and the roadmap.</p><h3>Explain what changes for the user</h3><p>A feature name rarely explains the full benefit or the conditions attached to it.</p><p>If you announce cross-chain transfers, specify which networks are supported. If you introduce account recovery, explain the recovery method and where users can read about its requirements.</p><p>You do not need to put the entire technical specification in the release. Include enough detail for someone to understand the capability, then link to the relevant documentation.</p><p>A useful question during editing is:</p><p>“What can the intended user do after this launch that they could not do before?”</p><p>If the answer involves a future release, make that timing clear. An announcement about work in progress can still be useful when it describes the scope accurately.</p><h3>Give every number a definition</h3><p>“10,000 users” could refer to registered accounts, wallets connected, monthly active users or participants in a testnet.</p><p>Those measures describe different activity.</p><p>If a number belongs in the announcement, include its definition, timeframe and source. For example, explain whether a transaction figure covers a day, a month or the entire test period.</p><p>Apply the same care to performance claims. A benchmark needs enough context for readers to understand the conditions: the environment, the method and any relevant limitations.</p><p>If you cannot explain how a number was produced, resolve that gap before distributing it.</p><h3>Describe partnerships by their scope</h3><p>A partner logo can attract attention, but the announcement still needs to explain the relationship.</p><p>Is the partner providing infrastructure? Funding development? Integrating a product? Supporting a marketing campaign?</p><p>Name the work and distinguish the agreement from its delivery.</p><p>For example:</p><p>“Acme Chain and Wallet B are building an integration that will let Wallet B users access Acme Chain applications. Testing is scheduled for November.”</p><p>This wording describes planned work. It does not imply that access is already live.</p><p>Have both teams approve the description before publishing. Check names, responsibilities, dates and any claims about what the relationship makes possible.</p><h3>Prepare the destinations before sending the release</h3><p>Read the announcement as someone who has never seen the product.</p><p>Open every link. Check whether the documentation is public, the access page explains restrictions and the launch date agrees with the release.</p><p>If the product requires a waitlist, the destination should say so. If the release describes a testnet, the linked page should not leave readers thinking they have opened the mainnet product.</p><p>Prepare the supporting assets at the same time:</p><ul><li>A product screenshot that shows the feature being announced.</li><li>Documentation or a release note.</li><li>An access guide, where needed.</li><li>A named contact for questions and corrections.</li></ul><p>Label sample screens and illustrative examples. They should help explain the product without being mistaken for customer results.</p><h3>Adapt the material to the audience</h3><p>The announcement provides a shared set of facts. Different channels may need different ways into those facts.</p><p>A technical publication may need architecture details and documentation. A community post may need a concise explanation of availability. A creator preparing a walkthrough may need a working demo and permission to use the supplied assets.</p><p>Keep the product stage, dates and central claims consistent across versions. Adapt the explanation to the audience.</p><p>Before choosing distribution, ask who the news is useful to and what that audience needs to understand it.</p><h3>Run one final comprehension check</h3><p>Give the draft to someone outside the launch team. Ask them to explain:</p><ul><li>What launched.</li><li>Who can use it.</li><li>What is available today.</li><li>What remains planned.</li><li>Where they can check the claims.</li></ul><p>Compare their answers with the product’s actual state.</p><p>If they describe a live integration that is still being built, fix the wording. If they cannot identify the intended user, make the audience explicit. If they cannot find access, improve the destination.</p><p>Then distribute an announcement whose details survive being read, quoted and shared beyond your own channels.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=27855fddc280" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>