Your Gemini and Google Business Profile checklist

The post Should You Let AI Manage Your Google Business Profile? appeared first on Koozai.com

Polished opening. Short paragraphs. Generic leadership lesson. Neat conclusion.

AI has made it much easier to produce content at scale, but LinkedIn is now taking steps to reduce the amount of low-value, repetitive material appearing in users’ feeds.

The platform refers to this kind of content as “AI slop”, and brands that rely heavily on LinkedIn should pay attention.

LinkedIn Isn’t Banning AI Content, but it is Targeting Low-Value Posts

LinkedIn isn’t banning AI-generated content.

Instead, it’s focusing on content that appears generic, repetitive or lacking a clear point of view. It has also said it wants to reduce automated comments, engagement bait and activity designed to artificially boost reach.

LinkedIn has introduced systems to identify generic AI content and says posts without meaningful insight may be less likely to reach people outside the author’s immediate network.

It has also added a “Seems like AI slop” feedback option, allowing users to flag posts and comments they think fall into this category.

The message is fairly clear: publishing more content isn’t necessarily going to help if that content doesn’t offer anything useful.

What LinkedIn’s AI Crackdown Means for Brands

For brands, the real risk isn’t simply using AI. It’s publishing content that feels generic, interchangeable or disconnected from genuine expertise.

AI can still be useful for organising ideas, editing copy, summarising interviews or turning longer content into social posts.

The problem comes when AI starts replacing the thinking behind the content.

If your LinkedIn strategy relies on prompts such as “write five thought-leadership posts about digital marketing”, you’re likely to end up with content that could have been published by almost anyone.

So what should brands do differently?

Give People Something Only Your Brand Can Say

The strongest LinkedIn content usually starts with something specific.

That might be:

• A trend your team is seeing in client work
• An interesting campaign result
• A common customer question
• An opinion from one of your specialists
• Original data or research
• A lesson from something that didn’t go to plan

AI can help turn that information into a polished post, but the insight should come from your business.

Ask yourself a simple question before publishing:

Could one of our competitors post this word for word?

If the answer is yes, it probably needs more personality or substance.

Turn Your In-House Expertise Into Content

One of the easiest ways to avoid generic content is to speak to the people doing the work.

Your specialists already have useful opinions, examples and experiences. You just need a process for getting those ideas out of their heads and into your content.

That could mean a quick interview, a few questions over Teams or a conversation after a client meeting.

You can then use AI to help structure or edit the material.

The important bit is the order.

Human expertise first. AI assistance second.

Focus on Content Quality Rather Than Posting More Often

Generative AI makes it possible to produce a lot more content.

That doesn’t mean you should.

Publishing 15 generic posts is unlikely to do more for your brand than publishing three genuinely useful ones.

LinkedIn’s changes reinforce something marketers should already know: posting frequency is only useful if the content deserves attention.

Rather than asking, “Are we posting enough?”, ask: “Are we giving people a reason to stop scrolling?”

Make Sure Your Comments Add Something Too

LinkedIn is also targeting automated and low-value comments.

Responses such as “Great insights, thanks for sharing” don’t add much to a conversation, whether they were written by AI or not.

Good comments add an opinion, ask a relevant question or share another useful example.

Treat engagement as a conversation, not another box to tick on the content calendar.

How Should Brands Use AI for LinkedIn Content?

The answer isn’t to stop using AI completely.

The smarter approach is to use it where it genuinely helps while keeping people responsible for the ideas, expertise and opinions behind the content.

AI can be useful for:

• Improving a first draft
• Repurposing existing content
• Organising expert interviews
• Suggesting different headlines
• Shortening or clarifying copy

But don’t expect it to create original expertise from nothing.

LinkedIn’s crackdown is less about whether AI touched a piece of content and more about whether that content gives the reader anything worthwhile.

Human Expertise Still Gives Brands an Advantage on LinkedIn

AI has made creating content easier.

Creating content people actually want to read is still the hard part.

For brands, that means prioritising expertise, experience and clear opinions over sheer publishing volume.

Use AI as a tool, not as your entire content strategy.

Because if your LinkedIn posts could have come from any brand in your industry, there’s a good chance your audience will treat them that way too.

Want to make your social content work harder without adding to the noise? Get in touch with Koozai to talk through your social media and content marketing strategy.

The post LinkedIn Is Cracking Down on AI. What Does That Mean for Brands? appeared first on Koozai.com

In March 2026, Rightmove became the first UK property portal to launch a dedicated app inside ChatGPT. Type “@Rightmove” followed by a plain English description of what you want, and it will search live listings for you, inside the chat window, without you ever visiting rightmove.co.uk.

That is quite the move [forgive the pun]. The UK’s largest property portal has just decided that meeting buyers inside an AI assistant matters enough to build for it directly. If Rightmove is showing up where the conversation is happening, the question for every estate agent, letting agent, surveyor and developer is simple: are you showing up there too?

It starts with a question, not a name

Somebody wondering whether now is a good time to sell in their town. A landlord trying to work out what new regulation actually means for them. A family quietly doing the sums on whether they can afford to move. None of them start by typing your business name into anything. They start with a question that feels personal, and increasingly, that question goes to ChatGPT, Google’s AI Overviews, Gemini or Perplexity before it goes anywhere near a search box.

Property is not formally classed by Google as a “Your Money or Your Life” topic in the way health and personal finance are. But let’s be honest about what buying, selling or letting a property actually is: for most people, it is one of the largest financial decisions they will ever make. Search engines and AI systems still lean on the same underlying question when deciding who to trust: is this experienced, is this trustworthy, is this safe. That is the heart of what is known as E-E-A-T, and it applies here just as much as it does anywhere formally labelled high stakes.

Property is not one journey. It is several, wearing the same website

This is the bit most property marketing gets wrong. A residential buyer, a portfolio landlord, a commercial tenant and a solicitor instructing a chartered surveyor are not the same person on different days. They ask different questions, check different sources, and are convinced by entirely different evidence.

• A buyer checks Rightmove, Zoopla and Google reviews, and is swayed by recent, specific feedback and a jargon-free fee structure.
• A landlord checks HM Land Registry sold price data and Propertymark guidance, and wants proof you actually understand current legislation, not just that you say you do.
• A commercial occupier checks RICS guidance and sector press, and cares about a named MRICS or FRICS surveyor with a track record on assets like theirs.
• A solicitor, lender or asset manager checks the RICS Regulated Firm register and published case studies, and barely glances at your star rating.

Treat all four like the first one, and you will only ever be found by the first one.

The moments that decide whether you get found at all

The question moment

Most journeys now begin with a natural language question to an AI tool, one that pulls its answer from HM Land Registry data, RICS and Propertymark guidance, and structured data on property websites, not from a Google ranking. Why Nobody Googles You First in Property Anymore covers this properly.

The sense check moment

Once someone has a rough idea, they go looking for the unfiltered version, checking whether the reviews match what the AI told them, whether a firm’s accreditations actually stack up. Is This Agent Legit? walks through what that looks like for each of the four audiences above.

The comparison moment

This is where AI stops summarising and starts judging, weighing your reviews, your accreditations, and how recently your content was updated against someone else’s, and deciding who gets mentioned. Why AI Recommends Some Property Brands and Not Others explains what tips that in your favour.

The enquiry moment

By the time someone requests a valuation or a fee proposal, the real decision has already been made elsewhere. From Question to Enquiry walks through two full search paths, a buyer’s and a facilities manager’s, side by side.

Why this matters more than a nicer website

Property is inherently local. Nobody asks “which is the best estate agent in the country.” They ask about a specific postcode, village or street, and AI tools are increasingly answering those hyper-local questions directly, without sending anyone to a search engine at all.

If you are invisible across the sources these tools trust, the enquiry form on your website simply never gets used, however well designed it is. Our AI team works with estate agents, letting agents, surveyors and developers to close that gap, and you can read more about how we support the sector on our property and real estate marketing page.

Want the full unabridged version? Download the Guide to AI Search for Property, a whitepaper covering everything in this series.

The post The Real Property Search Journey: What Your Customers See Before They See You appeared first on Koozai.com

It is easier to understand all of this properly by watching two very different people go through it, step by step. Neither of them looks anything like the tidy funnel diagram most marketing plans are still built around.

Path one: a buyer looking near Manchester

Step one. They open ChatGPT and ask, “Is now a good time to buy a family home near Manchester?”

Step two. A Google AI Overview answers a follow up question about average house prices in a specific postcode.

Step three. They open Rightmove and Zoopla, browsing listings and checking the Zoopla Estimate against what they are being told elsewhere.

Step four. They check Google Business Profile and Trustpilot reviews for the two agents they have shortlisted.

Step five. A local Facebook group or Reddit thread gets searched for one specific, nagging question: is this area actually safe, actually good for families.

Step six. Only now do they land on an agent’s website, to request a valuation or book a viewing.

If that agent was not visible anywhere in steps one to five, step six simply never happens. Not because the agent was not good. Because the agent was never part of the conversation.

Path two: a facilities manager facing a dilapidations claim

This journey looks completely different, and takes place somewhere entirely separate from the one above, but the underlying lesson is identical.

Step one. They ask ChatGPT, “How do I challenge a dilapidations claim at lease expiry?”

Step two. They read RICS guidance and sector press for background on the process and likely costs.

Step three. They search LinkedIn and sector directories to shortlist two or three chartered surveying firms.

Step four. They check those firms’ own websites and published case studies for a track record on comparable assets, and confirm professional indemnity insurance cover.

Step five. They search a specific surveyor firm’s name alongside “reviews” or “case studies” on Google.

Step six. Only now do they land on a surveying firm’s website, to request a proposal or a fee quote.

Same pattern. Same lesson. Different audience entirely, different questions, different platforms, but if you are not visible somewhere in steps one to five, step six never happens, whichever of these two people, or the others like them, you are trying to reach.

The point of putting these two side by side

Property marketing plans are still too often built around a single imagined customer, usually a residential buyer. But a landlord, a commercial tenant, and a solicitor instructing a surveyor are all running their own version of this exact same search path, through entirely different sources, looking for entirely different proof.

Map your own content and case studies against each of these audiences properly, rather than assuming one “about us” page will do the job for all of them. Write the pages that actually answer the questions each of these people has, in the places they are actually asking them.

If you would like help mapping your content against these different audiences and closing the gaps, our AI Marketing team can help, and you can find out more about our work in the sector on our property and real estate marketing page.

This four part series, together with the overview, covers the full journey from the first AI assisted question through to enquiry, across every kind of property audience, not just one. For the full version, download the Guide to AI Search for Property.

The post From Question to Enquiry: Two Real Property Search Paths appeared first on Koozai.com

Most property brands still think of their competition in terms of who else is on the high street or who else bids for the same instructions. Increasingly, the more important competition is happening somewhere they cannot see: inside an AI assistant, being weighed against a rival they may never have heard of.

People are now asking AI directly to compare:

• “Is Agent A or Agent B better for selling a flat in this town?”
• “Which chartered surveyor should I use for a dilapidations schedule in Leeds?”
• “Is this developer trustworthy for off-plan investment?”
• “Best commercial property manager for a mixed use portfolio in this city?”
• “Which firm handles rent reviews for industrial units near this motorway junction?”
• This is the step most property brands underestimate entirely.

An AI comparison is not a search result. It is a verdict.

Ask Google to compare two agents and you get two sets of links, left for you to sort through. Ask an AI assistant the same question and it will actually weigh things up and tell you which one it thinks is better. That is a fundamentally different kind of exposure, and a fundamentally different kind of risk if your signals are weak.

Here is roughly what is being weighed up in the background:

Review sentiment. Transparency of fees. Professional accreditations, RICS, Propertymark, ARLA. Structured data on your website. Third party citations and press mentions. How recently your content was updated. How clearly you explain who your service is actually best for.

Weak signals across these do not mean you rank lower. They mean you may simply never be surfaced at all, however good your actual service is. There is no consolation prize for being excellent but invisible.

Practical moves that genuinely change this

Keep key service pages updated every quarter with new facts, fresh reviews and clearer wording. Outdated content undermines both trust and visibility at exactly the moment an AI tool is deciding whether to mention you.

Add a short “at a glance” box to core service pages covering fees, patch covered, typical timescales and accreditations. AI tools read information laid out this clearly with particular ease, and can lift it directly when summarising you to someone.

Make sure your structured data is genuinely clean. LocalBusiness and Organisation schema with professional body memberships included, FAQ schema, review schema, and RealEstateListing or product schema for listings, developments or services. Keep it factual. This is too high stakes a decision for marketing fluff, and machines read structured data completely literally.

The truth underneath all of this

Clarity beats cleverness here, every time. The property brands getting recommended are not the ones with the most persuasive copy. They are the ones that made it easiest for a machine to be confident repeating them, which usually just means being specific, current and honest about who you are and what you charge.

If you would like help getting your structured data, reviews and service pages ready for this kind of scrutiny, our AI team can help, and our property and real estate marketing page has more on how we work with the sector specifically.

Want the complete version of this? Download the Guide to AI Search for Property which covers everything in this series.

Next: what does this actually look like end to end? Read “From Question to Enquiry”

The post Why AI Recommends Some Property Brands and Not Others appeared first on Koozai.com

An AI assistant has just recommended you. That should be the end of the story, but for most people, it is only the beginning of a much more sceptical second act. The questions at this stage sound noticeably more pointed than the ones that came before:

“Is this agent legit?”

“Do the reviews match what the AI told me?”

“Is this area actually safe, good for families?”

“Are these fees normal?”

And the checking happens across a genuinely wide spread of places: Rightmove, Zoopla, Google Business Profile reviews, Trustpilot, allAgents, local Facebook groups and Reddit threads, anywhere that feels human and current rather than curated.

BrightLocal’s 2026 Local Consumer Review Survey found that 97 percent of consumers now read reviews before choosing a local business, and that the share of people using AI tools such as ChatGPT to discover and vet local businesses jumped from 6 percent to 45 percent in a single year, now the third most used discovery channel behind only Google and Facebook. That is a US panel, but the pattern applies directly to how UK buyers and sellers are checking out agents and surveyors today.

This looks different depending on who is checking

For a residential buyer or seller, this is largely about recent, specific reviews and a fee structure that does not need decoding.

For surveying, commercial agency and property management, the checking process is just as thorough but looks for different things entirely. Buyers of these professional services want to know whether a surveying firm is RICS Regulated and carries adequate professional indemnity insurance. They look for case studies or a track record with similar assets, industrial, retail, listed buildings, mixed use blocks. They check how a managing agent handles service charge transparency and Section 20 consultation. And they look for evidence that a commercial agent has genuinely closed deals in that sub-sector recently, not just listed them.

Why journalism carries particular weight here

Muck Rack’s Generative Pulse research, based on more than a million AI cited links, found that earned media and journalism make up around a quarter of all links AI models cite, rising to roughly half when a query implies the need for recent information. A house price or rental market question is exactly that kind of query. “Is this still true today” is baked into almost everything people ask about property.

What actually earns trust at this stage

• Show your professional memberships clearly and consistently. RICS, Propertymark, ARLA, NAEA, wherever they apply, not buried in a footer.
• Keep review responses current, and reply properly to the difficult ones. A calm, specific explanation when something has gone wrong signals accountability, which matters more here than a spotless record ever could.
• For commercial and institutional audiences, replace the star rating instinct with a named, citable case study or client testimonial. “Advised on X rent reviews across a Y square foot portfolio” does more work for this audience than any number of five star reviews.

None of this needs to be flawless. It needs to hold up to a second look, because that second look is exactly what is happening before anyone ever picks up the phone. Our AI Marketing team can help you build the kind of review and accreditation presence that survives it.

Next: once the sense check is done, how does AI actually decide who gets mentioned? Read “Why AI Recommends Some Property Brands and Not Others”

The post “Is This Agent Legit?” How Buyers, Sellers and Landlords Check You Out appeared first on Koozai.com

Rightmove now has an app inside ChatGPT. Say that sentence to most property professionals a year ago and it would have sounded like science fiction. Today it is simply how a growing number of buyers are starting their search, typing “@Rightmove, three bed semi near good schools in Guildford” into a chat window rather than opening a browser tab at all.

That single fact tells you where the industry’s biggest player thinks attention is heading. This post is about what that means for everyone else.

The questions have not changed. Where they go has.

People are still asking exactly the questions they have always asked:

“Is now a good time to buy in Bristol?”

“What survey do I need for a 1930s semi?”

“What’s the average rental yield in Leeds?”

“How long does conveyancing usually take?”

What has changed is where those questions land first. Increasingly, it is an AI assistant, not a search engine, and certainly not a specific agent’s website.

Early UK data at this granularity is still catching up, but the direction of travel is clear enough elsewhere. In the US, NerdWallet’s 2026 Home Buyer Report found that 48 percent of people planning to buy a home in the next year say they have used, or plan to use, AI tools somewhere in the process, and a Realtor.com survey put overall AI usage for housing market information at 82 percent of Americans. Rightmove building directly into ChatGPT is the UK industry’s own signal that this shift is not a passing trend.

Where these tools are actually pulling their answers from

This is the part worth understanding properly, because it is genuinely different from how a Google ranking works. Tools like ChatGPT, Gemini and Perplexity draw on a fairly specific pool of sources:

HM Land Registry and Gov.uk data. Financial advice articles. Digital PR coverage. Comparison websites. Review platforms. Structured data sitting on your own website. Expert commentary. RICS and Propertymark guidance. Planning Portal information.

Notice how much of that list is official, professional body or government data, rather than marketing copy. That is a meaningful difference from a lot of other sectors. If your business is not referenced anywhere in that pool, an AI tool has nothing to work with when your name might otherwise be relevant, and you simply will not be considered later in the journey.

What actually helps here

• Make your content feel like you are explaining options to a friend who wants clarity, not a sales pitch. AI tools tend to reproduce that tone when they summarise you, which means the register you write in matters almost as much as the facts themselves.
• Write in plain English about the things people are actually confused by. Leasehold versus freehold. What a Section 21 notice actually means. How stamp duty bands work. This is where genuinely useful content earns its place in an AI answer, rather than a generic “why choose us” page.
• Update anything tied to legislation the moment it changes. The Renters’ Rights Act is a good live example of exactly the kind of compliance topic buyers and landlords are actively searching about right now, and stale content on a topic like that damages trust fast.

Getting this stage right is not about ranking for a keyword. It is about being one of the sources an AI system is confident enough to repeat when someone asks a question you could answer better than anyone. If you would like help making sure your content is set up to be found this way, our AI team can help.

Next: once someone has a rough idea, how do they check whether it actually holds up? Read “Is This Agent Legit?”

The post Why Nobody Googles You First in Property Anymore appeared first on Koozai.com

This time, we spoke with Jojo Furnival, Marketing Manager at Sitebulb, about why technical SEO fundamentals still matter in the age of AI search, how AI has changed the way she works and why brand building should be firmly on marketers’ agendas over the next 12 months.

Jojo has more than 15 years of experience across marketing, SEO and communications, having worked agency-side in London and Australia, as a freelancer and in-house in the charity sector. In 2023, she found herself, almost by accident, working in technical SEO.

As Sitebulb’s one-woman marketing function, she’s since built its marketing strategy from scratch, tripled its following and increased brand awareness in a key market from 46% to 65% in under two years. She’s also an active voice in the Women in Tech SEO community and a familiar face at industry conferences.

The Interview

For each Koozai Couch, we ask the same five questions to get a quick but useful snapshot of what people across digital marketing are seeing, using and rethinking.

Here’s what Jojo had to say.

1. What’s one thing the industry is getting completely wrong right now?

Getting caught up in GEO as the new shiny thing and treating it as a discipline that’s separate from SEO.

Remember, LLM bots can and do crawl indexed website content to seek information that helps them provide an answer to their users. This means the technical health, crawlability and indexability of your website are just as important as they’ve always been.

This falls under the remit of good old-fashioned technical SEO. And don’t you forget it!

2. What tool or resource has genuinely changed how you work lately?

Claude Cowork.

I can legitimately say that Cowork has transformed the way I repurpose content. With its help, I can turn one pillar piece of content into seven or more different formats and types of content with ease.

Manually, this pipeline would take me five days or more. Now, it can be done in a day.

As the one-woman marketing function at Sitebulb, I can’t tell you how valuable that is. Unfortunately, it doesn’t mean I can sit back and put my feet up. All the time I’ve saved has simply been filled with more responsibilities and projects!

3. What’s a take you’ve changed your mind on in the last 12 months?

AI workflows.

I’ve been working with AI for a while now, like most people, and the models have come a long way since the beginning.

For some time, I probably sat more in the “AI seems clever, but it’s actually the dumbest assistant you’ve ever had” camp.

However, since starting to use Claude Cowork and investing a lot of time in creating and continually improving genuinely smart Skills that work to my exact specifications, AI has become invaluable to my workflows.

4. What are you reading, watching or listening to that others in digital should know about?

I’m in a privileged position at Sitebulb because I’m talking to super-smart people in SEO and digital all the time.

They contribute articles to our site, which I edit. They speak at webinars that I organise. They put together training sessions for Sitebulb’s audience, and I work with them on the topics. I also attend conferences throughout the year, where I get the chance to engage with some incredible minds in our field.

But a lot of these people’s insights are accessible for free.

Anyone can watch Aleyda’s Crawling Mondays on YouTube. Anyone can sign up to receive Candour’s Core Updates newsletter. Anyone can join Good Signals’ Friday morning Office Hours. And any woman or person of a marginalised gender can join the Women in Tech SEO community.

So I heartily encourage anyone reading this to do exactly that because, alongside Sitebulb’s content, of course (!), this is where you can learn a huge amount from clever people who are actually doing the work.

5. Where do you think digital marketing is heading in the next 12 months?

BRAND BUILDING.

Sorry for the caps, but it deserves them!

SEOs have to adopt a brand marketing mindset. The modern SERP is a nightmare for organic results. The average searcher has to scroll basically to the fold before they see position one because of AI Overviews and paid search placements getting bigger. Thank you to Tom Capper for the data on that one.

And ever since featured snippets came out, we’ve been seeing drops in clicks.

So we know it’s not just about the click anymore. The way customers find and interact with brands has changed. For a growing number of queries, AI is providing the responses.

The question then becomes: how can our brand become visible in this environment?

There’s plenty of data showing that strong brands have an advantage, so I think brand building and digital PR are going to become much more important over the next 12 months.

Thank you, Jojo!

A huge thank you to Jojo for joining us on the Koozai Couch.

Her answers are a useful reminder that while AI is changing how marketers work and how people find information, the fundamentals haven’t disappeared. Technical SEO still matters, smart workflows can help small teams get more done, and building a recognisable brand is becoming increasingly important for visibility.

Catch up on previous Koozai Couch interviews

Enjoyed Jojo’s interview? There are more perspectives from across the digital marketing industry on the Koozai Couch.

Catch up with Sean Barber, SEO Manager at Macmillan Cancer Support and host of Search with Sean, as he shares his thoughts on AI search, SEO fundamentals and where the industry could be heading next.

You can also read our Koozai Couch interview with Erin Simmons, where Erin discusses trust, community and why doing less can sometimes support growth.

Fancy taking a seat on the Koozai Couch?

We’re always keen to hear different perspectives from people working across digital marketing.

If you’d like to feature in a future Koozai Couch interview, submit your answers via our Koozai Couch form. We’d love to hear what you think.

The post Koozai Couch: Interview with Jojo Furnival appeared first on Koozai.com

It is a Tuesday evening. The dishes are done, the laptop is shut, and somebody is lying on the sofa scrolling their phone with one thought running on a loop: we need a holiday. Not a specific one yet. Just a feeling. Somewhere warm, somewhere that will not wreck the budget, somewhere that does not require three evenings of research to find.

So they open an AI assistant and type something like: “Where’s warm in March for a couple’s holiday for under £800 each?”

Nobody in that moment is thinking about your brand. They are not thinking about SEO, funnels, or channels at all. They just want an answer that feels honest. And whether they realise it or not, that is the exact moment your relationship with them begins, quietly, somewhere you were probably not watching.

A global survey of 22,000 travellers by Skyscanner found that more than half are now comfortable using AI to help plan or book a trip in 2026. That number is moving fast, and it comes with a catch: the more people trust these tools, the less patience they have for anything that turns out to be wrong, outdated, or exaggerated. Travellers do not want to be sold to. They want to be told the truth, quickly.

Nobody researches a trip in a straight line

There used to be a tidy way of describing how people bought things: awareness, interest, desire, action, in that order. Travel planning has never actually worked like that, and Google now has a name for the real pattern: the messy middle. People bounce between apps, half-formed ideas, short videos, forum threads and reviews, circling a decision rather than marching towards it.

You will see a stranger’s TikTok of a beach, a friend’s Instagram story from a city break, a Reddit thread arguing about whether a resort is overrated, all before anyone types a single brand name into a search bar. If your visibility plan still assumes Google is the only place this happens, you are only seeing a fraction of the actual journey.

What a real 2026 trip search often looks like

Strip away the marketing language and a fairly typical path looks something like this:

An AI assistant for inspiration (“where’s warm in March”). Another AI tool to narrow it down (“compare 3-night city breaks for couples”). Google Maps for the unfiltered version, reviews, photos, local tips. A quick Reddit search to settle a nagging worry (“is Lisbon safe at night?”). Then, finally, a plain Google search for something very specific (“book hotel Lisbon centre”), and your website.

Notice how late your website appears in that list. Everything before it is where the decision actually gets made.

The four moments worth understanding

Dreaming out loud, to an AI

This is where the loose, aspirational questions live, and where AI assistants are pulling from travel articles, PR coverage, review platforms and destination guides to build an answer. How Travellers Are Using AI to Plan a Trip goes into this properly.

Checking the story holds up

The moment the dream needs a reality check: is this actually safe, do the photos match what people say, does the AI’s version match TripAdvisor’s. Is This Hotel Actually Good? covers what travellers are really doing here.

Letting AI make the call between two options

This is newer, and stranger, than most brands realise: people are now asking AI to decide for them between two operators or hotels. Why AI Recommends Some Travel Brands and Not Others explains what tips that decision one way or the other.

Handing over the money

By the time someone reaches a booking form, the emotional work is already finished. Making the Booking Easy covers what still manages to go wrong here anyway.

Why this matters more than it looks like it should

If a brand is invisible across the first three of those moments, the fourth simply never happens. A stunning booking page cannot rescue a trip that was never on someone’s radar in the first place.

Travellers are not looking for the loudest voice in the room. They are looking for the calmest, clearest path to an answer they can trust, and that path now runs through AI assistants and review sites long before it runs through your homepage. Our AI team helps travel brands build exactly that kind of presence.

Download the Guide to AI Search for Travel and Tourism, covering everything in this series.

The post The Real Travel Search Journey: What Your Customers See Before They See You appeared first on Koozai.com

Unfortunately, for anyone who has spent a lot of time perfecting a booking page, by the point someone actually lands on it, your persuading is mostly finished. They decided somewhere back in the AI conversation, the review scrolling, the comparison. What is left is not a sales moment, it’s admin.

You can hear it in the searches themselves.

“Book hotel Lisbon centre.” “Availability next weekend.” “Check in times.” “Cancellation policy.”

Nobody typing that is still weighing up whether to go. They are looking for the mechanics of making it happen.

None of what follows matters if a brand was invisible earlier in the journey. A gorgeous, frictionless booking page cannot rescue a trip nobody was considering in the first place. It just becomes a beautifully designed shopfront on a street nobody happens to be walking down.

Assuming the earlier stages have actually worked, though, this final stretch is exactly where a warm, ready-to-book lead can still slip away over something small.

Where it falls apart

A slow-loading page full of heavy imagery, right when someone has finally decided and just wants to confirm, is one of the most common and avoidable ways a booking gets lost. This is not the moment for more persuading. It is the moment for speed. Missing or buried details, what is actually included, what the cancellation policy says, what time check-in starts, send someone hunting through the site at the exact point they were ready to commit, which is precisely when you do not want to give them a reason to pause.

Silence after the click causes more anxiety than people admit. An instant confirmation email and a clear sense of what happens between now and the trip turns “I think I’ve just spent £800” into “I’ve just booked something I’m excited about.”

No way to talk to a human, for anyone still slightly unsure, especially on a longer or pricier trip, a visible phone number or live chat option earns more trust here than almost anywhere else on the site.

The one instinct worth resisting

It is tempting to use this last page to make one final case for why you are the best choice. Resist it. Nobody wants a pitch at the checkout. They want reassurance that the decision they already made was a good one, and then they want to get on with planning what to pack.

This stage is short, but it is the one people remember if it goes badly. A slow loading form or a buried cancellation policy right at the end of weeks of research can undo trust that took a long time to build. Treat it as a handshake, not a pitch.

This four-part series, together with the overview, covers the full journey from the first AI-assisted daydream through to booking. For the full, practical version, download the Guide to AI Search for Travel and Tourism. If you would like help putting any of this into practice, our AI Marketing Agency team can help.

The post Making the Booking Easy: What Travellers Need at the Final Step appeared first on Koozai.com

We’ve previously written an article on SEO audits but in this article, we’ll go through more specifically what a technical audit consists of, how that feeds into overall technical SEO, some of the tools that we use and maybe even a few little tips if you’re lucky. If you’re not an SEO, you can definitely pick up some tips and help gain an understanding but it’s definitely a task that’s better left to a professional SEO.

So what is a technical audit?

A technical audit is often done either as a standalone process or part of an overall SEO audit of a website. Like many things in SEO, what is reviewed in these audits varies from SEO to SEO as they may have particular expertise or interests in certain areas compared to others. Regardless of who is performing this, a typical audit breaks down various aspects of the website from crawlability and indexability, site architecture and data tracking to on-page aspects. The goal of this audit is to identify how do machines perceive your site, is the site healthy and stable from a coding aspect and does it offer a positive user experience for users.

What do I need to perform a technical audit?

Performing a technical audit is a fairly intermediate level task and therefore requires a good toolset to perform. There are a variety of tools out there both paid and free that can help you perform this but here’s a few we just can’t live without.

Crawlers

In order to audit a site, you need a crawling tool such as Sitebulb and Screaming frog. This will become a bit of a reoccurring theme throughout this section but the answer to “Which do I pick” is down to personal preference. Whilst Screaming frog is great for diving deeper into the granular data, Sitebulb is my personal preference because I believe its UI is much friendlier to view and digest with far more automation options. For example, using Sitebulb we have created automatic monthly crawls that will crawl our clients sites and update the findings to a shared area that the whole team can review at any time. Without a crawler, you simply won’t be able to gather the data you need to perform an in-depth audit across the site.

Google Search Console & Google Analytics 4

Before you kick off a crawl with your crawling tool, it’s always preferred that you have access to Google Search Console and Google Analytics 4. Not only can you connect these to your crawler, therefore providing more insight, but you can also manually review these yourself. You can find out if the website has Core Web Vital issues, and if so, for how long and is it a sudden spike that can be attributed to a developer change? You can also see if Google is struggling to connect to the site (therefore signaling an unstable server or server outage issue) by following steps:

1. Open up Google search console and select the correct property
2. Click on ‘Settings’ in the left hand side menu
3. Click on ‘open report’ in the Crawl stats area
4. Click on the correct host (You may have only one option in this area)
5. Click on the ‘host status’ bar and open up ‘server connectivity’

If you find that there are high or consistent levels of failed requests, this could be an indicator of a struggling server that needs to be raised with the developers.

Pagespeed Insights

Pagespeed Insights an amazing free tool that will quickly become your best friend. It allows you to quickly review the Core Web Vitals of an individual page alongside a Lighthouse test that highlights potential performance issues and opportunities to improve the user experience. One thing that’s important to remember is that the aim isn’t necessarily to chase a perfect score of 100 as a website can still perform perfectly well without reaching that magical green number. Instead, use the tool to identify genuine issues such as slow-loading images, content moving around whilst the page loads or elements taking too long to respond when a user interacts with them.

Javascript extension

Having an extension that allows you to disable javascript is a must have as it gives you a very simple way of understanding how reliant a website is on JavaScript and what happens when it isn’t available. Disable Javascript, refresh the page and have a look around then ask yourself “Does the main content disappear entirely” or “Does the navigation stop working” the goal here is to see if there is anything not working or not visible that you absolutely want a crawler to see.

Redirect Path

This is another extension that has become a personal favourite. Redirect path allows you to see the full chain of the page should you land on a redirect which helps identify any urls behaving in a way that they shouldn’t be alongside discovering potential chains and redirect loops.

Chrome DevTools

Chrome DevTools can look a little intimidating the first time you open it, particularly if you aren’t a developer, but you really don’t need to understand every section to get plenty of value from it during an audit. Right-clicking anywhere on a webpage and selecting inspect gives you access to the HTML behind the page alongside useful areas such as the Network and Console tabs. The Network tab is particularly handy because you can refresh the page and see all of the images, scripts, fonts and other files being requested by the browser alongside how long each one is taking to load and whether any have failed entirely. DevTools will take some time to get to grips with but your confidence will grow as you use it.

Schema Validator and Rich Results test

In order to see if the website is using structured data otherwise known as SCHEMA, we use both Schema.org’s Schema markup validator alongside Google’s Rich Results Test as they test for slightly different things. The Schema Markup Validator essentially reviews if structured data is actually present and if there’s any missing or incorrect details. Rich Results test is more focused on whether the structured data can be used by Google for featured snippets such as product information. Though this only tests one page at a time, it’s good to test a number of pages to understand if structured data is being used at all, and if there’s any trends such as no product pricing on all product templates.

So what is actually looked at in a technical audit?

With your shiny new toolset and your completed crawl, you’re likely going to be faced with a long list of scary looking errors and warnings which is where the auditing part comes in. It’s now the SEO’s job to manually review all of these to separate the genuine issues and opportunities from the false flags and low impact items. A crawling tool may tell you something is deeply wrong or isn’t present, but just like us mere humans, it tends to make mistakes. Here’s just a few things we recommend you look at.

Crawlability and indexability

One of the first things we want to establish is whether search engines can actually find and access the important pages on the website and, once they find them, whether those pages are able to be indexed. This means reviewing areas such as the robots.txt file, XML sitemap, internal links, canonical tags and noindex directives alongside the indexing information available within Google Search Console. You may find that an important page has accidentally been marked as noindex, that Google is being directed towards the wrong canonical URL or that thousands of low-value URLs are being generated through filters, parameters or other website functionality. Not every URL on a website needs to be indexed and in many cases Google choosing not to index things such as feeds, account pages or certain filtered URLs is perfectly normal. What we are trying to establish is whether the pages that actually matter are easy for search engines to discover and whether Google is spending its time crawling useful pages rather than getting lost in areas of the site that offer very little SEO value.

Broken pages and redirects

Broken pages are another area you will almost certainly come across during an audit, but it’s important not to fall into the trap of assuming that every 404 error automatically needs fixing. If a page has genuinely been removed and there is no suitable replacement, returning a 404 can be completely appropriate, but problems start to appear when important pages disappear without any thought being given to what should happen to the old URL. That page may still have internal links pointing towards it, external websites linking to it or users finding it through old bookmarks and search results, so simply removing it can create a poor experience and potentially throw away value that the URL has built over time. This is where redirects become important, but again they need to be used sensibly rather than sending every deleted page back to the homepage. Where a relevant replacement exists, redirect the user there, and whilst you’re doing that make sure you aren’t creating unnecessary redirect chains where a URL passes through several different pages before eventually reaching the correct destination.

Site architecture and internal linking

Site architecture is one of those areas where technical SEO starts to overlap quite heavily with user experience because a website that is easy for a search engine to navigate will usually be easier for a person to navigate as well. Important pages shouldn’t be buried away several layers deep with very few links pointing towards them, particularly if those pages represent the services, products or areas that are most valuable to the business. As part of the audit, review how the website is structured and how easily you can move between related areas, looking at navigation menus, breadcrumbs, category structures and contextual internal links within the content itself. Internal linking is particularly useful because it helps search engines discover pages and understand how different topics relate to each other whilst also giving users a natural route towards the next piece of information they might need. If you’ve written several useful articles around a particular service but none of those articles link back to the main service page, you’ve missed a fairly simple opportunity to bring those areas of the website together.

JavaScript and content accessibility

Earlier we talked about disabling JavaScript and this is where that check starts becoming particularly useful because it allows you to understand whether important parts of the website are being delivered in a reliable way. Modern websites rely on JavaScript for a huge range of things and there is absolutely nothing wrong with that, but if disabling it causes the main content, internal links, navigation or essential product information to disappear completely then it’s worth investigating how those elements are being delivered. Google can render JavaScript, but rendering adds an additional step compared to information being immediately available within the HTML, so we generally want the most important content and links to be as accessible as possible. This becomes even more important if the content relies on several different scripts or third-party services all successfully loading before it appears, because every extra dependency introduces another potential point of failure.

Site speed and Core Web Vitals

Site speed is another major part of most technical audits, although it’s important to look beyond simply reporting the score that PageSpeed Insights gives you. Core Web Vitals are designed to measure parts of the real user experience, such as how quickly the main content becomes visible, how responsive the page is when somebody tries to interact with it and whether elements unexpectedly move around whilst everything is loading. When you identify a problem, try to understand what is actually causing it rather than simply telling a developer that the website has failed Core Web Vitals, as that doesn’t give them a huge amount to work with. The issue could be an oversized hero image, too many scripts running before the main content appears, fonts causing the layout to shift or the server simply taking too long to respond. The more you can narrow down the likely cause and explain what is happening, the more useful your audit becomes to the person who eventually has to fix it.

Mobile usability

Mobile usability sounds like something that should be taken for granted nowadays, but you’d be surprised how many issues can still slip through unnoticed until somebody actually starts using the website properly on a smaller screen. Rather than simply resizing your browser and deciding that everything looks alright, spend some time navigating the site as a real user would by opening menus, clicking buttons, completing forms, using filters, viewing products and trying any other functionality that is important to the site. Sometimes an element technically works on mobile but is frustrating to actually use, such as buttons being too close together, filters covering the screen or content shifting around as the page loads. Google primarily uses the mobile version of the site for indexing, but even putting SEO aside, if a large proportion of users are visiting from mobile devices then a poor experience can have a very real impact on enquiries, purchases and overall performance.

Structured data

Structured data is another area worth reviewing during an audit because it provides search engines with clearer information about what the content on a page represents. Depending on the website, you may come across Organisation, Product, Article, Breadcrumb, LocalBusiness and a whole range of other schema types, and part of the audit is making sure that what has been implemented is accurate and actually reflects what users can see on the page. Ecommerce websites are a good example because product schema can quickly become more complicated once you introduce different variations, prices, stock levels and reviews, so it’s worth checking that the information being supplied to search engines matches the actual product. Schema isn’t about covering every page in as many different types as possible, it’s about helping machines understand the content and entities already present on the site in a clearer and more structured way.

Tracking and analytics

Tracking is one of those areas that can easily be forgotten during a technical audit because it doesn’t always feel like traditional technical SEO, but there isn’t much point improving a website if you can’t reliably measure whether those improvements are actually doing anything. Review the GA4 and Google Tag Manager setup and make sure the important actions on the website are being recorded correctly, whether that is purchases, enquiries, form submissions, downloads, phone calls or any other action that matters to the business. Look for obvious issues such as duplicate events, missing conversions or revenue figures that don’t seem to match what the business is actually seeing. If you’re going to use analytics data to help prioritise recommendations or report on the impact of your work later on, you need to have confidence that the information you’re basing those decisions on is accurate in the first place.

Don’t blindly follow your crawler

This is probably one of the most important lessons when performing a technical audit because a crawler is there to provide information, not make the final decision for you. You might finish a crawl and be presented with hundreds of warnings ranging from missing meta descriptions through to canonical issues and broken internal links, but those problems aren’t necessarily equal just because they appear in the same report. A missing meta description on an old article that receives virtually no traffic is unlikely to be as important as a canonical issue affecting every product on an ecommerce site, even though both technically need attention. When you review each issue, think about how many pages it affects, whether those pages are commercially important, whether it prevents crawling or indexing, what impact it has on users and how difficult it is likely to be to fix. A good audit doesn’t simply tell somebody everything that is wrong with their website, it helps them understand what needs fixing first and why.

Use the website yourself

One of the easiest mistakes to make when performing an audit is spending all of your time staring at crawling tools, spreadsheets and dashboards without actually using the website yourself. Spend some time browsing around like a potential customer would, use the navigation, search for things, apply filters, complete forms, view the site on mobile and click through important journeys from start to finish. Crawlers are excellent at identifying technical patterns but they don’t always tell you that applying a filter sends a user to a completely different section of the website, that an important button disappears at a certain screen size or that a form technically loads but is incredibly frustrating to complete. Some of the most useful issues you will find during an audit often come from simply spending time on the website and asking yourself whether everything behaves in the way you would expect it to.

Compare what your crawler finds with what Google finds

Another useful exercise is comparing the URLs your crawling tool discovers with what Google Search Console tells you Google has actually found because the numbers can sometimes be very different. If your crawler finds 5,000 URLs but Google has discovered 30,000, that immediately gives you something to investigate because those additional URLs have to be coming from somewhere. You may uncover historic URLs, filters, feeds, parameters, old site sections or functionality that is generating large numbers of unnecessary pages that aren’t obvious when following the normal internal linking structure. The opposite can also be useful, as you may have pages that your crawler can easily access but Google appears to be ignoring or rarely crawling, which could point towards internal linking, crawl priority or indexing issues. Looking at the same website through several different sources of data often gives you a much clearer picture than relying on a single tool.

Speak to the people who manage the website

Finally, don’t forget that technical SEO rarely happens in isolation and there are usually developers, marketing teams or internal staff who understand parts of the website far better than you do when you first arrive. If you find something unusual, speak to them before immediately deciding it is wrong because there may be a perfectly sensible reason why something has been implemented in a particular way, or they may already be aware of the issue and have a fix planned. Developers can also tell you whether a recommendation is relatively simple or whether changing it would require a significant rebuild, which can completely change how you prioritise the work. The best technical SEO work usually comes from collaboration rather than throwing a huge audit over the fence and expecting somebody else to work through it, so involve the people who actually build and manage the site and make sure your recommendations are realistic, understandable and tied back to what the website is trying to achieve.

So, is that everything?

Not even close. Technical SEO can go incredibly deep and depending on the website you may end up looking at everything from server logs and international targeting through to faceted navigation, ecommerce platforms, migrations, rendering and huge enterprise websites containing millions of URLs. The important thing when you’re getting started is not trying to learn everything at once, but understanding the fundamentals and becoming comfortable asking why something is happening rather than simply accepting what a tool tells you. Every audit you perform will introduce you to something slightly different and, more often than not, you’ll come across an issue you haven’t seen before which sends you down another technical SEO rabbit hole. That’s part of what makes technical SEO interesting, although you may not feel quite as enthusiastic about that when you’re three hours into trying to work out why Google has discovered 400,000 URLs on a website that only has 5,000 actual pages.

The post SEO Unpacked: What really is a Technical audit and how do you perform one? appeared first on Koozai.com

Somewhere in the middle of planning, the question changes shape. It stops being “what are my options” and turns into something closer to asking a well-informed friend to just decide for you.

“Is Brand A better than Brand B for couples?” “Compare these two tours.” “Best value hotels near Lake Garda?” “Is this operator trustworthy?”

That last shift matters more than it seems. Earlier in the journey, people were gathering. Here, they are outsourcing a judgement call to a machine, and the machine is going to give them one, whether or not you had any say in the outcome.

What is actually happening when AI “decides”

Unlike a search engine, which hands back ten blue links and lets the person sort through them, an AI assistant asked to compare two operators genuinely weighs things up and gives an opinion. It is not neutral. It is combining signals, quietly, in the background, before it ever produces a sentence.

Here is roughly what it is grading you on:

• Review sentiment. Not just star ratings, but the tone and pattern of what people are actually saying.
• Pricing visibility. Whether the cost is genuinely clear, or something you have to dig for.
• Brand mentions. How often, and how consistently, you turn up across the wider web, not just your own site.
• Structured data. Whether your site is actually machine-readable, or just human-readable.
• Third-party citations. Whether anyone independent has vouched for you.
• Recency. Whether your content looks like it was written last month or three years ago.
• Safety information. Whether you have actually addressed the “is this safe” question, rather than assumed it away.

If these signals are strong, you get named, sometimes recommended outright. If they are thin or inconsistent, the AI simply leaves you out of the comparison altogether, which is a much quieter and more total kind of rejection than losing a click. You never even know the conversation happened.

What tends to actually move the needle

Tidy structured data does a surprising amount of work here: FAQ schema, tour or product schema, local business schema, and accurate opening hours, pricing bands and locations all give an AI tool something unambiguous to work with, rather than something it has to guess at.

Keeping information current matters more in travel than in most sectors, because so much of it is genuinely time-sensitive: seasonal availability, local prices, opening hours. Stale content does not just look neglected, it actively erodes trust.

Being specific about who a trip is actually for beats being impressive. “This tour suits confident hikers looking for a two-day route” gives an AI tool something concrete to match against a real question. Vague, aspirational copy, the kind most travel marketing defaults to, gives it nothing to grab onto.

Clarity outperforms cleverness here, consistently. The brands that get recommended are rarely the ones with the best adjectives. They are the ones that made it easiest for a machine to be confident repeating them.

If you would like help getting your structured data, reviews and content ready for this stage, our AI Marketing Agency team can help.

Download the Guide to AI Search for Travel and Tourism for the complete version of everything in this series.

Next: by the time someone’s ready to book, what still manages to go wrong? Read “Making the Booking Easy”

The post Why AI Recommends Some Travel Brands and Not Others appeared first on Koozai.com