Most SEO reports lead with rankings, and rankings have quietly become the least informative number in the document. Position is personalised, so there is no single ranking to report — and a growing share of searches are now answered above the results entirely, which means being third on a list fewer people scroll to describes less every quarter.
What is wrong with leading on rankings?
Two problems, and the second is the one that has changed recently.
Position is not a single number. Results vary by location, device, search history and the exact phrasing used. A ranking report shows one measurement from one context, presented as though it were the state of the world. It is closer to a sample than a fact.
The list is no longer the whole page. When we checked 980 Singapore keywords about SEO services, keywords carrying 68% of the search volume returned an AI Overview above the results. Your position beneath that answer is a real thing, but it is a smaller share of the outcome than it was, and a report that measures only it will look stable while your actual visibility moves.
What should be in the report instead
| Measure | Why it matters | How often |
|---|---|---|
| Enquiries from organic search | The only number that settles whether the work paid | Monthly |
| Visibility across a basket of terms | Baskets move honestly; single positions move on noise | Monthly |
| Presence in AI answers | Increasingly where the answer is given | Quarterly |
| Technical health | Regressions happen silently after site changes | Monthly |
| Pages entering year two still producing | Proves the compounding is real | Quarterly |
| Rankings | Useful as an input, misleading as a headline | Monthly |
How to check whether AI answers name you
This takes twenty minutes and almost nobody does it.
- Write down the five or six questions a buyer would ask before choosing someone in your category.
- Put each into ChatGPT, Perplexity and Google, and record which businesses get named.
- Note whether you appear at all, and if so whether the description is accurate.
- Repeat the same queries monthly. Individual sessions vary, so only the trend is meaningful.
The results are frequently uncomfortable. Businesses that rank well often find they are absent from the answers entirely, usually because their pages have nothing specific enough to quote. That gap is exactly what answer engine optimisation addresses, and it is why we treat it as part of doing SEO properly rather than as a separate product.
The trap in traffic numbers
Organic traffic is a better headline than rankings and still not the answer. Traffic can grow from terms that never convert — informational searches that bring readers with no intention of buying anything.
If traffic is climbing and enquiries are flat over two quarters, that is not slow progress. It is a signal that the content is aimed at the wrong questions, and more of it will not fix the direction.
The reporting habit that causes the most damage
Weekly ranking reports. At that resolution, ordinary fluctuation looks like a trend, and the predictable response is to change something. Then something else. Within a quarter the strategy has been revised four times and none of the changes had time to take effect.
We have seen more programmes harmed by over-reaction to weekly noise than by any single wrong decision. Monthly for activity, quarterly for strategy, and a fixed rule that nothing changes direction on one month’s data.
If your traffic is up and your enquiries are flat for two quarters, that is not slow progress. It is the wrong direction, executed well.
What a good report lets you do
A useful report answers three questions without interpretation: is this producing business, is the trend real or noise, and what changed that we should respond to?
If the document in front of you cannot answer those, the problem is not that your agency is underperforming. It is that neither of you can tell, which is a worse position to be in and considerably more common.
Frequently asked questions
What should an SEO report actually contain?
A baseline, enquiries attributed to organic search, visibility across a basket of terms rather than single positions, technical health, and whether AI answers in your category name you. Rankings belong in it, but as one input rather than the headline.
Why are ranking reports misleading?
Because position is personalised by location, device and history, so there is no single ranking to report. And because a growing share of searches are answered above the results, meaning position on the list describes progressively less of what actually happens.
How do I measure whether AI answers mention my business?
Ask the engines your main commercial queries and record who gets named. Do it consistently — same queries, same intervals — because individual sessions vary. The trend across repeated checks is more reliable than any single result.
Is organic traffic a good measure of SEO success?
Better than rankings, worse than enquiries. Traffic can rise from terms that never convert. If traffic climbs while enquiries stay flat, the content is attracting the wrong readers.
What is a reasonable reporting frequency?
Monthly for activity and enquiries, quarterly for anything strategic. Weekly ranking reports encourage reacting to noise, since normal fluctuation looks like a trend at that resolution.
Send us your last three months of SEO reporting. We will tell you what it is not measuring, whether or not you engage us. See what we report on, message us on WhatsApp, or send an enquiry.
Sources
Written by Charlotte Zhang, Operations Partner at Kaizenaire Pte Ltd (UEN 201932071D). Charlotte co-founded Kaizenaire with Ken Tan in 2019 and runs client delivery, after four years in operations at OCBC and HSBC.