Searchlight is Aztek's marketing news roundup that brings together the week’s most relevant developments in marketing, search, AI, and digital strategy, all in one place. We update this article throughout the week with news we think is worth your time, along with context to help you understand what changed, why it matters, and what it could mean for your business.
This week's topics:
04/27: Reviews as Training Data: How Local Businesses Can Stand Out in AI-Generated Search Results
When someone searches on Google, AI Overviews can now pull together a quick summary of the options that seem most relevant to their question. Those summaries are built from information across the web, and customer reviews can play a meaningful role in what gets surfaced.
That means a strong review is doing more than helping a person feel confident about your business. It can also give Google’s AI more evidence to understand who you are, what you do well, and why you may deserve to show up when someone is looking for a business like yours.
Why Reviews as Training Data Change Local SEO
Generative search is looking at more than your website pages. It’s trying to understand your business as an entity, including what you offer, what people say about you, and how clearly those signals match what someone is searching for.
That’s where reviews can make a real difference. If customers repeatedly mention something specific, like “gluten-free wedding cakes,” that gives AI search tools stronger context than a basic business profile that only says “cakes.” In other words, reviews don’t just influence whether someone trusts you after they find you. They can also shape whether you show up, how you show up, and what gets said about your business when you do.
The Review Signals AI Overviews Parse First
Three factors rise to the top when models weigh review data.
- Volume: A steady flow of fresh feedback tells the model the business is active and relevant.
- Specificity: Reviews that contain product, service, or location detail give the model text it can quote.
- Recency: Newer reviews carry more weight than dusty praise from two years ago.
Secondary signals matter too. Author credibility (long-time Local Guides, for example) and owner responses both feed the ranking math.
How to Encourage Reviews AI Can Quote
Generic five-star ratings help, but rich language helps more. Ask customers to mention the problem you solved, the service they chose, and the city or neighborhood. Email follow-ups work well when they feel personal. A short prompt can read:
“Thanks for trusting Bright Smile Dental. A sentence about the procedure and how it went helps neighbors choose a dentist they can count on.”
Point-of-sale QR codes and texting platforms fill gaps when email open rates sag. The goal is usable detail, not scripted fluff.
Optimize Your Google Business Profile for Review Visibility
A complete profile gives AI a fact sheet to match against review claims. Take ten minutes to verify categories, services, hours, and high-resolution photos. Populate the Q&A section with common questions and clear answers. When the model sees the same service phrase in the profile and in multiple reviews it gains confidence and lifts the business above competitors that send mixed signals.
Respond to Reviews Like a Human, Train the Model
Owner replies create fresh, structured text that models crawl. A good response thanks the customer, repeats the service keyword once, and adds a next-step cue. Poor replies rely on canned “Thank you for your feedback” lines that provide no learning. Keep the voice real and concise. A quick checklist helps:
- Address the reviewer by name
- Reference the specific service
- Note how the feedback will be used
- Invite the reviewer back
These steps show expertise and attentiveness, qualities language models flag as trust signals.
Measure Success When Clicks Shrink
AI overviews often hand users a phone number or directions link inside the answer. Session counts may fall even while leads rise. Shift the scorecard to metrics that reflect this reality:
- Calls and direction requests from the Google Business Profile dashboard
- Booking or order actions that fire from direct integrations
- UTM-tagged visits from AI surfaces to confirm downstream behavior
Impressions only tell half the story. Focus on actions that actually book revenue.
Practical Next Steps
- Audit current reviews: Look for volume gaps, dated feedback, and missing service keywords.
- Refresh request flows: Add prompts that invite detail and rotate them quarterly.
- Tighten profile data: Align categories and services with the language customers use.
- Template owner responses: Build a three-sentence framework staff can personalize fast.
- Update reporting: Pull call and direction data weekly alongside organic traffic.
Reviews have always mattered. AI search has turned them into primary training material. Local businesses that craft a steady stream of authentic, detail-rich feedback shape the very systems that decide their visibility.
04/28: AI Visibility Score: What It Is, Why It Matters, and How to Boost Yours
Yesterday, we talked about how customer reviews are starting to influence whether businesses show up in AI Overviews. Today, we’re zooming out from reviews and looking at the bigger visibility picture. AI search is quickly changing how people find answers, vendors, products, and recommendations. Instead of scrolling through page after page of Google search results, users are asking ChatGPT, Gemini, Copilot, Perplexity, and Google’s AI Overviews for direct guidance. That means your brand may still rank in Google, still have solid organic traffic, and still be missing from the answers your next customer actually sees.
That’s where the AI Visibility Score comes in. It’s not perfect, and it’s definitely not a magic marketing number, but it is one of the earliest ways to measure whether AI tools can find, understand, trust, and cite your brand.
What Is the AI Visibility Score?
An AI Visibility Score is typically a 0 to 100 benchmark that estimates how visible your brand or website is inside AI-generated answers. Depending on the tool, that can include:
- mentions
- citations
- links
- share of voice
- sentiment
- prompt performance
- whether AI crawlers can access and interpret your site
SearchScore, for example, describes its score as a 0 to 100 model based on multiple layers and scoring categories, including technical access, content extraction, and trust signals. Semrush’s AI Visibility Toolkit measures how brands appear in AI-generated answers and compares those mentions against competitors. RainmakerRank positions its score as a quick scan across platforms like ChatGPT, Perplexity, Google AI, and Bing Copilot, with a per-platform breakdown and gap analysis.
The wake-up call is the benchmark data. SearchScore’s April 2026 SAVI report audited more than 866,000 websites and found the average AI Visibility Score was only 34 out of 100. Even more eye-opening, 74.2% of audited sites were classified as “Invisible or Low Visibility.” In other words, most websites are not exactly crushing it in AI answers right now.
Why Marketers Should Care
Traditional SEO rankings still matter. They’re just no longer the whole visibility story. AI tools are creating summarized, conversational search journeys where users may never click through a traditional results page. If your brand is absent from those answers, you may be losing consideration before your website ever gets a chance to help.
A site can have healthy technical SEO and still perform poorly in AI discovery. SearchScore found average technical scores of 70.1 out of 100, while AI platform readiness averaged 34.1. What does this mean? Good SEO is necessary, but it’s not the same as being AI-visible.
This is why platforms like Semrush are adding competitive AI dashboards. Marketers need to see which prompts mention their brand, which competitors show up instead, and where their content is being ignored. It’s keyword tracking’s more chaotic cousin, and yes, it comes with a learning curve.
How the Score Is Calculated, and Its Limits
Most vendors follow a similar pattern. They:
- sample prompts
- check whether your brand appears
- analyze citations or mentions
- weigh supporting signals like structured data, topic authority, technical accessibility, and brand trust
The tricky part is that every vendor has its own formula. SearchScore’s current methodology uses a three-layer model: foundation, extraction, and reinforcement. Foundation includes crawl access, schema, llms.txt, and knowledge graph signals. Extraction looks at things like original data, topic depth, FAQs, and direct answers. Reinforcement includes press, reviews, backlinks, partner links, NAP consistency, and social proof.
Semrush focuses on brand mentions, prompt monitoring, competitor visibility, sentiment, and technical blockers. RainmakerRank offers a faster scan that checks AI visibility across four platforms and turns the results into a score with recommendations.
These tools are directional. AI platforms change, prompts vary, and scoring models evolve. A score should help you spot trends and gaps, not send the whole marketing team into a Slack spiral over a two-point dip.
Common Reasons Scores Stay Low
Low AI visibility usually comes down to one of a few boring but important issues. AI tools may not be able to crawl your site. Your structured data may be missing or weak. Your content may answer broad marketing questions without giving AI systems specific, quotable, well-organized information. Your brand may also lack enough outside trust signals for AI tools to feel confident recommending you.
SearchScore’s April report points to structured data as one of the biggest gaps, noting that structured data helps AI identify what a business does, who it serves, and why it should be trusted. The same report found that a smaller CBD retailer outscored major brands because its advantage was structural, not based on size or name recognition.
That’s the lesson: don’t respond to a low score by cranking out 25 generic blog posts. Find the actual problem first and address it with quality content.
Practical Steps to Improve Your AI Visibility Score
Lock Down the Technical Baseline
Start with the things that help AI systems access and understand your site. Check robots.txt for AI crawler access, add or review your llms.txt file, validate schema markup, make sure important pages render cleanly, and use structured data to define your organization, services, people, locations, and key content.
Strengthen Content Authority
AI tools reward clear answers, depth, and evidence. Build content that explains who you help, what you do, why it matters, and how someone should evaluate their options. Add expert input, original data, FAQs, comparison content, definitions, and use-case pages where they make sense. Thin “SEO content” probably won’t carry much weight here.
Build Brand Signals Across the Web
Your website is only part of the picture. Consistent business listings, strong reviews, social profiles, PR coverage, partner mentions, backlinks, and knowledge graph signals all help reinforce that your brand is a real, trusted entity. SearchScore’s methodology specifically includes press, reviews, community presence, backlinks, NAP consistency, and social proof as reinforcement signals.
Keep the Metric in Perspective
Treat AI Visibility Score the way marketers used to treat Domain Authority: helpful for benchmarking, dangerous as gospel. The number matters less than what it reveals. Are AI tools finding your brand? Are they citing competitors instead? Are they misunderstanding what you do? Are technical issues blocking access?
The goal isn’t to “win” a vendor score. The goal is to show up in the right AI-assisted buying journeys, with accurate information, for the people most likely to become qualified leads.
04/29: Ask YouTube: How YouTube’s Conversational Search Pilot Could Rewrite Video SEO
YouTube has always been part search engine, part entertainment platform, and part how-to library. Now, it’s testing a feature that could make the search side of the platform feel a lot more like an AI assistant.
The feature is called Ask YouTube, and it gives users AI-written answers to natural-language questions, with videos cited as supporting sources. That may sound like a small interface change, but for brands that rely on YouTube for education, product discovery, thought leadership, or top-of-funnel visibility, it could be a much bigger shift.
If this experiment sticks, YouTube visibility may depend less on traditional watch-time optimization alone and more on whether a video is clear, structured, and easy for AI systems to understand.
What Is Ask YouTube?
Ask YouTube is a conversational search experience currently being tested with U.S. YouTube Premium users 18 and older, with the pilot expected to run through June 8. Instead of typing a short keyword phrase and scanning a list of thumbnails, users can ask YouTube a more specific question and receive an AI-generated response.
The experience appears to include a written answer, a primary cited video at a relevant timestamp, and follow-up prompts that keep the search going in the same thread. In one example from YouTube’s announcement, the feature helps build a road-trip itinerary by pulling together a step-by-step plan and citing relevant videos along the way.
That is a meaningful change in user behavior. A person may get the answer they need before they ever click into a full video. The video still plays a role, but it acts more like a source inside an answer than a destination someone must choose from a results page.
Why This Could Reshape Video Visibility
For years, YouTube SEO has revolved around familiar signals: titles, descriptions, thumbnails, retention, engagement, and relevance. Those basics are not going away. Ask YouTube simply adds another layer to how content may be discovered.
In this format, videos become evidence for AI-generated answers. That means the best-performing content may not just be the video that earns the most clicks. It may be the video that most clearly answers a specific question, explains a topic in plain language, and gives YouTube enough structure to identify the right moment.
For brands, this could complicate reporting. A product demo, educational explainer, webinar clip, or customer story may influence an AI answer without producing a clean increase in views or watch time. That does not mean the content failed. It may mean the content is doing a different kind of job.
This is the same broader shift marketers are seeing across search: AI-generated answers can reduce the need for a click while increasing the value of being cited. Visibility is still valuable, but it may not always show up in the same metrics.
How To Optimize For Answer-Mode Discoverability
The good news is that brands do not need to throw out everything they know about video SEO. They do need to make the content easier for both people and machines to understand.
Start with transcripts and captions. If YouTube’s AI is using video content to build summaries, the spoken words matter more than ever. Clear captions, accurate terminology, and clean explanations give the system better source material to work with.
The structure of the video also matters. Instead of burying the main answer after a long intro, brands should consider adding an early answer segment that directly addresses the question the video is meant to solve. This does not mean every video needs to feel stiff or overly scripted. It does mean the point should be easy to find.
Chapters and timestamps can help too, especially for longer videos that cover several subtopics. They give YouTube a cleaner map of the content and make it easier to connect a user’s follow-up question to a specific section.
Metadata still has a role. Titles and descriptions should reflect the way buyers actually search and ask questions. A title like “How To Choose the Right Industrial Filtration System” is much more useful than a vague branded title that only makes sense to the internal team.
New Metrics Marketers Should Watch
Ask YouTube could make reporting messier before it makes it more useful. If AI summaries cite videos without generating a full view, marketers may see changes in visibility that do not line up neatly with traditional performance metrics.
That means teams should pay closer attention to YouTube Search traffic, suggested-video traffic, key moments, retention patterns, and click-through rate. A lower CTR may not always point to weaker content. In an AI-answer format, it may mean users are getting what they need earlier in the search experience.
This is where stakeholder expectations matter. If leadership is only looking at views and watch time, they may miss how video content is influencing discovery. Reporting will need to account for both direct engagement and indirect visibility.
What To Watch Next
The biggest question is whether Ask YouTube stays limited to Premium users or expands into a broader YouTube search experience. YouTube has indicated it aims to expand access beyond Premium users, which would make this a much more important shift for marketers to watch.
Reporting will be the next pressure point. If AI-generated citation impressions become part of video discovery, marketers will need better tools to understand when their videos are shaping answers, even when users do not click.
TL;DR: video SEO is moving closer to answer optimization. Brands that make their videos clearer, better structured, and easier to cite will be in a stronger position if YouTube decides this experiment is worth keeping.
04/30: Instagram Originality Policy: Why Consistent Original Content Is Now Non-Negotiable
Meta is making it much harder for low-effort recycled content to ride the recommendation algorithm. Its originality rules, which previously focused more heavily on reused Reels and duplicate videos, now also apply to photos and carousels on Instagram and Facebook. That means unedited screenshots, reposted tweet roundups, recycled infographics, and third-party images without meaningful edits may no longer be eligible for recommendation surfaces like Explore and Feed.
For brands, this is a very clear reminder that consistent original content is becoming a requirement for organic visibility, not a nice little bonus when there’s time.
What Changing
The Verge recently reported that Instagram is expanding its originality enforcement beyond video content to include photos and carousels. To be eligible for recommendations, accounts need to post content they created, designed, photographed, or materially edited. If your social calendar is mostly:
- screenshots of other people’s posts
- lightly repackaged trend graphics
- borrowed visuals with no meaningful brand input
Meta may decide that content does not deserve extra distribution.
Those low-lift posts wont necessarily disappear from your profile, but the real issue here is discovery. If a post is excluded from Meta's recommendation feeds, it becomes much harder for new people to find you. That is where the policy starts to matter for businesses. A brand can keep posting, keep filling the calendar, and still lose visibility because the content itself is not considered original enough to earn reach.
Why Meta Is Betting on Originality
Original content keeps platforms more useful, more interesting, and less cluttered with copycat posts. In March 2026, Meta said it had updated its original content guidelines to give creators clearer direction on how to get their work seen and recommended in Feed and Reels. The company also said views and time spent watching original Reels on Facebook approximately doubled in the second half of 2025 compared with the same period the year before.
That tells us two things. First, Meta believes originality is tied to engagement quality. Second, the platform has data that supports continuing down this path. Recycled content may still get quick reactions, especially if it borrows from something already popular, but Meta is trying to reward the account that actually made the thing in the first place.
For brands, this should sound familiar. Search engines, AI answer engines, and social platforms are all moving in a similar direction. They want clearer signals of authority, originality, usefulness, and trust. The platforms may all measure those signals differently, but the direction is the same: thin, copied, generic content is getting less room to hide.
Who Loses and Who Wins
The obvious losers are aggregation and meme accounts that rely on screenshots, viral reposts, quote graphics, and reused clips. That model was already risky from a brand and copyright perspective. Now it may also be a reach problem.
Brands on autopilot are in the danger zone too. We all know the kind of feed: press release reposts, generic holiday graphics, supplier content copied over with a caption, and “just checking the box” updates that technically count as posts but do very little to build the brand. Those posts may have always been underperforming, but this policy makes the problem harder to ignore.
The winners are not necessarily the brands with the biggest production budgets. The winners are the brands with a clear point of view and a repeatable process for creating their own material. A simple behind-the-scenes photo with a useful caption can carry more originality than a polished infographic copied from somewhere else. A quick Q&A carousel from your team can be more valuable than another recycled trend post with your logo dropped in the corner.
How to Adapt Your Social Strategy Now
Start with a simple audit of the last 30 days. Look for posts that rely heavily on someone else’s content, screenshots, or visuals your team did not create. Those are your risk areas. From there, define what “meaningful edit” means for your brand. Adding a logo is not enough. Adding commentary, context, original design, data, customer insight, or a clear brand perspective gets you closer.
Then shift your calendar mix toward lightweight original formats your team can actually sustain. Think quick staff photos, short videos, practical carousels, customer questions, simple motion graphics, mini case study takeaways, or commentary on industry changes. The goal is not to turn every post into a production, it's to stop relying on content that could belong to anyone.
Finally, update how you measure success. Reach still matters, but it should not be the only number in the room. Watch saves, shares, profile visits, branded search lift, assisted conversions, and the quality of engagement. Once the recycled filler comes out of the calendar, you may get a cleaner read on what your audience actually values.
Original Content Is the Price of Admission
Meta’s expanded Instagram originality policy is a wake-up call for any brand still treating social media as an afterthought. The algorithm is putting more weight behind original work, and that pressure is unlikely to slow down. Brands that build a consistent habit of creating useful, distinctive content will have a better shot at earning reach and staying memorable. Brands clinging to recycled filler may find themselves posting into a much smaller room.
