Web analytics will show what works well on the site and what rather scares away visitors. Gradually, you will develop hypotheses, you will understand what else needs to be improved on the site.
Website development is an ongoing process. Implementing something new and then evaluating how it impacts advertising results.
And this is also a process that should not take place in a vacuum. Consider the following:
web analytics data;
sales department data (what questions do customers ask? What is whatsapp number data important to them? Is this reflected on the website?);
regular monitoring of competitors.
This also applies to contextual advertising.
4. There have been no calls for 2 days! Something needs to be done
(this is again a question of statistics)
But is it necessary? It may be a matter of dispersion. Dispersion is a measure of the spread of a random variable around the mathematical expectation. More details about it are well written in the article: "Burn Your Analytics".
Let's say you get 30 leads per month on average. This doesn't mean the graph will look like this:
Because it might look like this:
That is, the distribution will not be such that with 30 applications per month you receive 1 application per day.
Conclusion: it is impossible to evaluate any results based on short periods of time when sufficient statistics have not been collected.