#Articles

Funnel optimization

Data has become the main source of business to find new opportunities in their products or services. Something that has been there for a while, now is fast growing due to the evolution of technology around it. There are very different uses of data, for example: strategic decision making, searching for new markets, creating specialized products, or optimize the user experience which is the focus we will get on this article.

Nowdays is possible to know the path of every user from their first digital touch point to the final acquisition of our product, recording all the actions of that users to achieve a goal inside our website or application, this path is known as a funnel.

What is a funnel

Understanding a funnel could have a significant learning curve, each funnel has variables that will result in a unique behavior: users, product, media and traffic channels, communication and marketing, temporality and available information. It is not the same to sell loans than insurance, smartphones than cars, not even the same launch a campaign in January than December, or maybe users that came from Facebook use to act different to user who came from a Google search.

Let’s go from the very basic to the most complex so that we can see the potential of a funnel.

At a simplest point, a funnel is the path or steps that users will take to complete a task. Buying a product, asking for reports, renting a bus, or perhaps reading an article, all this actions are in a funnel. Tagging our flow we can know how many users go from one stage to another while they are browsing or trying to perform an action. A typical funnel would look like this:

But what if we segment users according the device they use to navigate. Or if we start to track users who arrived through an organic channel as SEO behave paid media. Here we begin to have a more detailed funnel that allow us to have more concrete findings.

For example, it is very different how users act if they have a strong motivation (desire), or if they already know our offer (interest), than just being informed for the first time (awareness). Also, understanding what is the segmentation and marketing strategy is essential for the analysis, for example it is normal for the funnel to fall at times when the user has to make an important decision such as “ accept a quote ”or“ grant personal data ”, but these“ falls ”are related to the type of segment that the funnel is going through the funnel, so when the most interested users are making the process the falls in those steps will be less than when users are there just to know the offer.

Analyzing a funnel

Analyzing the data is only the beginning of a design process aimed at optimizing a journey. The data will serve for: identify what is happening, generate hypotheses and measure the impact that our decisions will have on the funnel. But an optimization process requires more steps, very similar to the scientific method where observation is only the beginning.

It is important that quantitative analysis could be complemented with qualitative analysis, since a hard data does not faithfully represent the mental process of users and we do not know if their response is based on what they want or is only based on what we offered. Example: we have three types of packages for the user with different benefits, and one of those packages is chosen by 1% of users, using just the data perspective we could conclude that the package is not attractive to users, but what if the fact is that they are not understanding the benefits from it or some other element is getting all the attention.

WE NEED TO BE SURE.

Doing an experiment to test the hypotheses is a fundamental part of the process. Although in most cases they usually take form of A / B testings, we will only get a visible result in data, without understanding so much the cause of whatever is happening. We need more types of tests, not just sit and watch what people do, find the why. And there is when usability testings are more effective.

And like any experiment, is equally important to record what works well(to preserve it) and what does not (to stop trying to do that) and in both cases why is it happening. This is the point when the funnel analysis has a greater benefit than just optimizing to increase sales: a funnel is a source of learnings.

The funnel as a source of information

Just as scientists keep all their findings even if their experiments failed, it is important that that knowledge could be saved and shared to avoid the same mistakes and creating better strategies as more understanding we gained.

In recent months I have worked closely on funnels, and I have learned that the greatest benefit of data analysis and optimization, is learning. We can design in an iterative process indefinitely but that process would be useless if the lessons are lost. That is why I decided to start creating a blog where we could save the hypotheses that we are generating, and the results of these after the experiments. So after trying several tools (excel, paper cards), we conclude in a hypothesis journal using a blog tool, in which we record what we discover as funnel.

So, in conclusion, optimization should not be seen only on the side of the value we deliver to the customer, but on the how much effort costs to deliver it. Not only optimizing the journeys that the user goes through, but optimizing our processes so we can avoid the waste, and keep learning from a funnel becomes an additional but rewarding benefit in what we do. Isn’t funnels interesting?