Be a "data geek"

This is a guest post from Joca Torres. Joca is the director of product development and product management at Locaweb, Brazil’s leader in web hosting, cloud servers and SaaS applications like email marketing and online stores, serving more than 250,000 customers. His first startup experience was in the early 1990s when he founded and ran one of the first Brazilian ISPs. He has been working with internet related software ever since.

Joca is also the author of The Startup Guide: how startups and established companies can create and manage profitable web products. The book is in Portuguese, but Joca has been kind enough to translate a few sections of it for us. This is Joca’s fourth guest post in this series. Joca also posted all of the book’s content on the “Guia da Startup” blog (in Portuguese).

—————————————————————————————————————————————

During David Cancel’s 2011 “Data Driven Business” he mentioned how important it is to know your numbers. He even mentioned that “In Data We Trust” paraphrasing US official motto “In God we trust”. His talk inspired some chapters of my book “The Startup Guide” (in Portuguese) that I’ll translate below.

What data is important?

All data is important, but depending on what you’re looking to understand, certain data is more important than other. Knowing your data is an ongoing task as each new knowledge you acquire from data will motivate new questions that will need more data or a different perspective on the existing data to be answered.

The first information that you will want to know is how many visits you receive on your product web site. To know these numbers you can use some statistics report included in the hosting package of your hosting provider. Another option is the well known Google Analytics.

With a report like this you get some important information such as number of visits, number of unique visitors, page views and much more. Depending on the system that you are using, you may also see the first and last pages visited during an access to your site, which country and city your visitors come from, if your visitors have accessed your site coming from a campaign you are running on Google AdWords, Facebook or somewhere else, or if they found your site organically, or by directly typing the address, or searching for something in a search engine. It is important to remember that you should have reports not only for your web site but also for your web application.

But be careful! These report systems normally provide a huge amount of information and it is easy to get lost in this sea of data.

Along with the amount of views and hits your site has, there are other important data that you need to know about your web product:

• Number of people who learn that your web product exists: It is possible to differentiate the ways people learn that your web product exists classifying them into two categories, paid and organic. Paid are those people who learn about the existence of your web product because you have invested some money to inform them. You can invest money in Google AdWords, Facebook ads, ads on content sites, preferably sites with content related to the subject of your web product, magazine ads, also preferably magazines related to the subject of your web product. On the other hand, organic are those people who learn about the existence of your web product because you have invested time and effort to become known by creating relevant content on the topic of your site, interacting on blogs that talks about the topic of your product, making it easier for people to recommend your product to their friends and so on. The return in this case takes longer, but has the advantage of having no financial cost.
• Number of clicks generated by your ads or by other means: this information is a little bit more difficult to obtain because depending on your strategy to attract people to your site, this information is not available. The online ad systems like Google AdWords, Facebook and ads on content sites usually have this information available and the price they charge you is normally based on a price per click.
• Number of unique visitors: this is the number of new visitors your site receives. It is different from the number of visits since the same person can visit your site more than once before deciding to buy or to become a user.
• Number of visitors who become users: from the total of unique visitors, some will become a user of your system. If you offer a free trial or free version without an expiration date, this number may be fairly large.
• Number of users who become customers: at the end of the trial period, some of your users will want to become a customer by paying to continue to use your service. If you’re offering a free version of your product, with no expiration date, you must have a paid version that will encourage your users to upgrade from the free version to this paid version.
• Number of users and customers who cancelled: some users and customers may decide to no longer be your user or customer. It is important to know how many and the reasons why this happens since this information will provide you a lot of information to improve your web product.

Conversion funnel

Napoleon Bonaparte, French military and political leader known by the Napoleonic Wars, through which he was responsible for establishing the French hegemony over most of Europe in the early nineteenth century, had a major defeat in 1812, the Invasion of Russia. This invasion was a huge military operation brought by the French and their allies who had a major impact on the outcome of the Napoleonic Wars, marking the beginning of the decline of the First French Empire. In this invasion Napoleon used 580,000 combatants. Of these, only 22,000 survived, the rest perished on the way from France to Moscow due to difficulties such as cold, rain, rivers, and other natural and man generated obstacles.

This image resembles a conversion funnel of a web site that can be done with the data discussed above. The conversion funnel shows us how we’re losing potential customers in the way of attracting people to the site to the point where a person pays to become your customer:

The conversion funnel helps us see where we must work to get more paying customers. For a “data geek”, this should be the primary focus in terms of data to be obtained and analyzed. Each piece of this funnel requires a different type of action to be enlarged.

Example funnel

ContaCal is a web product I created in 2011 during nights and weekends with no connection with my day job in order to experiment with building a startup using the most recent methodologies and best practices. ContaCal is a calorie counter system with a twist. Besides telling you the amount of calories you’ve ingested it also tells you the quality of these calories. If you want to read more about it, you can read my previous posts. Below is an example of a conversion funnel based on ContaCal’s data.

Sorry for the Portuguese, but here’s the translated column names in the same order they appear above:

• clicks
• unique visitors
• unique visitors per clicks
• users
• users per unique visitor
• paying customers
• paying customers per users
• cancelled paying customers
• cancelled paying customers per paying customers

Note that this is a monthly funnel. It is important that you also have a daily funnel so you can measure everyday the impact of changes and experiences that you do to enlarge the funnel.

Below is another example of ContaCal’s daily funnel showing the daily cohort of users to paying customers. Cohort, as defined in Wikipedia, is a group of subjects who have shared a particular event together during a particular time span. I have a 5 day trial period, enough to test ContaCal, so I measure what happens with users since signup until the 5th day:

Again, sorry for the Portuguese, but here’s the translated column names in the same order they appear above:

• date
• # users
• # used in first day
• # used in second day
• # used in third day
• # used in fourth day
• # used in fifth day
• # became a paying customers

Next steps

In my next post I will talk about the long-term numbers we also have to track.