any marketer and product manager would give his soul for this
How to make in-depth statistics that will reveal certain patterns in people's perceptions and behavior?
Let's remember the story about the planes with holes that flew to the base.
I close with a spoiler for those who are familiar with the story.
Hidden text
Abraham Wald, who worked at Columbia University's statistical research group during World War II, helped figure out how to strengthen the fuselage of airplanes so they would suffer less damage during combat.
He realized that those parts of the machine that received the most damage did not need reinforcement – after all, the plane was able to return to base. It is necessary to pay attention to the parts of the hull and wings that have survived. This decision saved more than a dozen pilots and avoided serious economic losses.
Just as in the example above, the statistics of any project contain data with varying degrees of value for making the right strategic decision.
For example, among marketers and sales managers there is an opinion that when analyzing requests, correspondence and calls with potential clients, you need to look for objections that can be closed with new scripts. But experienced pros know that it is more profitable to study the profiles of clients who buy quickly and easily, without a bunch of questions and long negotiations.
Of course, it is important and necessary to take into account lost leads, but this task should not be in first or second place in importance.
Ways for a webmaster/marketer to collect statistics:
Polls.
Scrolling maps by page.
Buttons and UTM links.
Trends in visitor behavior – can be found in the Google Analytics tool.
Product analytics services such as Mixpanel.
…
No, there is no point in continuing the list, because even a first-grader knows this. Therefore, I propose to think about one observation that for some reason no one talks about.
Query Expansion
Any webmaster who knows how to analyze Yandex.Metrica statistics knows that over time, for certain queries, a sinusoidal graph is observed. That is, even if a page is firmly in the TOP for certain queries, traffic to it decreases and grows over time. This may be due not only to seasonality, but also to the expansion of the request itself.
You can be in the TOP of Yandex for bold queries, but this has no meaning if the topic expands.
The more (more popular) a certain query becomes, the more synonyms and related words users use when searching for information on a topic. The topic “cloud” itself expands and deepens. This means that webmasters have to study more data, analyze and structure information more carefully, and look for relationships between different indicators.
If in 2017 it was enough to write a guide “How to buy cryptocurrency”, get to the TOP and collect hot referrals to some exchange, now you will have to create a whole section on the site about how a beginner can get into crypto, which exchange is worth going to , which one is dangerous.
Important! It is worth understanding when a topic is “expanded” and when it is “inflated with air” due to search engine boosts by competitors.
How to get maximum statistics:
We collect statistics on low-frequency (or even single) queries from Metrica that led to our website.
We look for publications on sites on our topic and parse the comments under them.
We collect reviews about competitors on marketplaces if we work with goods.
Using AI, we transcribe popular videos on the topic and parse comments on them.
We select the most frequently used words and n-grams in each pool and in the general array.
We repeat the procedure every 2-3 months and monitor the emergence of new and narrow, but growing requests on the topic.
We implement all this into the structure of the site, descriptions of public pages on social networks, posts, hashtags, comments, video descriptions – in general, wherever possible.
And if you really want to go crazy, we set up an AI statistician who will look for some connections for us that are difficult to even guess.
Getting traffic just got easier
Today there are many more “easy” ways to get organic traffic. If previously everything depended on hard and careful work on the site, now for three kopecks you can search for Telegram posts, publications in Telegraph, articles in VK and in blogs of popular online magazines.
Interesting! A common mistake of managers is to find one, the most powerful, advertising tool and finally start sleeping peacefully. However, you understand that there are no such people anymore.
Теперь только в ритме танца…
Each marketing tool has its own limitations and now you have to turn on an internal combinator who selects the ideal balance between investments in different channels for attracting traffic.
What other ways are there to collect statistics?
What platforms are ideal for collecting information useful to a marketer? Those on which users can share their materials themselves. That is, they can publish themselves. This could be publications on social networks, reviews (you have to be careful here – there are a lot of fake ones) and data from negotiations/calls.
Example:
If we want to target our advertising to owners of marketing agencies, we analyze their social media accounts and personal blogs and try to understand “what language they speak.”
Then we go to the places where they live and begin to speak their language. It sounds simple, but behind this lies a colossal amount of work in collecting statistics, analyzing and synthesizing data.
From a marketing point of view, they have the greatest value words spoken by the man himself. By the way a potential client speaks, you can judge what he is thinking about.
To prevent the article from turning into a longread, I will write a continuation in the next one. I share other thoughts on the topic of analytics and online promotion in my channel. I will be glad to see your comments.
