SEO Tracking

I start with Google Search Console and Google Analytics because they are free, they cover the two halves of the story, and they are the only sources Google itself trusts. Search Console gives me clicks, impressions, average position, and CTR at the query level. GA4 gives me sessions, engaged sessions, and conversions. The first thing I do is define three to five conversion events in GA4 that match the business: purchases, form submissions, demo requests, phone calls, or newsletter sign-ups. Without those events, you are measuring activity, not ROI.

Next I connect both tools to [Looker Studio](/looker-studio/) and build a single [SEO dashboard](/seo-dashboard/) that lives on a shared drive. The dashboard has three tabs: daily, weekly, and monthly. The daily tab shows total clicks and impressions from Search Console, plus top-20 queries by impression change. I check it every morning with coffee — if impressions dropped more than 15% overnight I investigate immediately. The weekly tab adds CTR trends by device and a table of pages that lost more than 10% of clicks. The monthly tab compares current month to the same month last year to strip out seasonality.

For rank-specific work I run a [rank tracking](/rank-tracking/) tool against 50 terms that drive the most revenue. I update it daily and look at the distribution of positions — how many are in top 3, top 10, top 20 — rather than the average. I also export Search Console query data weekly and upload it to the dashboard so I can see which non-targeted queries are stealing or gifting traffic. The whole setup takes about half a day to build and then ten minutes a day to review.

The key decision rule: if a metric cannot be connected to a conversion, I drop it from the dashboard. Domain Authority and Trust Flow are interesting but they do not tell me why a page stopped converting last Tuesday. I prioritise impression share, click share, and conversion rate by landing page. That forces me to act on data that affects the business, not the vanity graph.

The biggest trap is measuring what is easy instead of what is useful. Average position in Search Console looks clean but it is a weighted average that hides the distribution. A keyword that ranks 1 for 1000 impressions and 50 for 10 impressions averages to 1.5, which tells you nothing. I now look at the impression-weighted CTR and the number of clicks at each position band. If I see a query with high impressions but low CTR, I check whether a rich snippet is missing or the title tag does not match the intent.

Another trap is relying on rank-tracking data in isolation. Tools that simulate SERPs often show different results than what real users see because of personalisation, location, and device. I cross-reference every rank tracker with the actual clicks from Search Console. If my rank tracker says position 3 but Search Console shows 50 clicks from position 5, the tracker is wrong and I adjust the tool's settings or accept the discrepancy.

Attribution in GA4 is another blind spot. Organic traffic often assists rather than closes. If you only credit the last click, you undervalue SEO. I set up a model comparison report that shows how many conversions were assisted by organic even if the final click was paid or direct. Without that, you will cut SEO budget because the last-click numbers look lower than they should.

Finally, do not ignore data sampling. GA4 samples reports when query volume exceeds thresholds. If you are a large site, your organic traffic report might be based on a fraction of sessions. I always check the small 'sampled' label and, if present, switch to a longer date range or use the unsampled report in Google Ads if you have that linked. Sampling hides volatility and makes your [SEO reporting](/seo-reporting/) unreliable.

I treated SEO tracking as a one-off setup task for my first year in the industry. I installed Search Console, connected GA4, built a dashboard, and then checked it once a month. I missed a slow decline in impressions for six weeks because I was not looking. By the time I noticed, a core update had already hit and I had no baseline to compare against. That mistake cost the site 40% of its organic traffic and took three months to recover. I now check my dashboard daily.

I also ignored keyword cannibalisation until it was too late. I had three different pages targeting the same head term, and Search Console showed them competing for the same impressions, each with a different average position. The overall position for the query looked decent, but none of the pages accumulated enough authority to rank in the top 3. I now run a quarterly [SEO analysis](/seo-analysis/) using Screaming Frog and Search Console data to catch pages that target the same query and consolidate them.

Another error: I did not track conversions from organic for the first six months. I was obsessed with rankings and traffic volume. I thought more traffic meant success. Then the client asked for revenue attribution and I had nothing. I had to retroactively set up goals, losing three months of data. I now define all conversion events before I connect Google Analytics. That single change made my SEO reports defendable in board meetings because I could point to actual leads and sales.

The lesson is this: tracking is not a setup, it is a habit. You have to look at the numbers every day, question them, and verify them against what real users do. The tools change, the metrics change, but the habit of watching your data closely never goes out of date.