When we look at our Google Analytics traffic sources, we are often misled by the sheer volume of traffic coming from some of the popular domains- Facebook, Google, Reddit, and so on.
But most often marketers don't track the quality of that traffic beyond a certain point. And when they do, they most likely uncover some counter intuitive truth about their businesses.
In this post I'll go through a case study where I'll look into:
- Biggest traffic sources to an analytics-based SaaS business
- Demonstrate how the traffic sources that actually convert are different than the ones that bring massive traffic
- Explain the significance of non-converting channels (there is a reason why they exist)
- Suggest some corrective steps so that one can assess the user signals better
Biggest Traffic Sources Are Often Misleading
Here's the problem with a new online business- they don't have enough traffic data to understand traffic behavior. I experienced it first hand with some of my initial businesses.
Whenever I had my website ready with basic content, the first thing I'd do is to get Google Analytics set up to track where my users were coming from and what was their behavior. I'd also set goals in order to track users who reached a certain point- like opted for the newsletter or registered with their emails.
But the biggest problem with default settings in Google Analytics is that it can only tell you the websites that brought traffic to your pages, which pages got the maximum views, on-page metrics like bounce rates, time on page and traffic sources.
GA can't tell you which channels converted the most unless it's properly set up.
The biggest challenge for a marketer is to know which channels converted most and why?
Understanding the channel preference of users, picking the right channel with best ROI and matching users from a certain channel with the right landing pages are the most crucial aspect of success for any online business.
Without knowing that they can't focus their attention on the most optimal channel. And that's something Google Analytics can't tell us.
Traffic Sources That Actually Convert
Many new marketers suffer from shiny object syndrome when it comes to tracking audience. They often go for channels that can send them maximum traffic, hoping that all channels will have the same conversion rate which is the conversion rate for the entire business.
So for example, if there are 10,000 monthly visitors to our site and 500 of them sign up for a free trial, many marketers would like to assume that 5% is "the conversion rate" for all our visitors. Whereas in reality, every channel has its own conversion rate. And to the surprise of many, it can vary between 0.1% to 10% .
These are just a ballpark figures for majority of the sites and it may again be different for new websites or sites with less traffic.
So the fact is some channels convert 100 times more than others.
When we look at few layers under our top level traffic, we get a completely different picture of our traffic.
Our most loyal users may be coming from some obscure website which we ignore when it appears at the bottom of our traffic sources (unless Google Analytics glosses it over). For example, traffic from Duckduckgo might convert 10 times as much as Google.
But these are some nuggets that can save us years of effort and hundreds of thousands of dollars worth of marketing dollars and at times completely turn around a business from negative to positive ROI.
Channel strategy is the most important factor that determines the success of a business after product strategy. And understanding of the traffic patterns and other important characteristics of traffic plays a key role in the success of a channel.
For our case study, Plausible's top 5 traffic sources are:
- Hacker News
But the top 5 sources that resulted into maximum trial subscriptions are:
- Hacker News
From what we can find here, Facebook which is the 4th most important traffic source is not in the top 5 when it comes to sign-ups. And if you dig even further, it's not even in the top 10.
On the other hand, Github, which one would have completely ignored as it contributes just 6% of the traffic compared to the top source, Hacker News, comes to 3rd place when it comes to subscribers.
To add some context, Plausible is an analytics software that provides data about your website visitors with a lightweight code in a privacy-friendly manner. When you sign up for their free trial, they send you a code that needs to be installed in your website so that you can track your web traffic.
It means at least one of the following two conditions need to be fulfilled for someone to click on your link from social media:
- The person is interested in analytical data
- The person owns a website or app
- A person is just curious
The following needs to be fulfilled for them to sign up for a trial:
- They need a privacy friendly/light web analytics software
- They are able and willing to take steps to install the new software
The difference in the expectation level of the first 3 conditions and the next 2 conditions will determine the conversion rate. The higher the difference the lower the conversion.
For a platform like Facebook and Twitter, it's easy to get lots of people who are explorers. They might click on a product just because they are curious (satisfying point number 3 at the top of the funnel). But when it comes to installing a code (or take any action that takes more than a click) they'll most likely give up.
On the other hand, people from Google and Github are way more focused. People coming from Google are already searching for a solution, so they have a high level of intent. Plus Google sends you targeted visitors, and they already apply a match before sending anyone on your way.
People coming from Github have the technical knowledge and they are probably aware of the product and they are coming to the website to explore further. It will be interesting to see if they make a final purchase.
While deciding on a channel strategy it's always important to find channels similar to the ones that are already working in terms of sign-ups and not just traffic.
Now I'll bring another list that tells us about the top 5 sources by the percentage of conversion. And this is important because, if we are to focus our energy on just a few platforms, we should prioritize the ones with maximum conversion rate because they bring us the best quality users.
Top 5 sources by conversion rate:
Different Conversion Rates For Different Traffic Sources
Unfortunately, we don't have the data for final subscriptions (paid subscription) which I can bet would be very similar to the 3rd list.
Why There Is So Little Conversion From High Traffic Sources
To answer this question, you have to understand the nature of traffic coming from these sources and their motivation to visit the page.
When we post something in social media, people there are more prone to click something new and interesting. But when they visit a page that curiosity dies down.
As we keep adding more hurdles to this journey by asking them to fill out a form, install a code, complete a few steps in a process, they keep giving up on this curiosity. In this regard Twitter is little different from Facebook or Instagram as people will not click something just out of curiosity.
In Twitter, the discussions are much more contextual and you don't follow people just becuase they are "friends" but because you share similar interests with those people. This difference in user's inherent motivation is apparent in the conversion rates mentioned above.
On the other hand, people coming from github (and to some extent