Once you’re starting to get consistent traffic on your site and landing pages, you’ll want to start looking at ways to optimize conversion funnel performance.
It’s great to have a lot of people visiting your website, but it’s even better when they convert.
Your website’s conversion funnel will be impacted by:
- Your traffic acquisition sources: Are you acquiring more high-intent bottom-of-the-funnel (BOFU) traffic or more low-intent top-of-the-funnel (TOFU ?) traffic?
- A site objective: Are you optimizing for sales, signups, leads, demo requests or brand awareness? Different objectives will lead to different kinds of funnels and conversion rate windows.
- The structure of your website: Is your site a single landing page? Is the conversion flow sequential? Is the structure hierarchical? Or does it have a network structure?
In this post, we look at how to map and monitor your conversion funnel to optimize it.
Understanding your Site Structure
Sequential sites are the easiest to optimize because the experience is the conversion funnel.
The more dynamic the website is, the harder it is to hone in on a single funnel to optimize (e.g. A > B > C > ?).
Most hierarchical and network-structured sites will have both planned and unplanned conversion funnels.
Google Analytics’s Users Flow functionality (Audience > Users Flow) is a good way to uncover unplanned conversion funnels, but first you must know what it is you’re trying to optimize for.
Optimize Conversion Funnel Performance, but for what?
To focus your marketing efforts improving the performance of your conversion funnel, you have to know your business goal.
Whether it’s sales, signups or leads, you can define the key steps in your customer journey by looking backward from it. In fact, a lot analytics tools can help you understand the most common paths to reaching your conversion goals.
For example, at Highlights our main goal with the marketing website is signups. We get a lot of our traffic from search engines and through content promotion. This means that a common path to signups looks like this:
- Visitor views a blog post on social media or on a website;
- He/she reads the post and decides to visit our homepage;
- After careful investigation, he/she decides to signup;
- From the signup screen, he/she decides to go through the registration;
- Voilà! ?
To optimize conversion funnel performance, we need to understand and monitor all of these micro-goals to see how they’re helping (or hurting) the over-arching goal (signups).
How to Monitor Conversion Funnel Performance
The best way to track conversion funnel performance is by using investor Dave McClure’s A.A.R.R.R framework (Originally from Startup Metrics for Pirates).
There are five steps to his framework:
- Acquisition: Visitors entering the funnel from various channels;
- Activation: Visitors enjoying a first experience with the app/site;
- Retention: Visitors coming back and becoming retained;
- Revenue: Users purchasing the product and becoming customers;
- Referral*: Visitors or buyers referring the site or product to friends and colleagues.
* I suggest inverting the Revenue and Referral phases.
Although we’re optimizing for a specific objective, it’s a good idea to track the entirety of the customer journey from Acquisition all the way to Referral.
By doing so, and evaluating the performance of cohorts – a group of people who share a common characteristic over a certain period of time – you’ll be able to find the break points.
For example, at Highlights, the main signup flow looks a bit like this:
The question that this cohort answers is: Of the people that visited the blog (1,000), how many visited the homepage (150), clicked sign up (25) and actually completed the registration (17).
By looking at the drop from one step to the next and comparing data to benchmarks, you can find the bottlenecks and identify opportunities to optimize conversion funnel performance.
Although you can track these funnels in several analytics tools when the data is available, I like to use Google Sheets to visualize the entire funnel performance:
Deciding on an Evaluation Window
Startups with weekly goals and metrics outperform other startups.
– Mark MacLeod, Investor
With conversion funnel optimization, there’s always a balance to be struck between speed (velocity) and data accuracy (statistical significance).
Depending on whether you’re in the Good Enough, Head Above Water or Accuracy phase of your business, you’ll certainly want to aim for different levels.
For example, in the early days of a business when there are a lot of things that need to be figured out, velocity is often more important than accuracy. In that case, it’s more important to be directionally accurate than precisely accurate.
That said, we wouldn’t recommend doing the same thing when the business is at maturity.
At Highlights, we’re on 3-week learning / optimization sprints for our funnel performance. It’s the best balance we found in terms of data accuracy, velocity and resource availability (someone’s gotta make the changes!).
Regardless, we look at performance every week:
This helps us identify focus areas and opportunities for improvements.
How to Optimize Conversion Funnel Performance
Once you’ve identified a breakpoint in your conversion funnel performance, you’ll want to analyze it from various angles:
- Do users from different traffic sources behave differently?
- Are there countries, regions and cities performing better than others?
- Can users on mobile, tablet or desktop complete the same tasks?
- Are there considerations in terms of site performance?
- Are there issues with payment or anything else you’ve not foreseen?
This will help you create hypotheses for improvements and experiments.
We’ll cover how to run efficient conversion funnel optimization experiments in another post. Sign up to our newsletter to make sure you don’t miss out.
Once you’re starting to see sufficient improvement to the core metric you’re working on, you can move on the next break point. By following this process, you’ll be able to continuously grow the performance of your conversion funnel.