For years, the monthly SEO report had a simple centerpiece. A table of keywords, each with a ranking position, and a few arrows showing who moved up and who moved down. Clients understood it at a glance, and agencies could point to it as proof of progress. That table is now one of the least trustworthy items in the report, and many marketing teams are quietly discovering that the numbers they relied on no longer mean what they used to.
Rank tracking tools are under pressure from changes Google has made to how its results are served, and the way people find information has also moved beyond a single list of ten links. This matters for every business investing in SEO in Dubai, because reporting built on shaky numbers leads to shaky decisions. This guide explains what has gone wrong with rank trackers, what to measure instead, and how to build a reporting approach that combines Google Search Console, GA4, and AI citations into a clear picture of progress. At The Share of Voice, this is the reporting model we now use with clients.
What Has Happened to Rank Trackers in Search Engines
To understand why reporting needs to change, it helps to know what has been going on behind the scenes.
Rank trackers work by requesting Google results pages in bulk and recording where each website appears. For a long time, they relied on a parameter that let them load a hundred results in a single request. In September 2025, Google stopped supporting it. Tools that had been collecting deep ranking data cheaply suddenly needed many more requests to gather the same information, and many reduced how far down the results they checked, often limiting tracking to the first couple of pages.
The change had an unexpected side effect on Google’s own data. Many site owners saw a sharp fall in Search Console impressions around the same time. A large analysis of several hundred sites found that most lost unique ranking terms and most saw fewer impressions. The most widely accepted explanation is that automated tools loading deep results had been inflating impressions, and when that stopped, Search Console data became closer to the reality of human searches. This means any comparison of impressions across that date needs care, because the drop reflects a change in measurement, not necessarily a loss of visibility.
Through 2026, the pressure continued. Google has added measures that make scraping harder, including redirects in results links and requirements that make automated requests more expensive. Industry reports in recent weeks described third-party rank data as unreliable, with some vendors reporting that a share of the results pages they collected were not genuine. Many tools have since made fixes, but the episode showed how dependent rank tracking is on a window that Google can narrow whenever it chooses.
Why a Single Rank Number Was Always Fragile
Even before these changes, a ranking position was less solid than it looked. Results vary by location, device, language, and personal search history, so two people searching the same phrase can see different orders. Features such as map packs, shopping results, videos, and AI-generated summaries push organic listings down the page, which means position one does not always deliver the traffic it once did.
For a market like Dubai, with a diverse, multilingual population and strong mobile use, these differences are even more pronounced. A keyword that ranks third in one neighborhood on a phone may rank differently elsewhere, and a rank tracker that checks from a single location is only sampling one version of reality.
None of this means rankings are meaningless. It means they are one directional signal among several, not the measure of success.
The New Reporting Stack
A better approach is to build reporting in layers, using data that comes directly from the sources rather than from scraping. Four layers work well together.
Layer One: Google Search Console
Search Console is the foundation because its data comes directly from Google. It shows what actually happened, how often your pages appeared, how often people clicked, and the average position at which they appeared.
Several practices make it more useful for reporting. Focus on clicks as your primary measure, since impressions were distorted by the measurement change described above. Look at click-through rate alongside position, because a page that ranks well but attracts few clicks may have a title, snippet, or intent problem, or may be losing attention to features at the top of the page. View position as a distribution rather than a single average, for example how many queries sit in the top three, the top ten, and positions eleven to twenty, since movement across those bands is more informative than a headline number.
Separate branded from non-branded queries. Branded searches, where people look for your company by name, are usually stable and highly clickable, and they can hide weakness in the non-branded traffic that reflects true SEO reach. Compare periods of equal length, and annotate your reports with the date of the measurement change so that trends are not misread. For a Dubai business, also filter by country and, where relevant, by page language, so you can see how English and Arabic content perform separately.
Layer Two: GA4
Search Console tells you how people arrive from search results. GA4 tells you what they do next. This is where SEO progress becomes business progress. Artificial intelligence can help companies improve SEO faster by spotting patterns in data at scale and in real time.
In GA4, look at organic search sessions to your key landing pages, but do not stop at traffic. Track engagement, and above all track key events and conversions that matter to the business, such as enquiries, bookings, calls, form submissions, or purchases, attributed to organic search, because AI tools are built to give users a relevant, useful, and personalised experience, so conversions matter more than raw traffic. A trend of flat traffic with rising conversions can be a positive sign, since a more qualified audience may be arriving.
GA4 also helps you identify traffic from AI platforms. Referrals from tools such as ChatGPT, Perplexity, Gemini, and Copilot can be grouped into a custom channel so that you can track them separately from standard organic search. This traffic is still small for most sites, but it is growing, and it tends to be higher intent because the visitor has already received a recommendation. Keep in mind that not every AI visit carries a referrer, so this number understates the true effect.
Layer Three: AI Citations and Generative Engine Optimization
The third layer addresses what rank trackers cannot see at all, whether AI systems mention, cite, or recommend your brand. This matters because a growing share of research now happens inside ai generated answers, where there is no list of positions to track, and these tools aim to provide a user experience that is relevant, useful, and personalized.
Measuring it requires a different method. This work is generative engine optimization (geo). Build a representative set of questions that real customers would ask, covering your services, comparisons, and common concerns, in both English and Arabic if relevant. Effective work here starts with user intent and clear, direct answers to the questions real customers ask. Run those questions through the main platforms, including Google’s AI features, ChatGPT, Perplexity, and others, and record whether your brand is mentioned, whether your website is cited, and how you are described. Repeat the exercise on a regular schedule, because AI answers vary between runs, and a single result is not a reliable measure.
From this data, you can report citation frequency, share of voice compared with competitors, the pages that are being cited, and the sentiment of the description. Strong visibility usually depends on quality content that AI systems can retrieve, understand, and reuse, and AI can help individual companies do SEO better and faster by analyzing patterns in data at scale and in real time. There are tools that automate parts of this, and a well-organized manual process is a perfectly sound way to begin.
Layer Four: Business Outcomes
Finally, connect all of this to results the business cares about. In ai search, the customer journey can compress because consumers may discover, compare, and choose a provider within a single interaction before ever clicking through, which is why content that is easy for AI systems to retrieve, understand, and reuse is more likely to appear in generative AI responses. Track leads, sales, and revenue from organic search using your CRM or sales records, not only analytics. Monitor branded search volume over time, since rising brand searches often indicate growing awareness, including awareness created by AI answers that do not generate a click. If the sales team reports that leads are better or worse, that is evidence that no dashboard should override.
What to Do With Rank Trackers Now
Rank trackers have not become useless. They are best used with realistic expectations. Treat them as a directional sample, not a precise measurement. They also cannot show whether your brand appears in generative engines or other AI answer surfaces. Focus on the top ten to twenty results, where data is most reliable. Look at trends over weeks rather than daily changes, which can be noisy. For local businesses, consider tools that check from multiple locations, so you can see how visibility varies across areas. And never act on a ranking drop reported by a tracker alone.
A simple cross-checking rule helps. Before reporting a decline, confirm it in Search Console, check whether traffic and conversions in GA4 moved in the same direction, and ask whether leads have changed. If the tracker says rankings fell, cross-check traditional search signals alongside AI visibility before concluding there is a real decline. If clicks, conversions, and enquiries are steady, the problem is probably with the tracker.
Rank in LLMs: A Different Kind of Visibility
When people talk about how to rank in LLMs, it is worth being clear that the concept differs from both traditional search and traditional search engines. There is no position one in a ChatGPT answer. An AI system selects and synthesizes sources, and your brand may be named, cited, paraphrased, or left out, depending on the question, the platform, and even the particular run.
This means that to rank in LLMs, you are really aiming to be a trusted, frequently cited source across a range of relevant questions. The signals that help include clear, well-structured content that answers specific questions, consistent information about your business across the web and across multiple sources, credible third-party mentions, regularly updated pages, and structured data as a supporting cue rather than a standalone fix. Reporting on this should therefore focus on frequency and context, how often you appear across a set of prompts, in what role, and next to which competitors, instead of any single position. When they align with your site, review sites, reviews, and social platforms can also help AI systems discover and evaluate a brand.
It also helps to remember that the different platforms do not draw on identical sources, and the overlap with traditional Google rankings is partial. A strong position in one does not guarantee presence in another, so report each platform separately.
How to Rank in SEO Strategy Without Relying on One Number
The goal of ranking in SEO and ai search has not changed. You still want to be visible when people search for what you offer across search engines. What has changed is how you demonstrate it. Google now shows AI-generated answers directly in search results for many English searches in the MENA region, so reporting should reflect both classic rankings and AI visibility. Instead of one ranking figure, a convincing report shows a set of linked signals: Search Console clicks and the share of queries in strong positions, GA4 organic conversions, AI citation frequency, and business outcomes. When these move together, you have strong evidence of progress. When they diverge, you have a prompt to investigate.
What a Good SEO Report Looks Like Now
A practical monthly report might contain a few clear sections. A headline summary describing what changed and why it matters, in plain language. A Search Console view of clicks, click-through rate, and position distribution, with branded and non-branded separated. A GA4 view of organic conversions and the quality of traffic, including any AI referral traffic. Some companies also pair this SEO view with separate search engine marketing dashboards. An AI visibility view showing citation frequency, competitors, and notable changes. A business outcomes section linking SEO to leads and revenue. And a short list of actions for the next period, based on what the data revealed.
Keep the report short and focused. Clients do not need a hundred keywords. They need to understand whether the work is producing results and what is being done next. If the client is also running SEM campaigns, the report should track cost per click as a key measure of financial efficiency and return on ad spend as a vital KPI in a dedicated paid-media section rather than mixing them into SEO results.
Common Reporting Mistakes
Several mistakes are common in this transition. Treating tracker positions as facts, and reporting drops without verifying them, is the most frequent. Comparing impressions across the September 2025 measurement change without noting it creates false stories of decline. Reporting only traffic and ignoring conversions can hide improvements in quality. Ignoring branded versus non-branded splits can mask weakness. Tracking AI visibility once and treating it as a fixed number overlooks the natural variation between runs. And presenting too many metrics without interpretation leaves clients confused rather than informed.
Why This Matters in Dubai
Dubai’s market adds a few specific considerations. Search behavior is bilingual, and voice search is growing, especially on mobile devices, so reports should separate English and Arabic performance. Many customers move from search to messaging apps or phone calls, so conversions can occur outside the website, which means tracking calls, chat clicks, and CRM data is essential. Voice queries are often longer, more conversational, and frequently tied to local intent, so businesses need local SEO basics in place to connect with their target audience. That includes keeping an up to date Google My Business profile and consistent local citations so local search visibility improves for people nearby. The audience is also an early adopter of AI tools, and 37% of active AI users in the UAE start searches with AI rather than Google, so AI visibility may matter sooner here than in many other markets. And competition in sectors such as real estate, healthcare, and hospitality is intense, so credible reporting that identifies real opportunities is a genuine advantage.
How The Share of Voice Reports Progress
At The Share of Voice, our reporting starts with Search Console and GA4 as the source of truth, supplemented by AI citation testing and tied to client CRM data wherever possible. We use rank trackers as directional inputs, never as the sole evidence, and we apply a verification rule before reporting any significant change. We separate English and Arabic performance and annotate reports with relevant measurement changes so that trends are read correctly.
For clients investing in SEO in Dubai, this gives a more honest and more useful picture, one that connects visibility to enquiries and revenue and identifies where to focus next, especially in a market where AI adoption is unusually high.
In the UAE, 97% of businesses and residents use AI, investment remains strong across public and private initiatives, and Arabic-centric models are advancing how large language models interpret local queries, which makes this reporting approach more important now and for the future.
Final Thoughts
Rank trackers are struggling because the data they depend on has become harder and more expensive for them to collect, and because a single ranking number never captured the full reality of search. At the same time, discovery is spreading from traditional search into AI search answers, where there is no ranking to track at all. The response is not to abandon measurement but to build it on better foundations.
A reporting approach combining Search Console, GA4, and AI citations, as a practical GEO strategy for tracking whether your brand is cited or recommended across AI platforms, anchored in business outcomes, gives a more reliable view of how you rank in SEO and how you rank in LLMs. It is also easier to explain, because it shows what customers did, not just where a tool says you stand.
Get in touch with The Share of Voice to find out how your SEO progress looks when measured with data you can trust, and how to build a reporting model that holds up in the new search landscape and prepares you for the future of search.
Frequently Asked Questions
- Why are rank trackers less reliable than they used to be?
Google removed the option to load a hundred results in one request in September 2025, making deep tracking much more expensive, and it has since made automated scraping harder. As a result, many tools limit how deep they check, and some have reported incomplete or unreliable data in recent weeks. Rankings also vary by location, device, and personalization, so a single position was never a perfect measure.
- Why did my Search Console impressions drop in September 2025?
Many sites saw impressions fall after Google stopped supporting the hundred-results parameter. A widely accepted explanation is that automated tools loading deep results had been inflating impression counts, and that stopped. This means comparisons across that date should be treated with caution, and clicks are a more dependable measure.
- How can I measure SEO progress without a rank tracker?
Use Search Console for clicks, click-through rate, and the distribution of positions, GA4 for organic conversions and traffic quality, AI citation testing for visibility in AI answers, and your CRM for leads and revenue. Together, these show whether SEO is producing real results, even if you do not track individual keyword positions.
- Is it possible to track AI referral traffic in GA4?
Partly. You can create a custom channel group that captures visits from platforms such as ChatGPT, Perplexity, Gemini, and Copilot based on their referrers. However, not every AI visit carries a referrer, so the figure will understate the real effect and should be treated as a minimum.
- How do I rank in LLMs?
There is no single ranking position. You aim to become a source that AI systems trust and cite, which is the core idea behind geo, by publishing clear, well-structured content, with strong extractability and clear entities that help systems interpret your brand, that answers specific questions, keeping information about your business consistent across the web, earning credible third-party mentions, and updating key pages regularly to optimize visibility. Measure progress by how often you appear across a set of relevant prompts over time.
- Should I stop using rank trackers altogether?
No, but use them as a directional sample rather than a precise measurement. Focus on the top ten to twenty results, watch trends over weeks, and never report a drop based on a tracker alone. Confirm any change in Search Console, GA4, and your leads before drawing conclusions.
