A Beginner's Guide on Using Data to Assess Business Performance

September 1, 2015

Running an online business that’s growing slower than projected is never an ideal scenario. What can tremendously help diagnose the problem is to have data and know how to gain insights from it. It is only through the collection and analysis of data where you can free yourself from guesswork, start validating assumptions, and gain insights on how you should be operating your business.

But there’s a caveat: data that wasn’t measured is data that’s lost forever [1]. Things will seem hopeless if you haven’t been collecting data all the while. It’s not too late. You can still turn outcomes in your favor by knowing exactly what data you need to collect moving forward. Read on and I’ll show you how.

Discovering new efficiencies and identifying root causes of problems requires collecting the right data from both a conceptually comprehensive (the whole funnel or customer lifecycle) and a rigorously detailed (what and how to measure) perspective.

Where are your visitors coming from?

There’s a saying that traffic is the cure-all. Time and time again, I’ve found companies stumble on a lack of traffic as the root of many of their problems.

High-quality traffic needs to be earned, not bought.

It’s very tempting to spend money right now to boost your traffic. But if your site isn’t optimized and you don’t have a strong retention program, the only thing that your marketing budget can do is to temporarily increase your site’s visitors or one-off sales. This isn’t sustainable in the long term when you’ve generated maximum reach in your target market due to perpetual spending on marketing. High-quality traffic needs to be earned, not bought.

I can’t stress this enough, but avoid being impulsive in spending your marketing budget to quickly increase your traffic if your site isn’t optimized (technically, about 2% conversion from visit to purchase). More importantly, if you’re not even collecting data, cut marketing spend as much as possible since there’s no way to tell whether your site is optimized or not.

Conversion Rates: What Keeps You Up At Night

Conversion is defined by two things:

  1. A situation that a visitor is in. For example: landing on a webpage containing all the features your product or service is capable of doing.
  2. An action they’re expected to make under that situation. For example: signup for a free trial of your product.

And here are some common examples of conversion:

Now let’s talk about conversion rate. It’s defined as the percentage of people who converted among those who were given a certain situation or offering on your site. Conversion rates are what makes or breaks businesses. You may have a million monthly visitors but only four thousand monthly paying customers [2]. The conversion from visit to purchase is 0.4%, which is generally a very bad conversion rate.

Low conversion rates keep the top business executives up at night. The reason being there are many variables at play that contribute to an increase or decrease in conversions. Some examples are seasonality, the quality of traffic, assortment, user experience, navigability, promotions, and customer service. Adding to the complexity of figuring out the causes of conversion rates is how these variables affect each other [3].

There are situations where improving the conversion rate in one step of the funnel causes a decrease in conversion rate at another step of the funnel. For example, you may be optimizing your conversion rate from free trial to purchase by asking people to input their credit card information upon signup. But doing so can add friction to the ease of creating an account for a free trial. As a result, you’ll probably see a decrease in your signup conversion rate.

If you’re constantly making changes in the hopes of optimizing conversion rates, the key is to monitor the overall effect of the fluctuations across affected steps of the funnel. Say your funnel involves three steps: A, B, and C. Below is a simplified example of the results of a split test where you tried to optimize conversion from steps A to B.

Current Version

Experimental Version (optimized A to B)

At first glance, it seems reasonable to go with the experimental version since the 40% lift in conversion from A to B (50% to 90%) have “outweighed” the 30% decrease in conversion from B to C (50% to 20%). But it’s the conversion rate from A to C that should ultimately be considered. The current version’s A-C conversion is (100% x 50% x 50%) = 25%. While the experimental version’s A-C conversion is (100% x 90% x 20%) = 18%. Thus, the current version is still better than the experimental one.

How effective are your landing pages?

A landing page is the first page viewed for any given visitor. The optimal setup is to have several landing pages, each suited to welcome incoming traffic driven by marketing campaigns targeting specific demographics. Landing pages usually get the biggest share of traffic across all of your site’s pages. They’re usually located at the start of the funnel and represent the last point of contact for marketing. From then on, it’s already the function of the site that’s responsible on delivering what was written and promised on the marketing content.

The most important conversion rate to monitor on a landing page is its bounce rate. The bounce rate of a page represents the percentage of visits where the first page viewed is that page and it was the only page viewed by that visit. In simpler terms, the bounce rate of a page is the percentage of single-page visits on that page.

Bounce rates are important because it gives us a measure of interest (or disinterest) on your landing pages. Ideally, your landing pages should be able to address two things:

  1. Capture the interest of people who’ve never heard of your site before
  2. Give returning visitors useful content to consume [4].

It’s all about making people who just entered or returned to your site react positively to your offering. If they’re interested, they go to another part of your site to know more about your offer, and they won’t count as a bounce. The bounce rate is a simple metric to start with in order to quantify engagement at the start of the funnel.

How engaged are your visitors?

Engagement is all about the degree in which people who come to your site become convinced of your calls to action, love your content, share it with others, explore other parts of your site on their own, contribute their own content, or constantly use your service.

How do you measure engagement? I’ll give you some examples:

An often overlooked mindset that online businesses make when it comes to figuring out how to measure engagement is to relate those metrics to revenue. Engagement should ultimately be a measure of the propensity of your visitors to eventually perform actions that contribute to your revenue. You shouldn’t be happy with a highly engaged audience if you’re not really converting enough of them towards revenue. We’re not here to play the vanity metrics game.

A million people visit an online retail site. No one buys anything. Press declares retail site a massive success!

Every online business must achieve profitability at some point, and be sustainable moving forward. Which brings us to the most important metric: revenue.

What user actions give you revenue?

We’ve finally come to the part of the funnel that’s the lifeblood of every business. This is where we can fully encourage activities that drive revenue. When you’re only starting out, you need to monitor aspects of your site that you think may encourage visitors to take action towards a purchase, i.e. conversion to revenue [5].

For e-commerce businesses in particular, conversion to revenue means anyone who checks out and pays for the items in her cart. For SaaS businesses, it’s a user who subscribes to a paid plan, along with paid add-ons or enhancements to the plan.

It’s possible to drive revenue at the start of the funnel and it seems intuitive to do so because landing pages get the most traffic. But this is far from ideal. Forcing visitors to put out their wallets immediately after landing on your site will probably be a disaster. Your best shot in converting visitors to buyers is to condition them or guide them towards making a purchase, i.e. a purchase that they’re willing to make.

Make every conversion to revenue the result of an informed decision.

This may seem contrary to typical sales advice wherein the more you hustle, the more deals you close. But do what you can in your power to make every conversion to revenue the result of an informed decision. Giving your users enough guidance for them to be willing to pay is to your advantage. First, it builds brand loyalty and awareness. You’ll get really happy customers who would be willing to tell the good news to others. Second, they’ll have you on top of their minds when they need to purchase the same product again. Third, you can create mechanisms to receive very helpful feedback from visitors who had a pleasant experience but in the end still refrained from making a purchase.

It’s a wrap!

So there goes your overview! The items above aren’t exhaustive, but should help you think about data and how your business as-a-whole is performing in terms of actual events happening on-site.

I hope this article gave you a sense of some key metrics that you should be measuring, and will encourage your business to discuss further what other data points need to be collected. My hope is that such data should provide a richer picture of how your business is performing and help you become more data-driven in coming up with business strategies.


[1] These are mostly time-based or time-sensitive data that needs to be collected at the moment of the event. On a lighter note, there are types of the data that weren’t measured, but can still be inferred from currently available data.

[2] Normally, customer acquisition cost, churn, and projected lifetime value are also taken into account.

[3] For data scientists, monitoring correlations among conversion rates across the funnel is a pastime activity.

[4] Even for SaaS, returning visitors to your landing pages are mostly people who are still assessing whether they’ll use your software or not. Active or signed up users of your product will more or less be already logged into your app’s dashboard, or will enter your site on the login page. Even if they enter your site via your landing page just to get to your login page, they won’t count as bounce.

[5] In general, it’s any user’s actions that contribute to revenue. Some online businesses, like high-traffic content sites, do not depend on purchases for revenue. They generate revenue by selling ad space on-site and when readers click on their ads. For our purposes, we’ll use the term “buyer” to denote a user whose actions generated revenue for your business.

A Beginner's Guide on Using Data to Assess Business Performance - September 1, 2015 - Adler Santos