gshc2020.com

The 41-year-old striker, a regular in Portugal's starting line-up since 2003, did not play a minute of Sunday's win over Norway and did not train with his teammates on Monday or Tuesday. The Portuguese Football Federation (FPF) has not reported any injury or physical issue with the player.

Ronaldo's future in Portugal's international team has increasingly become a key talking point in recent years.

• Got Sky? Watch now on the Sky Sports app 📱
• Not got Sky? Get instant access with no contract📺

An automatic starter for two decades, the five-time Ballon d'Or winner was dropped from the starting line-up by then-head coach Fernando Santos for Portugal's 2022 World Cup knockout matches against Switzerland and Morocco. Roberto Martinez, Santos's successor, reinstated Ronaldo and selected him from the start for all 10 of Portugal's matches across the 2024 European Championship and this summer's World Cup finals.

While Ronaldo's long-term rival Lionel Messi, 39, will make his final Argentina appearance next week against Benin, there is no sign that the Al Nassr forward has plans to call time on his Portugal career.

Last month, he said he would "probably" retire at the end of the club season next June but before this international break he insisted he could still help the national side.

"I'm a thinker. You have to think," Ronaldo said when asked if he had considered retiring after this year's World Cup. "This is why you see me here now. It means I will continue.

"I had a conversation with the (FPF) president, with our coach. I still believe that I'm able to help the national team to reach what they want and will finish when I feel like I'm not doing anything here."

Ronaldo had seven shots without scoring in Portugal's 1-0 victory over Wales before being substituted off in the 65th minute for Goncalo Ramos. "I thought it was important to have fresh legs," Jesus said after the game. "Someone who could press the Wales center-backs and stop them playing out, someone with more speed in the first line of defence."

Twitter

This content is provided by Twitter, which may be using cookies and other technologies. To show you this content, we need your permission to use cookies. You can use the buttons below to amend your preferences to enable Twitter cookies or to allow those cookies just once. You can change your settings at any time via the Privacy Options.

Unfortunately we have been unable to verify if you have consented to Twitter cookies. To view this content you can use the button below to allow Twitter cookies for this session only.

For the win over Norway three days later, Ronaldo was an unused substitute. "Whether we like it or not, Cris isn't like Joao Neves, who is 20 and can play every three days," Jesus explained.

The head coach had said before that match that he planned for Ronaldo to participate, but that he was unlikely to feature from the start. When the final whistle went in Oslo, Ronaldo made his way directly from the substitute's bench and down the tunnel, not joining his team-mates in acknowledging the visiting Portuguese fans.

After the game, Jesus said it "wasn't the right moment for Cris to come on" and the state of the match had meant he changed his mind.

Sources, speaking on the condition of anonymity to allow them to do so candidly, have told The Athletic that Ronaldo was irritated by not featuring at all in the game. It was just the third time the forward was an unused substitute for Portugal; following a Euro 2008 group game (when Portugal had already qualified) and a friendly match against Croatia ahead of Euro 2024.

Jesus and Ronaldo have shared a strong professional relationship. Prior to his appointment this summer, Jesus coached Ronaldo for a year at Al Nassr and won the Saudi Pro League. The 72-year-old said his decision to coach in Saudi Arabia was due to Ronaldo's invitation and his willingness to coach the striker.

However, Portuguese sports newspaper Record described the situation between the two as an "internal earthquake". On Tuesday, Ronaldo shared an image of himself in national training gear alongside a heart emoji with the Portugal flag.

Please use Chrome browser for a more accessible video player

On Wednesday, sports newspaper A Bola led their Wednesday front page with the headline "situacao de crise" (crisis situation, translated from Portuguese). Portuguese publication B24 additionally said an "emergency" meeting had been scheduled between Ronaldo, Jesus, and FPF president Pedro Proenca.

The aim of this meeting was to clear the air. The FPF have indicated this was not a meeting, but rather a direct conversation between Jesus and Ronaldo, who share a strong mutual respect.

Publicly, Jesus has said Ronaldo "is an important figure for Portugal, playing or not". At the start of this international break, Ronaldo suggested to media that some in Portugal no longer wanted him to be part of the squad. "I feel the Portuguese fans like me, but I do understand that some TV channels want to criticise me," he said. "They've tried to kill me, but only with a shotgun. They are dying for me not to come to the national team anymore."

No men's footballer has made as many international appearances (234) or scored as many goals (146) for their nation as Ronaldo. The striker has featured in 12 major tournaments for his nation, winning the European Championship in 2016 and the Nations League title in 2019 and 2024.

He is unquestionably an icon for Portugal, but the debate on his future and participation with the national team continues to dominate.

This article originally appeared in The Athletic.

© 2026 The Athletic Media Company %!s()

It resulted in Dame Esther setting up Childline, a national telephone helpline for children suffering abuse or neglect.

She later became a renowned campaigner for older people and, in the last years of her life, for the legalisation of medically-assisted death.

Esther Louise Rantzen was born in Berkhamsted, Hertfordshire on 22 June 1940. Her paternal ancestors were Polish Jews.

Her father, Henry, who worked as a broadcast engineer for the BBC, had moved the family out of London to escape German bombing.

The family returned to London when the war ended. She later recalled she hated the grey streets of the capital after the semi-rural idyll in which she had spent her early childhood.

After a spell at a private school, run by what she recalled was a sadistic headmistress, she went to the North London Collegiate School in Edgware, and then to Somerville College, Oxford, where she read English.

The relationship caused huge tensions in her department, not least because Wilcox's wife Patsy, one of Rantzen's best friends, also worked at the BBC.

By 1968, she was working on Braden's Week, a consumer programme fronted by Canadian presenter Bernard Braden.

A decision to allow the researchers a short spot on air gave Rantzen her first on-screen TV experience, where she made a favourable impression on the producers.

In 1972, Braden was sacked by the BBC after he appeared in a commercial for Stork margarine and his show was dropped.

A year later, the programme was resurrected, given a new title, That's Life! and Rantzen was selected as the main presenter.

Braden was sanguine about Rantzen's promotion but his wife, TV presenter Barbara Kelly, was publicly contemptuous of the decision and accused Wilcox of being instrumental in her appointment.

Launched in May 1973, That's Life! became an almost instant hit.

Recorded just before transmission on Saturday nights, it was initially a consumer protection programme investigating viewers' complaints about shoddy service and faulty goods.

Over the years it grew to include more quirky items.

It initiated the Jobsworth award for over-zealous bureaucrats and council officials, and Heap of the Week, in which failing domestic appliances would be destroyed on screen in a variety of ingenious ways.

Over the years, Rantzen was joined by a series of co-hosts, including Paul Heiney, Chris Searle and Bob Wellings, while the comedian Cyril Fletcher read his Odd Odes.

Rantzen also went on to make an appearance on the ITV show I'm a Celebrity, Get Me Out of Here in 2008.

She wrote a number of regular columns for newspapers and magazines and also involved herself with more than 50 charities.

Following the MPs' expenses scandal in 2009, Rantzen announced she would stand for parliament in Luton South.

She pitched herself against the sitting Labour MP, Margaret Moran, who had claimed expenses in excess of £23,000 for her second home in Southampton.

Moran stood down two days later but Rantzen fought the seat in the 2010 General Election, coming fourth with just over 4% of the vote.

Rantzen was made an OBE in 1991 and CBE in 2006. In the 2015 New Year Honours list, she was appointed a Dame Commander in recognition of the two helpline services she set up.

The following year she warned of the social media pressure on children in the digital world, urging families to "be aware" of the issue.

Speaking on Desert Island Discs in 2019, Dame Esther told presenter Lauren Laverne that her own broadcasting career would not be as successful if she had joined the industry now as a young woman, because she would not be "pretty enough".

In an interview with The Mirror in June 2023, she spoke about her diagnosis with stage-four lung cancer.

It "made me realise how very lucky I've been in my life," she said, "working with Childline and the Silver Line, and meeting so many fascinating and inspiring people".

"I'm not good at regrets. What I treasure most are the fantastic friendships I have made thanks to That's Life! during the last 50 years, the people I met, and the team who worked so hard, and laughed so hard, together for so long."

Six months later, she told The Times she did not expect to live much longer.

"Last January," she explained, "I assumed that I would probably not be around for my birthday in June, and that I'd definitely miss this Christmas."

"But when I met my oncologist I realised how ignorant I was and what gargantuan strides medicine has made recently."

"Has my diagnosis taught me any lessons valuable enough to pass on? The cliché is live each day as if it's your last and treasure every moment."

In December 2023, Dame Esther said that she was taking a "miracle" treatment for her lung cancer, adding that she might "buzz off to Zurich" - where assisted dying is legal - if it did not work.

A year later, she witnessed the House of Commons vote in favour of the first stage of a Bill that might make a medically-assisted death legal in England and Wales. A similar bill was introduced to the Scottish Parliament.

Although Dame Esther knew that she would not live long enough to take advantage of any new law, she said she wanted to protect other families from "the same agonising experience".

"In the olden days, doctors would ease you out of life and the difference was it wasn't necessarily with your consent," she said.

In March 2025, her daughter confirmed that the drug that had kept Dame Esther alive since her cancer diagnosis was no longer working. They had always known that its effects would be temporary.

Last December, she told a newspaper that she was planning "an 'official' Christmas with my children and five grandchildren, slightly ahead of the real Christmas so that there will be more chance that I am actually alive to enjoy it with them!"

In September 2026, the House of Commons voted against the Assisted Dying Bill, a decision Dame Esther described as "a tragic day".

She added that she was now too frail to travel to Switzerland to end her life.

"Everything I didn't want to happen I'm having to go through now," she said, "which means that the people I love and care about either cannot visit me, or when they do, see me suffering pain or indignity which will inevitably get between them and our shared happy memories, and may even overwhelm the way they think of me."

The broadcaster and campaigner was once asked about what drove her to get involved in a multitude of causes.

"I need to be needed," she said. "It's what makes you get up in the morning, and makes you feel when you go to bed that the day has been well spent.

"There's a real pleasure in feeling somebody needs your help and that you actually have made a difference." %!s()

Salah Sarsour, a US permanent resident for nearly three decades, was detained by US Immigration and Customs Enforcement in March. He was released after a federal judge found he had probably been detained in retaliation of his advocacy for Palestinian rights, which the federal judge said would render his detention unlawful.

Despite that ruling, the administration pressed forward and fast-tracked Sarsour’s immigration case – a separate legal track – and on Tuesday night an immigration judge ruled he is “removable”.

Judge Jayme Salinardi, who was brought in from out of state to oversee the case, announced he would be retiring shortly after issuing the decision.

In a statement, Sarsour’s attorneys called the ruling “a setback for free speech rights”.

“The case we presented compels only one conclusion, which is that Salah is not deportable, he has done nothing wrong,” Patrick Taurel, one of Sarsour’s attorneys, said during a press conference shortly after the hearing. “He should be given back his green card with an apology for what’s been done to him.”

The Department of Homeland Security did not immediately respond to a request for comment, but claimed before that Sarsour was “suspected of funding terror organizations and lying on immigration forms”. He has denied both claims.

In a statement issued shortly after his arrest, the DHS cited Sarsour’s decades-old military court convictions in Israel for, it said, “throwing a Molotov cocktail at the homes of Israeli armed forces and illegally attempting to possess weapons and ammunition”.

Sarsour said those accusations stemmed from an arrest as a teenager by Israeli soldiers in 1989, during the first intifada. He also says he was forced, under torture and threat of rape, to sign documents in Hebrew without understanding them.

He ultimately spent two years in an Israeli prison after being convicted in a military court system that has a conviction rate of over 96% for Palestinians. Sarsour was again detained by Israeli authorities for nearly three months when he last visited Palestine in 1995.

His lawyers detailed those facts over four immigration court hearings, during which they noted that the government had for years known about Sarsour’s arrests in Israel.

In addition to his involvement with several local organizations, Sarsour is a board member of American Muslims for Palestine (AMP), a national advocacy group that Trump administration officials have accused of countering US foreign policy interests.

“They’re retaliating against him for speaking out about Palestinian human rights,” said Luna Droubi, another attorney. “That should terrify all of us.”

“It’s a principle of this country that people should feel that they can speak freely, and that’s being eradicated”.

In a statement, Sarsour, who has six children and nine grandchildren, all US citizens, called the decision “deeply disappointing” but vowed to “continue fighting”.

“We have a mission: let’s continue to make our voice clear, to protect the first amendment, to support justice,” he told the Guardian in July. “I’ll continue with that message whether I’m inside or outside.” %!s()

Search behavior now spans traditional search results and AI-generated answers. Buyers use ChatGPT, Gemini, Perplexity, Google AI Mode, and other answer engines for conversational research, so the right tool stack depends on the problem you need to solve — baseline visibility, ongoing monitoring, crawl diagnostics, first-party platform reporting, or content execution.

This guide compares those jobs, the tools that fit each, and the measurable outcomes to consider when deciding what is worth paying for.

How I evaluated: I work with enterprise and scaling brands on AI search visibility, and the recommendations here combine that experience with current product documentation and published research from Ahrefs, Vercel, Microsoft, Google, HubSpot, and other primary sources.

Table of Contents

• What is AI search optimization, and why does it matter?
• How AI Search Optimization Differs From SEO
• AI Search Optimization Tools Landscape and Jobs to Be Done
• Make your site accessible to AI systems.
• Structure content for citations with schema and direct answers.
• Optimize content differently for ChatGPT, Gemini, Perplexity, and AI Overviews.
• Track mentions, citations, and sentiment with AI search optimization tools.
• Choose AI-powered SEO software with an outcomes-first framework.
• Run a one-week pilot, and prove impact.
• Avoid common pitfalls with AI search optimization tools.
• Frequently Asked Questions About AI Search Optimization Tools
• Start with what you can measure, then fix what you can see.

What is AI search optimization, and why does it matter?

AI search optimization is the practice of making a brand and its content more likely to be mentioned, cited, and accurately represented in AI-generated answers.

You’ll see related work referred to as answer engine optimization (AEO), generative engine optimization (GEO), and AI SEO. The labels differ, but the goal is similar: improve how answer engines understand, represent, and cite your brand. When a brand launches an AI search strategy, the goals are:

• Citations. Clickable links to your pages in an AI-generated answer.
• Mentions. References to your brand in answer text, even when no link appears.
• Extractable answers. Clear, self-contained passages an answer engine can understand and quote accurately.

HubSpot’s State of AEO 2026 found that 44% of marketers have made a business purchase based on brands they discovered through answer engines.

AI-referred traffic remains an emerging acquisition channel, but early datasets indicate high intent. Similarweb’s 2025 ecommerce analysis estimated that ChatGPT-referred visits converted at 11.4%, compared with 5.3% for organic search.

How AI Search Optimization Differs From SEO

AI search optimization complements traditional SEO; it does not replace it. SEO still supports crawlability, indexing, rankings, clicks, and organic traffic. AI-search measurement adds a second question: When an answer engine builds a response, does it mention your brand, cite your content, and describe you accurately? That changes how you evaluate tools.

• Measure visibility, not just rank. Track citation frequency, share of voice against a defined competitor set, sentiment, and description accuracy alongside conventional search metrics.
• Expect output variation. AI answers can change with prompt wording, context, and the system used. Traditional rank position therefore does not map one-to-one to answer-engine visibility. Understanding semantic search helps explain why meaning and context matter more than an exact keyword match.
• Track cited sources, not only owned pages. Answer engines can cite publishers, communities, review sites, videos, and other third-party domains. A citation report should show which sources repeatedly appear for your priority prompts.
• Use SEO and AEO together. Clear definitions, crawlable text, strong evidence, and useful structure can support traditional search while also making content easier for answer engines to interpret.

AI Search Optimization Tools Landscape and Jobs to Be Done

Plenty of AI search optimization tools are on the market, but the right choice depends on the job you need the tool to do.

Job 1: Tell me where I stand right now.

Before you buy recurring monitoring or change a page, establish a baseline for what AI search engines are already saying about you. This is a one-time diagnostic that shows how answer engines currently represent your brand.

• HubSpot’s AI Search Grader is free and requires no account. It returns a composite score out of 100 across sentiment, presence quality, brand recognition, share of voice, and market competition using ChatGPT, Perplexity, and Gemini.
• Ahrefs Brand Radar adds paid market-level research across multiple AI platforms when you need broader competitive context.

Best for: Establishing an AI-visibility baseline and building the internal case for recurring measurement.

Job 2: Track my visibility over time.

Recurring monitoring is useful when you need comparable visibility trends across a stable prompt set and competitor set.

Current self-service prices span a wide range. Otterly.AI starts at $29/month, while HubSpot AEO costs $50/month standalone or $45/month when paid annually. Peec AI and Profound start at higher levels, while AthenaHQ offers a free entry tier. Engine coverage, prompt capacity, and reporting depth change by plan, so compare the tier you would actually buy rather than the vendor’s maximum advertised coverage.

The goal of this job is consistent measurement: run the same priority prompts repeatedly, track mentions and citations against the same set of competitors, and monitor how visibility and answer accuracy change over time.

Job 3: Show me what the engines see when they fetch my site.

Teams need to determine whether AI crawlers can reach their pages, use server logs, Cloudflare AI Crawl Control, or another crawler-diagnostics workflow. These sources can show whether GPTBot, ClaudeBot, PerplexityBot, and OAI-SearchBot are reaching your site and whether their requests succeed.

Crawler diagnostics are less glamorous than a visibility dashboard, but they can isolate access problems before you spend time optimizing content.

Job 4: Tell me what the platforms themselves report.

Third-party AI visibility tools estimate how answer engines represent your brand from the outside. Microsoft and Google now also provide first-party AI-search signals, but they measure different things.

• Bing Webmaster Tools AI Performance: Microsoft launched the report in public preview on Feb. 10, 2026. It reports citations across Microsoft AI experiences, cited pages, citation trends, and sampled grounding queries. Microsoft added Intents, Topics, Citation Share, and Compare in preview in June 2026.
• Google Search Console’s Search Generative AI performance reports: Google announced dedicated generative-AI impression reporting on June 3, 2026, including views by page, country, device, and date for participating sites.

Pro tip: Bing’s grounding queries and Google’s generative-AI impressions answer different questions. For conventional Google query analysis, use Google Search Console. Keep these datasets separate instead of treating them as interchangeable measures.

Job 5: Do the work.

Getting a baseline tells teams where their brand stands, and recurring tracking shows marketers how visibility changes. But measurement alone does not write content, refresh pages, or resolve the problems behind the data. For this job, content operations platforms that connect visibility signals to execution are useful.

• Writesonic combines AI-driven search visibility tracking with content creation, site audits, SEO tools, and AI workflows to help teams act on visibility gaps.
• AirOps combines AI-search visibility tracking with content creation and refresh workflows, Brand Kits, knowledge bases, and CMS, SEO, AEO, social, and project integrations.

In my work, I still see teams monitor visibility in one tool and carry the findings into Google Docs, a CMS, and a project board to do the actual work.

Best for: Teams that already know what they need to change and want to reduce the handoff between visibility data and execution.

Make your site accessible to AI systems.

Vercel and MERJ’s crawler analysis found that the major AI crawlers it tested did not execute JavaScript. GPTBot fetched JavaScript files in 11.5% of requests and ClaudeBot in 23.84%, but neither executed them. Google’s Gemini was a notable exception because it uses Googlebot’s rendering infrastructure.

That means critical content rendered only in the browser can be unavailable to crawlers such as GPTBot, ClaudeBot, and PerplexityBot even when Google can render the page successfully.

Tool to use: Cloudflare AI Crawl Control

Source

Why I like it: Cloudflare AI Crawl Control is available on all Cloudflare plans and shows which AI services access a site, plus crawler activity and request patterns. It also lets you set crawler-specific access policies and monitor compliance with robots.txt.

If AI crawler traffic is near zero on pages you want answer engines to access, investigate your crawl rules, security settings, and response behavior first. Healthy crawler traffic rules out one class of access problem, but it doesn’t prove that every system can retrieve, interpret, or cite the content correctly.

Pro tip: If you already use Cloudflare, AI Crawl Control gives you a cleaner interface for this check. Otherwise, server logs filtered by the relevant user agents can answer many of the same access questions.

How to Make Your Site Readable to AI Crawlers

Open a key page and view its source rather than relying only on the DevTools Elements panel, which shows the DOM after JavaScript runs. Search the source for a sentence from your main body copy. If the text is present in the initial HTML, you have cleared an important basic accessibility check for crawlers that do not execute JavaScript. Also check these blockers:

• robots.txt rules. Vercel and MERJ found that robots.txt directives were effective for the crawlers measured in their study, so an accidental disallow can remove content you meant to expose.
• Bot management and WAF rules. Ensure security rules don’t block or flag AI crawlers you intend to allow.
• Gated or interstitial content. Avoid putting content you want cited behind access barriers that a crawler cannot clear.
• Failed or slow responses: Check crawler requests for errors, timeouts, and redirects that keep the intended content from being retrieved.

Does llms.txt help with AI search visibility?

As of August 2026, the evidence does not support treating llms.txt as a Google Search visibility lever. Google says llms.txt files aren’t required for its generative AI features and don’t affect Search visibility either positively or negatively.

An Ahrefs study of roughly 137,000 sites found that 97% of valid llms.txt files received no requests in May 2026. The files that did receive requests were more likely to be accessed by coding agents and other agentic systems than by search-oriented retrieval bots.

If another service you use has a documented reason for consuming llms.txt, maintaining one may still be useful. I don’t consider publishing the file itself an AEO initiative.

Structure content for citations with schema and direct answers.

In May 2026, Ahrefs published a controlled study on schema and AI citations. The researchers identified 1,885 pages that added JSON-LD between August 2025 and March 2026, matched them against roughly 4,000 control pages from other domains with similar prior citation levels, and compared changes across Google AI Overviews, AI Mode, and ChatGPT.

The results were +2.4% in AI Mode and +2.2% in ChatGPT, both statistically indistinguishable from noise, and −4.6% in AI Overviews, a small but statistically significant decline. The authors cautioned that they could not confidently attribute that decline to schema.

In the Ahrefs study’s broader analysis of 6 million URLs, AI-cited pages were almost three times more likely to contain JSON-LD. That correlation matters, but it does not show that adding schema causes AI citations. Two caveats are especially important:

1. The controlled study looked at pages that were already receiving citations. In that sample, adding schema did not produce a measurable citation lift.
2. Google stopped showing FAQ rich results on May 7, 2026, and later removed the feature documentation. FAQ markup added solely to win that Google rich result no longer has that rationale. You can still explore other SERP feature opportunities.

What I’d do with schema: Use Organization, Article, Product, Review, and other relevant structured data when it accurately describes the visible content on the page. Structured data gives Google explicit information about a page and can enable eligible rich results. I would not sell schema as a direct AI-citation lever based on the evidence we have today.

Tool to use: Google’s Rich Results Test (free)

Source

Why I like it: The Rich Results Test shows which Google-supported rich results your structured data can generate and catches many technical errors and warnings. It does not validate every schema.org type, catch every quality issue, or tell you whether schema is earning AI citations.

What I Prioritize for Extractable Answers

Schema aside, these are the content patterns I prioritize when I want an answer engine to interpret a passage cleanly:

• Visible text that does not depend on client-side rendering.
• Answer-first paragraphs under relevant headings.
• Plain definitions before elaboration, with explicit tradeoffs in comparison copy.
• Specific numbers with sources and dates attached.
• Headings that closely match the question the reader is trying to answer.

Here’s how I check whether a paragraph stands on its own: Could someone paste it into a chat window as a complete answer with nothing added? If it needs three sentences of surrounding context to make sense, I rewrite it.

Optimize content differently for ChatGPT, Gemini, Perplexity, and AI Overviews.

ChatGPT Search, Google’s AI features, Perplexity, and Microsoft’s AI experiences don’t retrieve or present web information in the same way. A brand can therefore be represented differently across platforms, which makes engine coverage an important part of tool selection. Here are the platform differences I would actually use when evaluating a tool.

• ChatGPT: ChatGPT Search can use third-party search providers as well as content supplied directly by publisher partners. Do not assume its source set or results will mirror Google.
• Google AI Overviews and AI Mode: Both draw on Google’s Search index and core ranking systems, and both can use query fan-out to find supporting pages. Google says the two features may use different models and techniques, so the responses and links they surface can differ.
• Perplexity: Perplexity searches the live web and builds responses with citations and links to sources, making source-level monitoring especially useful.
• Microsoft AI experiences: Bing Webmaster Tools AI Performance provides first-party citation reporting across Microsoft AI experiences, which gives you a platform-native check against third-party monitoring data.

Tool to use: HubSpot AEO ($50/month standalone or $45/month when paid annually; also included with Marketing Hub Professional and Enterprise).

Why I like it: The prompt-level view shows the actual answer returned by ChatGPT, Gemini, or Perplexity, rather than reducing everything to a single visibility score. Prompts can be filtered by engine, buyer-journey phase, and product or service relevance. At the same time, citation analysis shows which domains and content types are driving mentions for your brand and your competitors.

For Marketing Hub Professional and Enterprise customers, connected CRM context makes AEO recommendations more specific, and HubSpot’s content tools give teams a way to act on those recommendations. That fits the broader HubSpot stack: HubSpot Smart CRM provides a unified system of record for customer information, while Content Hub provides tools for content creation and management.

Track mentions, citations, and sentiment with AI search optimization tools.

Manual spot-checking is a legitimate way to start an AI search strategy, but it is difficult for ongoing reporting because screenshots don’t provide a consistent trend line. Once you move past manual checks, a repeatable measurement setup needs:

1. A fixed prompt set. Use prompts that reflect real buying questions across problem-aware, solution-aware, and vendor-comparison stages. I usually start with roughly 25 to 50. Keeping a stable core matters because replacing prompts constantly makes period-over-period comparisons much less useful.
2. A defined competitor set. Share of voice only becomes meaningful once you have decided which competitors to include in the comparison.
3. Citation-level source parsing. Look for tools that identify the domains and pages behind an answer rather than merely detecting whether your brand appeared.
4. Sentiment and accuracy tracking. Negative mentions or inaccurate descriptions are a different problem than absence.
5. First-party corroboration. Where platform-native reporting exists, use it to check on third-party monitoring.

One expectation to set with leadership early: these datasets do not always reconcile one-to-one. Bing describes grounding queries as sampled, and a query-oriented view answers a different question from a page-oriented citation view. That distinction is also useful context when thinking about semantic search. Report trends rather than treating a single-day number as ground truth.

Tool to use: Bing Webmaster Tools AI Performance (free).

Source

Why I like it: Bing Webmaster Tools AI Performance is first-party data, not an outside estimate. Microsoft launched the report in February 2026 and added Intents, Topics, Citation Share, and Compare in preview in June.

The report includes total citations across Microsoft AI experiences, cited pages, citation trends, and sampled grounding queries — the key phrases Microsoft’s systems used to retrieve content referenced in AI-generated answers. Those grounding queries can help you understand which phrases are most often associated with your site being cited in AI answers.

Choose AI-powered SEO software with an outcomes-first framework.

When choosing AI-powered SEO software, it is easy to get lost in the features. Evaluate tools against the outcomes you need, not the longest feature list. Here is the weighted scorecard I use.

To get more out of a demo, ask two questions:

• Who owns this on Monday morning? AI visibility work can cross content, PR, product marketing, and web engineering. If nobody owns the dashboard after purchase, it is unlikely to change much.
• What decision changes based on this data? If the honest answer is “we would know our score,” you are buying monitoring rather than execution. That can still be useful, but price and staff it accordingly.

Pro tip: Bring 10 of your own prompts to every demo. Vendor sample prompts may not reflect your actual buying questions. Running the same prompts across competing tools exposes differences in coverage and results, which helps you decide how much confidence to place in any single visibility score.

Run a one-week pilot and prove impact.

You do not need a quarter to determine whether an AI search optimization tool fits your workflow. I would use five working days to establish a baseline, compare tool output with what I can observe manually, and decide whether the platform earns a place in the stack.

Day 1: Baseline

• Run a free diagnostic such as AI Search Grader.
• Pull Bing Webmaster Tools AI Performance and Search Console Search Generative AI performance reports if you have access.
• Capture your starting data.
• Write down your 10 highest-value buyer questions.

Day 2: Technical Check

• View the source of your five most important commercial pages and confirm that critical body copy is present in the initial HTML.
• Check robots.txt for the AI user agents you intend to allow.
• Pull server logs or Cloudflare AI Crawl Control data and confirm that GPTBot, ClaudeBot, PerplexityBot, and OAI-SearchBot are reaching the pages you care about.
• Resolve obvious access problems before judging content or visibility recommendations.

Day 3: Manual Cross-Check

• Run your 10 prompts manually across ChatGPT, Gemini, Perplexity, and Copilot.
• Record which brands are named, which sources are cited, and how accurately your brand is described.

This is tedious, but it gives you a reference set to evaluate the reports from a paid monitoring tool.

Day 4: Two-Tool Trial

• Load the same 10 prompts into both.
• Compare the tools with each other and with your manual reference set.
• When a result differs meaningfully from what you observed, ask the vendor how its sampling, model access, location, personalization controls, and refresh cadence work.

Day 5: Decision and Selection

Choose the tool that best fits the decision you need to make and the workflow you can sustain. Then commit to one change:

• Rewrite the top of your three highest-value pages so the answer comes first.
• Fix a rendering or crawl-access problem.
• Pursue visibility on a third-party source that repeatedly appears in competitor citations.

Re-measure in 30 days against the Day 1 baseline.

In my experience, visible changes are more realistic on a weeks-not-days timeline because the underlying systems refresh on their own schedules. Description accuracy is often one of the first signals I watch before expecting citation volume to change.

Avoid common pitfalls with AI search optimization tools.

As an AI search visibility consultant, these are the mistakes I watch for most often:

• Buying monitoring before checking access. A visibility dashboard cannot tell you much if the crawlers or search systems behind the answers cannot reliably retrieve the pages you want represented.
• Stacking overlapping tools. Two monitoring platforms can produce different visibility numbers because their prompts, models, locations, sampling, and scoring methods differ. Pick one primary monitoring source and, where available, use platform-native data as a check.
• Treating the visibility score as the KPI. Visibility scores are vendor-specific composites. They are useful for tracking direction inside the same methodology, but prompt-level answers, citations, sentiment, and description accuracy are usually more actionable.
• Changing the prompt set every month. You lose comparability. Add prompts around a stable core rather than repeatedly replacing the core.
• Optimizing only owned content. If your citation reports repeatedly surface review sites, communities, trade publications, videos, or other third-party properties, the next action may be PR, partnerships, community participation, or distribution rather than another blog post.
• Expecting attribution to look like SEO attribution. Mentions without links produce no referral sessions. A click-only measurement framework, therefore, misses part of what these tools measure.
• Assuming what worked in January still works in August. Google’s FAQ-rich-result deprecation and the 2026 schema research both changed recommendations within months. Re-test platform-specific assumptions regularly.

Frequently Asked Questions About AI Search Optimization Tools

Do AI search optimization tools replace traditional SEO tools?

No. AI search optimization tools add a measurement layer for brand visibility, citations, share of voice, sentiment, and answer accuracy. They do not replace the technical SEO, crawlability, indexing, content quality, and authority work that still supports traditional search and many AI-powered search experiences.

How do I measure ROI from AI search optimization tools?

Measure ROI at three levels. First, track visibility signals such as citation frequency, share of voice, sentiment, and description accuracy against a baseline. Second, track referral sessions from AI sources in GA4; those capture clicked citations but not unlinked mentions. Third, connect AI visibility to pipeline signals such as self-reported attribution, CRM data, and branded-search movement where available.

Expect incomplete attribution. An unlinked mention can influence a buyer without creating a referral session, so no current measurement framework captures every AI-assisted touchpoint.

Which AI platforms should I prioritize first?

Prioritize the platforms your buyers actually use rather than relying on a single universal ranking. Start with your own referral data, customer research, sales feedback, and existing visibility.

Also distinguish what each platform can report directly. Google Search Console provides first-party impression data for participating sites in Google’s generative AI features, while Bing Webmaster Tools provides citation reporting for Microsoft AI experiences. For platforms without comparable webmaster reporting, you will usually need third-party monitoring or a controlled manual prompt set.

Can a small team benefit from AI for SEO tools?

Yes. A small team can benefit from AI-powered SEO tools by starting with a narrow task rather than buying a large stack. Use a free diagnostic to establish a baseline, define a small set of high-value prompts, and add recurring monitoring only when you need trend data you cannot manage manually.

For a smaller budget, a self-service monitoring tool may be enough. What matters most is whether someone on the team will use the data to make a specific content, technical, PR, or distribution decision.

Start with what you can measure, then fix what you can see.

AI search optimization tools are most useful after you know which problem you are trying to solve. Verify that the systems you care about can access your content, establish a baseline for how your brand is represented, choose a stable set of prompts, and use the resulting data to make a specific content, technical, PR, or distribution decision.

The tools and answer engines will keep changing, so treat your measurement framework as something you revisit rather than a permanent scorecard.

If you are still at the “where do we even stand?” stage, start with the free AI Search Grader. If you need ongoing monitoring, HubSpot AEO tracks 25 prompts daily across ChatGPT, Gemini, and Perplexity for $50/month or $45/month when paid annually, and is included with Marketing Hub Professional and Enterprise.

Recommended Stories

• list 1 of 3Northern Ireland protesters block Orange Order march in Portadown
• list 2 of 3Northern Ireland protesters block Orange Order march into Catholic area
• list 3 of 3Protesters block Orange Order march in Northern Ireland

Some pitched tents. Others brought food and stayed through the night, keeping watch.

“We’ll be here as long as we have to be,” said one local woman, who asked not to be identified for fear of violent reprisals. “I’ve seen the fighting when those Orangemen come along here, and I’m not having it back here – none of us are.”

The Orange Order, a Protestant organisation with a historically anti-Catholic tradition, has been banned from marching on Garvaghy Road for the last 28 years.

The High Court order has overturned that ban. Now, a small group of Orangemen gathers at Drumcree Church each day in their traditional regalia, ready to march.

But they have not been able to get beyond the end of the road. Police stand in their way, refusing them permission to pass and citing the risk of violence with so many protesters waiting a few hundred metres away.

This has infuriated the Orange Order, which has called for the Police Service of Northern Ireland chief constable, Jon Boutcher, to be removed. The group believe they have a legal right to march and accuse police of yielding to the protesters.

“We hold a positive determination issued by the Parades Commission and reinforced by the High Court judge,” Nigel Dawson, district master of the Portadown Orange Lodge, told Al Jazeera. “The chief constable must comply with the law.”

In the 1990s, there were protests and violence when the Orange Order tried to return from Drumcree Parish Church along the road.

When asked how much longer the Orangemen are prepared to wait, Dawson said, “I’ve been at Drumcree for 28 years.”

Dawson warned of a dangerous precedent in which “might makes right”.

If a protest is big enough, the police can refuse to uphold a court ruling, he said, adding, “We have the law on our side here.”

Ultan Larkin, a protester who travelled to Portadown from the neighbouring town of Armagh to show support for the Garvaghy Road residents, said, “The law is on their side, but justice is not on their side.

“There’s hurt here. There’s pain here. Just because you can doesn’t mean you should.”

The standoff has so far been peaceful. But history tells the story of why this confrontation has a real risk of escalation.

‘Justice is not on their side’

Garvaghy Road was one of the most potent flashpoints of The Troubles – a 30-year conflict between mostly Protestant unionists who fought for Northern Ireland to remain part of the UK, and predominantly Catholic republicans who want a united Ireland.

Orange marches through the Catholic area were usually accompanied by some of the worst sectarian violence Northern Ireland has seen.

The peace deal known as the Good Friday Agreement was signed in 1998, and it transformed Northern Ireland.

Large-scale violence largely ended, former enemies shared power in a devolved assembly at Stormont, and a generation grew up in an era of peace without soldiers manning checkpoints on their streets.

Following this peace process, the Orange Order was prevented from parading along Garvaghy Road – an effective ban that has lasted almost three decades, until now.

In an unexpected decision, the Northern Ireland Parades Commission allowed for a scaled-down Orange march to go ahead under strict conditions. Only 35 Orangemen could take part in the march, and they would have to do so in silence without their usual flutes and drums.

Residents of Garvaghy Road were appalled and legally challenged this ruling.

A High Court judge initially granted an injunction preventing the parade, but the Court of Appeal overturned that ruling late on Saturday, clearing the legal way for the march to proceed on Sunday morning.

A further High Court challenge was subsequently dismissed.

Larkin grew up during the Troubles. He was convicted for having a senior role in the Real IRA and was sentenced to four years in prison – a conviction quashed years later on appeal.

“We have to live together, and to do that we have to respect each other,” he said. “But where is the respect if you’re marching through a community that doesn’t want you and where the people of that community are asking you to march somewhere else?”

“On the Irish tricolour we have green, white, and orange. The green is for the Irish Catholics here, and the orange is for the Protestant community, with white symbolising peace. We all have our place.”

He said there are old wounds on Garvaghy Road.

“People here suffered a lot when the Orange marches came. They remember how bad things were here – the intimidation and that sort of thing – and some people just aren’t able to forgive that.”

Political fallout

The standoff has become a political crisis in Northern Ireland, affecting the power-sharing government that runs the country.

Sinn Fein First Minister Michelle O’Neill visited Garvaghy Road to stand beside the protesters, alongside Sinn Fein President Mary Lou McDonald.

The Democratic Unionist Party (DUP) accuses nationalist politicians of breaking the law by supporting a protest that blocked Garvaghy Road, and observers have questioned how the first minister can continue in her role when she has openly rejected a High Court ruling.

The confrontation is particularly awkward as the DUP and Sinn Fein are supposed to govern together, representing the two principal political traditions that emerged from the conflict as part of the peace process.

The language between them has become increasingly bitter, and in less than nine months, Northern Ireland faces an Assembly election.

For now, the DUP and Sinn Fein remain in government together, while accusing the other of threatening the political settlement that followed the Good Friday Agreement.

Meanwhile, attempts to find a resolution continue.

Northern Ireland Secretary Chris Bryant has met representatives of the two communities, but talks have so far failed to produce a breakthrough.

His future is also now on the line over this.

Bryant could revoke the Orange Order’s right to march, but he has not exercised this power. Unionists have threatened to call for his resignation if he tries. Irish nationalists say they will do likewise if this parade goes ahead. %!s()