SEO Reviews

An SEO review is not a report. It is a diagnostic session where you take raw data from tools like Google Search Console and [Google Analytics 4](/google-analytics/), validate it, and then decide what to do next. Many people generate fancy dashboards and call it a review, but that is like taking a photo of a patient and calling it a diagnosis. The real work starts when you question the numbers.

I start every review by checking that the data is real. I look for sudden drops or spikes that match a tracking change, not a real traffic shift. For example, last year a client showed a 15% drop in organic conversions. I almost panicked, then I remembered we had changed the [Google Tag Manager (GTM)](/gtm/) container two weeks earlier. The conversion tag was firing on the wrong button. Once fixed, the real number was flat. Without that validation step, I would have wasted time chasing a phantom problem.

After validation, I compare current performance to a meaningful baseline. I use the median of the last 6 to 8 weeks, not the previous month, because one week can throw off a month-to-month comparison. If I see a 10% drop in clicks from one week to the next, I check if the same week last year had a similar dip. If it did, that is likely seasonality, not a Google update.

Finally, I force myself to write down one thing I will do differently based on the review. It could be updating a title tag, disavowing a bad link, or pausing a content campaign. If the review does not change my next action, I wasted my time.

An SEO review matters most when you are about to spend time or money on SEO. If you are planning a content campaign, you need to know which pages are already performing and which are underperforming. I once reviewed a site that had been publishing blog posts weekly for six months with no traffic growth. The review showed that 80% of the traffic came from three old pages, and the new posts were all targeting the same low-competition terms that nobody searched for. The fix was simple: stop writing for keywords with zero volume and redirect effort to updating the three winning pages. That decision saved the client months of wasted work.

Reviews also matter when you suspect a problem. A sudden drop in rankings is obvious, but a slow decline over three months is easy to miss. I run a [rank tracking](/rank-tracking/) check every week, but I only review the trend every month. Last quarter, I noticed a steady drop in impressions for a set of product pages. When I dug into the [SEO analysis](/seo-analysis/), I found that Google had started showing a competing site's product carousel above our results. The fix was to add structured data for product variants. Without the review, I would not have caught it.

Another time a review matters is when you are reporting to a client or boss who wants to see [SEO results](/seo-results/). A raw number like "organic traffic up 12%" is meaningless without context. I always pair the number with a baseline, a segment, and a reason. For example: "Traffic is up 12% vs last month, but that is driven by a single blog post that went viral on Reddit. Core product pages are flat." That honest breakdown builds trust and leads to better decisions.

I used to think an SEO review was a summary of last month's performance. I would open Google Analytics, pull organic traffic, look at rankings, and call it a day. I missed the most important part: validation. I remember a client where I reported a 20% increase in organic traffic, but it was actually a tracking bug that duplicated pageviews for a week. I celebrated the win, they celebrated the win, and then we spent the next month wondering why the number dropped back to normal. I learned the hard way that you must verify your data source before you analyse it.

I also used to ignore segmentation. I would look at the aggregate organic traffic line and think everything was fine. But the aggregate hides everything. A 5% decline overall could be a 30% decline on your money pages balanced by a 20% increase on junk pages. I now segment by page type, by device, and by geography before I make any judgement. I once found that a site's mobile traffic was dropping while desktop was rising, because the mobile menu broke after a WordPress update. The aggregate looked stable, but the segment told the truth.

Another mistake was treating reviews as a one-way process. I would produce a [SEO report](/seo-reports/) and send it, expecting the client to act. Now I schedule a 30-minute call after every review to walk through the findings and agree on next steps. That conversation is where the real value happens. Without it, the review is just a document that gets opened once and forgotten.

Finally, I used to compare every metric to the previous month, ignoring seasonality. A retail site's traffic always drops in February. I now compare against the same period last year and against a rolling 6-week average. That small change stopped me from raising false alarms.