Google Analytics

The king of analytical software and something no website should be without.

Completely free, extremely easy to implement, there is no excuse for not having Google Analytics on your website.

This doesn’t mean understanding what is in your analytics reports will be easy so, again, my recommendations is to keep it simple.

1. Use Google Analytics to look at the key metrics that define website performance; number of visitors, conversion percentage, conversion value, and total revenue.
2. Drill down into one variable at a time - where the visitor came from (their channel), where they started out (their landing page) and look for patterns
3. Look for under-performing scenarios, such as one landing page that isn’t converting as many customers as others, and make a change to try and improve your website
4. Wait…
5. Re-check your analytics to see if your change has made things better… or worse.

Yes, it’s literally this simple. The trick, or the “art” if you prefer, is to spot the patterns in the first place.

I could write a whole book on spotting patterns in Google Analytics (I probably will, one day) but, for now, these are a few common factors you should always be looking into:

1. Traffic Source: Where did the visitor come from? Was it a search engine, a link in a tweet, your cost-per-click campaign?
2. Landing Page: What page did they start their journey on?
3. Device Type: What type of device (desktop, laptop, tablet, mobile, etc.) was the customer using?

Murder on the Analytics Express

If I had £1 for every time someone had said to me “our sales are up/down because of X”, where X was one thing and they hadn’t bothered to look at any other factors… well, I wouldn’t need to be writing SEO books anymore. I’d be on a beach somewhere, writing books about something else entirely.

The point is, it’s easy to latch onto the first link you find between (let’s say) site speed and conversions. “Page load times went up and conversions went down”. Seems logical. Is logical. But is it the whole story?

What if conversions had been slipping down for months, but the site had only slowed down since the last software update? What if traffic had also dropped, or a popular product had gone out of stock or been taken off the market, or you’d stopped spending money on CPC this month?

Much like the conclusion to Murder on the Orient Express (no spoiler apology, if you haven’t read it then you’ve brought this on yourself) - there’s often more than one culprit.

Don’t get sucked in by your first deduction - be sure that you’ve looked at the problem from more than one angle.

_“Unless you are good at guessing, it is not much use being a detective.”_- Agatha Christie

Land then Bounce, Exit or Convert

When looking at your analytics data, remember that there are three possible outcomes to every visit: Bounce, Exit, or Convert.

Bounce: The visitor leaves without going to another any other page - they literally “bounce off” your website.

Exit: The visitor leaves, but this is not the first page that they looked at.

Convert: Huzzah! The customer converted. They will now exit, but you got the conversion.

All three are normal states, but obviously we want to

1. Reduce the number of bounces to be as low as possible
2. Have the minimum number of non-converting exits

Attribution Models

Not every visit that ends in a bounce or an exit without a conversion is necessarily a failure. Depending on your particular industry or niche, it may require multiple visits to your website before the customer will convert.

Thinking more broadly, it’s a well-established marketing principle that it takes seven “touches” before someone will be ready to interact on a call to action. These touches can be online or offline, such as

• A physical connection, such as meeting at a networking event
• Seeing an ad, either physical or digital
• Seeing your logo, maybe as a sponsor or on a brochure
• Seeing your social media posts in a news stream
• Receiving your e-newsletter or other email marketing piece
• A phone call
• A word-of-mouth mention by a friend or colleague
• Visiting your website

Google offers a special type of report, the “Attribution Model” that allows you to see which online touches have taken place in the run up to a conversion.

If you’re unsure, for example, of what contribution your social media activity is having on sales the Attribution Model will be able to show you. 

Using this model you can control the timeline over which the interactions have taken place, my preference is for a 90-day window, and whether conversions are allocated to the first interaction, last interaction, or proportionally across multiple interactions.

When you are looking at a “normal” Google report you are already looking at a “Last Interaction” report - if the report tells you the traffic source is (for example) email then that is the last interaction. This is why I like to look at First Interaction on my conversions - this is the interaction that is doing the hard work of creating interest in a website/brand for the very first time with a customer. 

Advice on Time Travel

When conducting comparisons of any statistic, it is important to take into account the times that you are comparing.

Many businesses will have seasonal fluctuations in their sales - it’s not rational therefore to always compare this month to last month. Most businesses will want to look at month to month statistics, but comparing the past month to the same month in the previous year is more logical.

Don’t fall into the trap of comparing apples with oranges (or Mays with Decembers).

The Cookie Monster Paradox

I mentioned a little earlier the problem of running too many analytics platforms.

Not only is it a waste of time (unless you have an army of people to interpret all that data) but it’s also having a negative impact on your website performance. Most analytics tools need to deposit a cookie (a small text file of information) onto your computer to help them remember you from page to page and visit to visit. Google Analytics is no exception.

I’ve been using a tool called Ghostery (https://www.ghostery.com/) which tells me what tracking technologies websites are using and allows me to block them, if I want to, based on these cookies.

The most incredible thing that I’ve noticed using this utility is how much faster the web becomes when I’m not using my bandwidth to shovel data to umpteen third party tracking platforms that have been integrated into websites that I visit.

Like me, you’re probably already sick of cookie pop-ups asking you to complete complex preference arrangements on websites. Here’s my tip - install a cookie blocker like Ghostery and you’ll never need to worry about being tracked online again. You’ll also save bandwidth and have a faster internet.

“It's impossible to move, to live, to operate at any level without leaving traces, bits, seemingly meaningless fragments of personal information.”

William Gibson

Can’t someone else do it?

If your website provider cares about you, they will make sure that you have access to Google Analytics and they will make sure you are getting regular reports.

If your website provider is one of the snake-blooded sort who are secretly reptiles wearing human skin... then they will send you the reports once a month and charge you for the privilege.

Either way, make sure you have the ability to log in and at least explore your analytics. You wouldn’t run your business without checking your accounts on a regular basis, would you?