What is funnel analysis? In a sentence: A powerful method of online data analysis which shows you the most important conversion steps of a user one by one.
Something like this:
How do funnel analytics work in practice? What are the biggest questions and roadblocks? How could you extract „actionable” knowledge from them? How can you create it for yourself? By reading this article you’ll get all the answers!
Why is Funnel analysis good? When should you use it?
Funnel Analysis is typically useful if you want to measure a linear usage-flow. Basically, it is about going step by step and calculating how many users reached a certain point of a given process. Or on the other hand: how many users churn out to this point? The process itself can be a simpler thing. Like filling out a registration form…
… or it can be a bit more complex, like an onboarding process (e.g. the graph on the top of this article shows the onboarding steps of a note-creator app.) Eventually you can measure your whole product from the first page visit to the point of purchase and more…
In each case the keyword is linearity. With funnel analysis we try to model processes where users can’t skip through certain steps (those strictly follow one another), and where they can reach their goal on a straightforward path. On an e-commerce website for instance this path can be set up easily:
- Landing page.
- Browsing through the products.
- Visit a specific product.
- Place the specific product in the Cart.
- Filling out the purchase details.
- Placing the order.
- „Thank you for your purchase!”
IMPORTANT: Funnel analytics react to the experience of the users. We measure their development. So even if the user browsed many different products, we count it as one. Always think in “level-ups”: so if someone placed a product in their basket, tick the 4th step, and the goal will then be to get them to the 5th step. If someone puts 80 products in their basket, they still only get one tick for the 4th step. (The number of purchased products will be measured on other metrics anyway.)
SaaS (Software as a Service) and UGC (User Generated Content) websites are roughly as simple as e-commerce. The situation gets more complex with media pages, which is why we rarely use the funnel analysis with those.
A basic funnel analytics model to start with
For a start I would recommend the most well-known funnel model. Based on Dave McClure’s AARRR model a user goes through 5 steps when using your online product or webpage:
1. They land at the page
2. They start using the product
3. They return to the page
4. They make a purchase
5. They refer the page to friends
As you can see, at Dave McClure’s model linearity is not 100% (for example, someone can recommend you without making a purchase), but in most cases this will still be the real order.
How to create your first funnel analysis?
Step #1: Define the steps of your funnel!
For ease of understanding I would suggest making between 5 to 10 steps. If you have a lot more, you risk getting lost in the data, if you have much less, then you don’t have a funnel. 🙂
(Plus: you can always create sub-funnels at any point in time. So e.g. you have a main funnel overarching your whole website where one step is registering, then the process of registration can have a separated sub-funnel. This logically will be built into the main funnel, but on a visualization level it’s worth separating them, so you don’t get distracted.)
Step #2: Pick your funnel analysis tool(s)!
It can be literally any software. If you want to go with an easier setup process, you can use smart tools like:
- Google Analytics (eg. by the goal settings, or by enhanced ecommerce settings)
- Mixpanel (they have some built in functions for that)
- Hotjar (they also have built in funnel metrics)
And you can always analyze funnels with your own tools. Even Excel or Google Spreadsheet can do the job, but if you are on an advanced data coding tool like R, SQL, Python or bash, you can use some built-in packages or you can build your charts with some 3rd party data visualization tool like Tableau, GoodData, Chartio or anything, that fits for you.
Step #3: Visualize your funnel!
Once you are done with all of these above, choose the simplest, most easy to understand visualization. So NOT this:
And not this:
And only if it’s a must, then this:
The simplest and the best visualization for funnel analytics is the classy bar chart. It’s no accident that I picked this as an example above.
I highly recommend to use bar charts and nothing else for visualizing funnels!
How can you get actionable insights from funnel analysis?
This is a key question. Each online research and data analytics is done so we can make a change or improvement!
But how can we get useful information from a Funnel analysis? There are 3 major ways.
One is the so-called bottle-neck-check. Here we are looking for where the chart drops the most. If we see that during the registration everyone provides their name, email address and password, but almost everyone disappears at captcha, then we can suspect that there is a problem there (e.g. the captcha is not readable). But of course it’s not always this easy. There are steps where we actually expect our users to churn. E.g. entering bank card details during check-out comes always with a big drop. So we should not search for the „bottle neck” where most users disappear in absolute numbers, but where the drop out is the highest relatively – compared to our expectations. (Our expectations could be based on many things – we can use our „common sense”, but using market benchmarks or past data is more advised).
Another favourite metrics of mine is the time-delay between the certain funnel steps. How much time it takes for someone to put a product in the cart, starting from the first view on it… This tells us a lot about our product and our communication as well. The time it takes to purchase a $2000 laptop will never be as quick as buying a $2 battery.
The third recommended funnel analytics practice is segmentation. If we can find user-segments who are more successful at certain steps than others, then we receive great feedback on who to target in the future, as well as on why other users stuck. For example, if we can see that men buy female jewelries without a problem, and a large portion of women eventually don’t buy any, then it may be worthwhile to think about our main message to target men to be „buy your wife some jewelries”.
“But what can you do with your Funnel? Definitely not starting to heal the top of the Funnel so more people can come in through there. It’s not necessarily the best solution if you begin to fill the largest hole between two steps. I think the best option is if we begin our work at the bottom of the Funnel. Because if you begin to manically pack people to the top of the Funnel (e.g. with Google AdWords or Facebook Ads), those will drop out anyway. And those you load to the top and drop out will never come back. That’s a wasted User.
So it’s best to spend your time on those we know love us and have tried many of our products. Let’s see what can help them and heal the bottom of the Funnel for them. You don’t want to work with those who come and just take a peek at your product. So you gradually fix your Funnel upwards, and when it has reached a certain „thickness” where you say okay, this works, then you can start working on larger marketing costs and other good ideas. And bringing in the Users.”
- Funnel analysis is a nice tool with which you can measure the life cycle and development of your users steps by step.
- Define 5-10 steps and visualize your funnel in an easy to understand way!
- Analyze your funnel: Check at which step most people drop out at, and which segment is the strongest! Good Luck!
And if you want to be notified first about new content on data36 blog (like articles, videos, handbooks, etc.), sign up for the Newsletter!
1. According to Tamás Geiger : In GA premium there is a “custom funnels” feature as well!
2. According to Tamás Lindwurm: “if we want to broaden our funnel then it’s better to begin at the bottom (with the most committed clients) and moving up from there”.