When it comes down to measuring value or success we like numbers.
Choosing the right things to measure produces some good numbers that can define what is valuable and what is wasted time. Some numbers are worth a whole lot while others are more like pretty window dressing. There’s this whole book Measure What Matters that really breaks this kind of thing down.
The thing about numbers is that they are often parsed by people before anyone sees them.
This isn’t to say that numbers are fraudulent in nature, but when you see a pretty graph, someone went through the numbers and decided this was the pretty graph you’ll see.
Numbers don’t lie. They are recorded by people, who are not always perfect.
Two perfectly capable scientists can look at the same set of numbers and tell completely different stories. Propaganda is littered with this kind of thing. Almost every lobby group can find a scientist that can slice data up to say what they want.
I think people need some data literacy courses.
Stats class is a real thing and data analysis is an entire field of study
I went to John Abbott College and did Social Sciences with Psychology.
Because of the “with Psychology” part I had to take the following classes to get my DEC (the weird Quebec CEGEP degree, it’s a whole other article):
Quantitative Methods & Advanced Quantitative Methods - The math of stats
Research Methods - This one covered how to go to libraries and do research your peers will accept.
I can’t remember the name, but you either had to do a research summary paper or perform an experiment.
Most people choose to do the research paper summary. You read like 10-15 peer approved papers on a topic and write a report of your findings. I opted to go the other route.
I ended up working with people who were data analysts professionally. They made reports and compiled information. It’s a whole science in and of itself.
I’m an amateur by comparison.
I ran a bootleg experiment in Quebec’s version of college
When it came time for that course where I did my experiment I got lucky.
My teacher was so excited he did half my project for me. We were on a mission to test altruism. I asked 100 people for money and recorded my results.
The hypothesis was that if people were given a reason for the money, they would be more likely to donate if they were male and less likely to donate if they were female. Previous research (my teacher found me unreleased papers and everything) told me to expect this.
Now my experiment was flawed as I asked for money to get coca cola out of a vending machine. But I asked 100 men, and 100 women for either 25 or 50 cents. I provided a reason sometimes, I did not in another case. I had a fat stack of papers and I had to collect signatures of consent to be used as data.
I’m no pro or anything, but I have seen and tasted what the real thing is like.
One day I started to get paid to optimize things
Looking at excel sheets full of numbers doesn’t scare me.
I understand how to organize data, categorize it and turn it into something actionable for the people who are too busy to do that.
It’s like cleaning a room, but with data.
As a corporate trainer I had to figure out what knowledge was useful and what knowledge was pointless. To accomplish this goal we analyzed contact data and put more emphasis on subjects that generated more calls. Data started to seep into my life.
Inevitably I had to work with knowledge base content on a mission to reduce client contacts. This required going through article performance data and improving the content until people gave it more thumbs up than thumbs down. So much of my life went into analyzing not only the performance, but the actual fiscal impact of the effort.
What I learned is that people will always choose the numbers we see publicly.
Data driven decision making is how you really make money
When I started my career my ego got in the way.
Inevitably the lesson that data tells you what people want was drilled into my thick skull.
Search data reveals what people want articles people need. The original choices that were made for a knowledge base were not always on point. Then we may have overcorrected in a human driven decision to be prepared for user problems.
The data inevitably revealed the right amount of content needed to achieve the best results.
Almost always, human intuition was proven to be kind off the mark of reality.
A lot of the time you are just doing your best with the information you have.
Your main goal is to try and find stats that when you improve them, the end result is more money (or savings).
It’s okay to be wrong at first.
Sometimes people like vain stats that don’t mean anything
When you work in music and you go viral, this is a good thing.
For local retail marketers, virality is a mixed bag. Virality is achieved when your content goes beyond its target market. That means if you are targeting local Montreal vendors and trying to show your audience is local, virality can blur your stats.
It’s a weird look to be a Montreal based page but have random European countries be in your top 5 countries.
What this tells people who can read the data is that your audience is not local. Your content can get reach, but the people who see the content will never buy from the local merchants. That makes your audience kind of useless.
When it comes to statistics, certain metrics are vain. They look amazing but don’t translate into sales or even engaged followers. It’s worth it to think about what the stats mean.
If getting a lot of views does not translate into a larger core community, views is not the metric you should be measuring.
Measure things your actions can influence, that also validate your goals are being achieved.
Learn about the relationships of stats
Most of the time stats are relative.
You have a baseline, which is your current average without any changes. Then you have your goals, these determine which stats matter.
If you are looking to sell, your stats are going to be things like:
The click through rates across your sales funnel.
The average revenue per order
The total sales volume
All your actions are going to be to improve the number of people who buy, and then increase the amount you sell with each order.
If you are looking to grow a community you instead will measure:
Community growth
Engagement rate
Retention rate
Your goal is to keep people in the community and to keep them talking. The more they interact and the longer they stay, the more you are winning.
That being said, take some time to think about the “research” you see.
Question the sources used in the data presented to you
I had a peer send me a Kendrick Lamar vs Drake poll, held by a YouTuber, who’s audience was going to skew Kendrick no matter what.
The YouTuber’s content attracted the kind of person who would be Team Kendrick.
That means the results of that survey prove diddly squat to me except that guys fanbase does not like Drake.
I saw about 7 screenshots of Hip Hop DX’s X.com poll on the same topic and it went back and forth as time went on.
When you see people talking about their data, take a look at where the data comes from. Sometimes you end up learning the study in question was like mine, some college shit where a few hundred people were surveyed. Maybe it’s a sample size of 20’000.
While representative in the 5 figure range, in isolation it’s not proof of anything.
Don’t worry folks, the kids already get how data works
Let’s end this one with some positive news.
While many struggle to wrap their heads around this data driven computerized world, I think the little ones coming into it will be just fine. While I don’t think they are necessarily math wizards by virtue of formal education, when it comes to statistics, this is something they know.
The reality is the social media era ushers in with it a basic requirement of data analysis.
You need to be able to read your own numbers and understand how to optimize and grow. For the first time in history, the knowledge to do that is a few clicks away, on the same platforms. For free.
I think that as time goes on the dumbest technophobes will have a strong enough internet literacy that the sensationalism of the Boomers will fade away into oblivion.
Or it will find something better than weirdos quoting bad stats.
Maybe I’m just optimistic.
Live Long and Prosper Everyone
Comentários