Analytics fundamentals

Analytics is about tracking the metrics that are critical to your business. While business as a whole needs to track lots of different metrics (e.g. human resource cost, revenue per employee etc), in this course we only talk about analytics in the context of conversion optimization.

If you drive traffic via Google Ads to a landing page where they opt-in to your email list, and eventually buy something – the key metrics for you would be cost per lead (how much Google Ads dollars you spend to acquire a lead), number of leads and revenue you make out of those leads. These metrics combined tell you the effectiveness of your customer acquisition strategies, and the financial health of your business.

In the end it’s about how much money you’re making. My friend Brian Massey likes to say that if you want to know how your business is doing, ask your accountant. While that’s true in the big picture sense, you need to be able to look at your web analytics suite (e.g. Google Analytics) and have a clear picture how you’re doing.

You need to measure Key Performance Indicators (KPIs)

Your website has a business objective – to get (profitable) sales and/or leads. Every objective can have several goals – specific strategies you’ll leverage to accomplish your business objectives (e.g. reduce bounce rate, increase email opt-ins etc). You need to make sure your goals are DUMB (as per Avinash Kaushik): Doable, understandable, manageable, beneficial.

Configure your web analytics tools to measure each of those goals. Besides measuring simple goals, you also absolutely need to measure your Key Performance Indicators (KPIs).

What are key performance indicators? Measures that help you understand how you are doing against your objectives (e.g. getting people to buy from your online store).

For websites your top KPIs would typically be

  • Conversion rate (CR – the measure of your site’s effectiveness in persuading your visitors to take a desired action). 
  • Revenue per visitor (long-term view, how much money are you making per each unique person in a year). The strategy for improving this KPI is to attract more valuable visitors. Use this to critically examine each new visitor acquisition effort, and segment as necessary to identify which ones are performing.
  • Revenue per visit (RPV – on average, how much revenue you can expect to generate off of each incoming visit). Similar to the previous performance indicator, it’s a good indicator of how you’re doing right now in your marketing and conversion efforts. Compare revenue per visit to average revenue per visitor to see if your short-term efforts are paying off, but not really contributing to the lifetime value of a visitor.
  • Average order size (Sum of Revenue Generated / Number of Orders Taken). This critical measure for ecommerce businesses helps you optimize the amount of revenue you generate per transaction, and helps you optimize for higher order sums. You can also identify where the high-value customers come from, and how their behavior is different. Segmentation is key here (audience, traffic channels, campaigns etc).
  • Average items per cart completed (Sum of Products Purchased / Number of Completed Shopping Carts). One of the best strategies to increase average order value is getting customers to buy more items each time they purchase. By paying attention to this, you can optimize for it.
  • Checkout abandonment rate (the percentage of people who start checking out to those who complete that process). This KPI is important since it’s easiest to get money from the people who have already decided  (or considered) to give it to you.

Note that this is a non-exhaustive list, and different businesses need different KPIs. SaaS companies need to measure Revenue per User, and businesses that drive paid traffic need to pay attention to Cost per Conversion and Cost per Lead. Some might want to measure Average Visits Prior to Conversion and Percent New and Returning Visitors.

Think critically about your business model, and the type of KPIs you need to track.

When you measure them, you get metrics. Metrics are numbers – either counts (a total – like “revenue last month”) or ratios (e.g. bounce rate).

Metrics to pay attention to besides KPIs

  • Bounce rate. It shows the percentage of people who leave a page without a single click. Note that this is not a KPI, but it can offer insights. It helps you find campaigns and landing pages that need improvement. Ignore site-wide bounce rate, look at it page-by-page. 
  • Exits and exit rate. These metrics help you identify pages where most people leave your site from, so you can improve them.
  • Time to purchase. This metric helps you understand how quickly or slowly your visitors convert. Convert them at a pace they are most comfortable with. The cheaper and/or simpler the product is, the less time people usually need. For more expensive products or services (e.g. motorcycles or liposuction), people might need more time to think / do research. Adjust your visitor path through your site, copy and landing pages accordingly. If the Time to purchase is much longer, then you might want to use drip content email campaigns in your conversion strategy.
  • Assisted conversions. How many of your conversions had more than one channel / ad / marketing touch prior to converting? When you see which traffic sources convert the best, it’s not telling you the full story. Just because organic search click was the last one before conversion, it doesn’t mean it should get all the credit. They might have come via social media first. Aim to optimize for a portfolio of channels.

Business Objectives -> Goals -> KPIs -> Metrics -> knowledge we can act on.

Identifying bad metrics

Metrics are there to provide actionable insight. You need to be able to look at a metric, ask “so what”? – and have an answer.

  • “Our top ‘exit’ pages are blog posts”. So what? We need to do a better job directing blog traffic elsewhere after reading the post.
  • “Conversion rates for our top AdWords keywords are way up”. So what? We should increase our AdWords budget.

Bad examples:

  • “Our average time on site is up by 30 seconds!” So what? Em…err.. no idea.
  • “Our total traffic is up 15% from last month!” So what?

Averages lie, you need to go deeper

Web analytics reports are full of averages.

But in real life no one is “average” and no user experience is “average”. They’re often talked about since they are an easy way to aggregate information so that others can see it more easily. But they offer little to no insight.

A familiar sight:

avg

Always keep in mind that when looking at a metric, ask yourself, “what will I do differently based on this information?” So now when you look at the average visit duration, do you know any better what to do next? Any brilliant insights? Didn’t think so.

What should you do instead?

#1 Look at the data per segments.

Average revenue per visit is an important KPI for ecommerce stores:

per-visit-value-avg

But it’s more insightful if you look at it per traffic source, you get more context:

per-visit-value

#2 Look at distributions

Distributions show what makes up the average and look at the numbers in a much more manageable way. All web analytics suites show average pageviews per visit:

pageviews-per-visit

Are more pageviews better? Probably? Well, what about looking at it from the perspective of conversions:

depth-conversions

There’s quite a clear trend that more pageviews lead to higher conversions, thus increasing pageviews per visit would probably be useful.

Distributions will also be insightful in the case of “totals”. So instead of just looking at total transactions:

trans1You could look at the number when it’s distributed by “visits to transaction”, and learn that most people are ready to buy during their first visit:

trans

Rule out outliers & statistically insignificant stuff

Let’s say you want to identify your pages with the highest bounce rate. Should be easy: go to Site Content -> All Pages and sort by Bounce rate. You get this:

top-bounce

It’s useless. Why? Cause all the pages listed here have 1 to 4 visits. The data for these pages is not statistically significant. So we need to use filters to exclude all low traffic pages.

Click on ‘Advanced’ and let’s set the minimum number of unique pageviews to 1000:

advancedfilter

Now you get this:

topbouncefiltered

And you can actually check out what could be wrong with these pages – and compare top bounce product pages to lowest bouncing product pages to form a hypothesis as to what might be wrong.

Use total numbers next to ratios

What are our best converting traffic sources?

top5

Wow, look at that! Referrals are the best!

Let’s set some context around it by adding the number of visitors and transactions (click to enlarge):

ratetotal

So it seems like the “top 5” only account for 177 transactions out of a total of 2460. Where do the rest come from?

top-transactions

Having transaction numbers next to conversion rate will give us context. Referrals are actually a rather small source of revenue here.

If you’re new to Google Analytics, start with this Beginner level course (included in this CRO Minidegree). If you want to go all in, check out our Digital Analytics Minidegree.

Navigate to...

Current Lesson:

Analytics fundamentals

Get unlimited access

Get this full course and more with an All-Access subscription

Sign up now

CRO foundations

Prev