How to tell if your marketing actually worked
Traffic went up after you posted. That does not mean the post did it. How to read Search Console before and after an action without fooling yourself.
You published something, and a week later traffic was up. It is very tempting to decide the post did it. Most of the time you cannot know that, and the most useful thing a tool can do is tell you how much it does not know.
Populr measures what happens after every action it recommends, using Google Search Console. This is the method, including the parts that make the numbers weaker than they look.
Compare windows, not moments
A single day tells you almost nothing. Search traffic moves with the day of the week, the season, the news, and ranking changes you will never be told about. The useful comparison is a window before the action against an equal window after it, with the same number of each weekday in both.
Search Console helps and hinders here. Its performance data arrives a couple of days late, so the after-window is never quite as fresh as it looks. It also keeps sixteen months of history, which is enough to check whether the same weeks last year did the same thing without you.
Read the four numbers together
| Metric | What it tells you | How it misleads |
|---|---|---|
| Impressions | How often you appeared in results | Rises when you start ranking for searches nobody clicks, or the wrong ones |
| Clicks | How often someone chose you | Small numbers swing hard: ten to fifteen is +50% and means almost nothing |
| CTR | Clicks per impression | Falls when impressions grow faster than clicks, which can be good news |
| Position | Average rank where you appeared | An average over every query; one new low-ranking query drags it down |
None of these is the result on its own. A page whose impressions doubled while clicks stayed flat has probably started appearing for searches it does not answer. A page whose position got "worse" may simply have started ranking for more things. Read them together or not at all.
Association is not causation, and we say so
Even a clean before-and-after only shows that something changed after you acted. It does not show your action changed it. A competitor may have dropped out of the results. Google may have changed how it ranks. Your topic may have been in the news that week.
So we report an association score, not an attribution. In our product, 50 means no movement and above 50 means the metrics improved after the action, and every score carries a confidence that reflects how much data was behind it. A score built on thirty clicks is shown with low confidence because it deserves it.
Associations become useful in aggregate. If posts on one channel are reliably followed by improvement and posts on another are not, that pattern is worth acting on, even though no single post proves anything. That is the level we make claims at — channels, over time — rather than individual posts.
A checklist you can use without us
- Write down what you expect to change, and on which pages, before you act.
- Pick equal before and after windows of at least two weeks each, aligned by weekday.
- Wait for Search Console to catch up before reading the after-window.
- Look at clicks, impressions, CTR and position together, per page, filtered to the queries that matter.
- Check the same weeks last year for a seasonal pattern.
- If the numbers are small, write "inconclusive" and mean it.
Common questions
How do I know if a blog post increased my traffic?
Compare equal before and after windows in Search Console for that page, aligned by weekday. Even then you have an association, not proof — rankings, competitors and seasonality all move independently of your post.
Why did my CTR drop when my traffic went up?
CTR is clicks divided by impressions. If you start appearing for many more searches, impressions can grow faster than clicks, which lowers CTR even while clicks rise. Read the metrics together rather than one at a time.
How long should I wait before measuring SEO results?
At least two weeks after the change, plus a couple of days for Search Console's data to arrive. Shorter windows are dominated by day-to-day noise rather than by anything you did.
Populr is an AI CMO for founders without a marketing hire. Paste your website and it builds the plan.
Try it free for a month