Analytics SEO

I start with a single question: what business outcome does this SEO work need to move? Once I have that, I pick three metrics that act as proxies. For example, when I was working on a B2B SaaS site, the goal was more trial sign-ups from organic. I tracked organic sessions, trial conversion rate, and top landing pages by conversion volume. I connected [Google Analytics](/google-analytics/) and Google Search Console to a [Looker Studio](/looker-studio/) dashboard that showed only those three numbers. I used [Google Tag Manager](/gtm/) to fire a custom event for trial sign-ups so I could see them in GA4. The dashboard took me two hours to build and another hour to verify the data. I did not add any secondary metrics for the first month. This forced me to look at the same three numbers every week and notice patterns. After a month, I added a filter for new vs returning visitors only when I noticed that returning visitors were converting at a higher rate. That filter was a dimension, not a metric. The principle is: start with the minimum viable measurement. You can always add later, but you cannot unsee noise.

I also include a weekly action step in the dashboard: a text box where I write what I will do based on the data. This turns measurement into a decision. Without that, analytics is just a report. I have seen too many organisations collect data for months without acting on it. I avoid any metric that cannot be tied to an action. For example, impressions are interesting but not actionable unless you have a click-through rate problem. So I track impressions only when I am specifically working on CTR. Otherwise, I ignore them.

Before you trust the dashboard, you need to check that the data is correct. I always verify one metric manually by comparing a few days of Search Console data with a manual search. I also check that my Google Analytics property is collecting the right events. This sounds obvious, but I have found tracking errors in half the accounts I have audited. I use a checklist: does the event fire when I expect it? Does the goal match the business outcome? I call this the 'sanity check' and I do it every time I set up a new dashboard. I will not make a decision based on data I have not personally verified.

The biggest watch-out is the average position metric in Google Search Console. It is an average across all queries, so a page that ranks first for one query and tenth for another will show an average of 5.5, which is meaningless. I always look at the distribution instead. Similarly, the click-through rate in Search Console can be misleading because it is based on estimated impressions. I have seen cases where Google shows 1,000 impressions for a query that actually had zero visibility in the SERP because of a knowledge panel. So I cross-check with a [rank tracking](/rank-tracking/) tool to see the actual positions for my target queries. I use a simple rank tracker that checks the exact SERP. This gives me a second opinion.

Another watch-out is the attribution window. Google Analytics 4 uses a default attribution model that may not reflect the true value of organic search. I have had clients who saw organic conversions drop after switching to GA4 because the model changed. I now always check the model and, if necessary, create a custom model that gives credit to the first interaction. This is especially important for content that drives awareness but not the final click. Do not let a default attribution model fool you into thinking organic is underperforming.

I also see people make the mistake of comparing their analytics to competitors without adjusting for differences in traffic. For example, a competitor might have higher organic traffic simply because they have more pages. I use a tool like Semrush to compare share of voice, but I always normalise by number of pages or brand strength. I do not compare raw numbers. I use [SEO competitor analysis](/seo-competitor-analysis/) to understand the landscape, but I always question the data source.

Finally, I watch out for data sampling in Google Analytics. For large sites, GA4 may sample data when you run a report. I always check the sampling indicator. If it is sampled, I either use the unsampled report or export to Google Sheets for analysis. I do not make decisions on sampled data. I also keep an eye on the 'other' row in reports, which can hide important data. If a report shows a large 'other' category, I investigate further.

I spent my first year in SEO measuring everything. I had dashboards with 20+ metrics: impressions, clicks, average position, CTR, bounce rate, time on page, pages per session, goal completions, revenue, new users, returning users, device breakdown, location breakdown, hour of day, and more. I thought I was being thorough. In reality, I was drowning in data and making no decisions. The dashboard was a novelty that my manager looked at once and never again. I learned the hard way that you do not need more data; you need fewer, better metrics. I now measure only what I will act on.

I also got wrong the assumption that all organic traffic is good. I remember a client whose organic traffic spiked after a blog post went viral for a query that had nothing to do with their product. The traffic was high, but the bounce rate was 95% and no conversions. I was celebrating the traffic growth until I looked at the conversion data. That experience taught me to always filter by relevance. I now always check the landing pages that drive the most traffic and ask: does this visitor have a reason to stay? I use [SEO tracking](/seo-tracking/) to monitor which pages actually contribute to the goal.

Another mistake was relying too heavily on ranking reports. I used to report on keyword positions every week, but I never connected those positions to actual visits. A keyword can rank first and get zero clicks if it is a featured snippet or a zero-click query. I now use the Search Console click data to see which positions actually send traffic. I also stopped caring about the top 10 rankings for long-tail terms that have no search volume. I now focus on queries that have at least some impressions.

I also initially ignored the technical side of analytics. I assumed that if I installed Google Analytics, it would work. But I later found that I had duplicate tracking, missing events, and a bot-filtering issue that inflated my traffic numbers by 30%. Now I always audit my tracking setup before I trust any data. I use Google Tag Manager to manage tags and I regularly test events using the preview mode. I also check the 'all traffic' report in GA4 to see if the numbers make sense. If you have not verified your tracking, you are operating on a guess.