The Seotistics newsletter is written by Marco Giordano, a Data/Web Analyst with the goal of combining business and web data. Tired of the usual boring Analytics content without any business impact? Seotistics teaches you how to use Analytics, web data and even content in your workflow while helping you with Strategy.
Share
π The (Web) Metrics Guide
Published about 6 hours agoΒ β’Β 7 min read
β
Use Data Or Be Used By Data!
The July 15 issue of Seotistics is here for you!
Metrics and KPIs are in everyone's mouth but many barely know how to make the most out of them.
That's because the word KPI has been degraded.
It doesn't help that most content online only covers the basics (e.g. traffic, revenue, that's it).
Please move this email to your Primary inbox or reply to it. This is to prevent Seotistics goes into spam by accident. Gmail users can read this tutorial to do it.
I am updating my Discord community to make it a reference for my readers.
Since social media have lower reach and emails can bounce, communities are the only way to ensure you get my message.
We do have a forum too now and I will try to organize my thoughts there.
β
Oh, and we restarted our weekly calls ;)
Before We Start
Let's clarify what everyone gets wrong.
β
β
In general, the following are almost always wrong:
SEO/PPC/Marketing KPIs
Only considering financial metrics
The only KPIs are related to business success. The others are metrics.
This doesn't imply metrics don't matter, they are just not as important as your key performance indicators.
So you still want to monitor traffic but unless you are a publisher, don't obsess over it.
Yes, creating new metrics for AI traffic can be "OK".
β But those won't be KPIs by definition, let's be real please.
The obsession with financial metrics is a double-edged sword.
Revenue alone doesn't tell you anything about why you were successful.
What People Do Incorrectly
Metrics are relegated to reporting and often without interaction between them.
"Users increased by 10% and that's good"
Statements like this are superficial and out of context because:
It can be bot traffic
It didn't affect the bottom line
The solution to this problem is thinking in terms of business and outcome.
Why are you measuring something?
Well, in Web Analytics it's often because you want to make more money.
This brings us to the next point: giving context to metrics.
And what's the best way if not visualizing them?
β
Metric Trees & Indicators
Metric trees allow you to visualize and conceptualize how different metrics affect your outcome.
The outcome is your North Star metric, how you measure success in the long term for a business.
We've covered a similar concept before:
β
Leading indicators are your inputs, your decisions and the levers you pull to affect your outcome.
The outcome is also called the lagging indicator (revenue in the example above).
The assumption is that you can consider businesses as a set of inputs and outputs.
While this is a gross simplification, it works for many use cases!
Now, let's see a proper example of a metric tree:
β
The North Star (Revenue) is the outcome/lagging indicator.
The components are how you calculate the parent metric.
# Clients and Amount Invoiced are components of Revenue...β and Users and Conversion Rate get you the # Clients.
Influences are factors that affect a given metric, like seasonality.
These can be correlated to your metrics and often account for uncertainty, external factors and "intangible" stuff.
Despite what many claim, websites have the usual business models and metrics.
True, every business is different...
but not really.
Churn is an issue for all B2B SaaS and having a good margin is essential in Ecommerce.
Publishers rely on traffic and like ads money.
There are exceptions but in 99% of the cases, we already know the business models.
Expanding On Metric Trees
Metric trees are helpful companions for visualizing how a business works.
They are rarely static in reality and you will have to change them once in a while.
For example, you may notice that you are missing an obscure influence you forgot about.
Once you have created your nice tree, it's time to ask the right questions.
β
It's not as hard as you see, consider this as a mental framework to exercise your brain muscles.
The DuPont Matrix
If you have a background in business administration like me, you will recognize that this concept isn't exactly new.
In 1919, this model was created to break down ROE (Return on Equity).
Documenting Your Work
From this, I recommend building a dictionary with your metrics and other information:
β
Here's an even better example:
β
The calculation field is necessary to make it practical. β You can get more creative than what I showed and even explain which dashboards or use cases rely on those metrics. β The Breakdown field tells you which dimensions to group by.
Skipping this step will give you future headaches and increase complexity.
Once you are comfortable with a method, it will become much easier to work on projects...
and you will get more confident in what you do.
Remember that in some cases it's all filters.
E.g. Counting key events means you filter by event_name and then simply count the rows.
You apply a filter and simply count, that's it.
Metric Trees To Products & Context
Yes, you can use our nice findings to build dashboards!
Many of them focus on the what (e.g. total clicks, total users) and not so much on the levers.
The levers represent the reason you get to the outcome, as already discussed.
This is a nice mental shortcut to make your dashboards more useful.
For example, a nice thing to keep track of in SaaS is the number of new users.
Why? Because you should acquire new ones if you want to expand!
Avoid committing the fatal mistake of reporting on metrics without context:
β
Social media show many great examples of hype building that lacks important context.
How many times did you see posts bragging about crazy traffic or spectacular revenue (w/o a mention of the profit margin lol)?
β
And if you recall, I also provided more specific examples for Web Data, like:
β
Context can also be referred to as "Knowledge" if you recall the pyramid model I used some weeks ago.
It's the step before you take action!
After you have the context, you need to do something or you will have a list of interesting facts.
(And trust me, you don't want that or you lose the project).
β
Actual Knowledge and Wisdom don't happen in days but even years.
There are examples of companies that actually use data to their advantage and even more so with Web.
Think of all the aggregator websites or marketplaces.
They heavily rely on data to make decisions.
Other businesses aren't that sophisticated, think of a traditional B2B company.
In those cases you can actually learn something new and faces all the real data challenges.
Old VS New Mindset
The outdated mindset (which is still popular in Marketing) focuses on the tool.
You will hardly see these topics discussed.
That's because many agencies and practitioners rush to implementation and forget about what matters.
If you want to work less and have more impact...
you need to target what matters the most.
As I always say, domain experts with technical knowledge win.
β
If you are a marketer and have learned enough analytics, you will be more useful than a random analyst.
That's because you know where to look at the data and what drives value, there is no middleman!
Now the Web Analytics field is changing and is getting more oriented toward what people do in data in general...
but those won't make you special.
Some Python and Google Cloud knowledge isn't enough if you get the essentials wrong.
This is why I created "Think Like a Web Analyst", which is also under a 30% discount with code SUMMER25:
The Seotistics newsletter is written by Marco Giordano, a Data/Web Analyst with the goal of combining business and web data. Tired of the usual boring Analytics content without any business impact? Seotistics teaches you how to use Analytics, web data and even content in your workflow while helping you with Strategy.
Read more from Seotistics - Web Analytics + Business + Strategy
30% off on all of my products until July 29 (CET), be quick! Use the code "SUMMER25" at checkout. Analytics for SEO - Course (1-to-1 on demand, full support and exclusive content + free updates) P.S. The course was updated recently and will receive more content this week. Learn Analytics For SEO Now! If you are tired of being puzzled by Web data, I can help you: β Python notebooks and SQL code to learn the hard skills β SEO Processes and Examples to convince stakeholders and become actionable...
Use Data Or Be Used By Data! The July 7 issue of Seotistics is here for you! LLMs, AI, this and that, there are too many buzzwords but not much action. Was content actually affected? Yes and no. I will tell you exactly why and what you can do instead of reading another case study. Please move this email to your Primary inbox or reply to it. This is to prevent Seotistics goes into spam by accident. Gmail users can read this tutorial to do it. Read this in your browser π£π£ Important...
Use Data Or Be Used By Data! The June 30 issue of Seotistics is here for you! An Analyst is a superior thinker and a person who can dissect topics with precision. Some of my students and readers alike commit the usual mistakes that can be easily avoided. We will take some of the concepts from my last issue to see why. β This Seotistics issue may inflict scalds and burns, handle with caution. Please move this email to your Primary inbox or reply to it. This is to prevent Seotistics goes into...