Prioritising Product and UX Improvements

How do you eat a UX elephant?

The five-word sentence “To do: website UX improvements” is guaranteed to strike fear into the heart of any serious UX practitioner. Behind this sentence belies a many-headed beast of an errand which lurks indefinitely on task lists and never seems to get done.

When time is of the essence and there are a million other deadlines to hit, the key to attacking this to-do list monster is breaking the task down properly into manageable chunks and deciding which are the most important to tackle first. In order to do this, we’ve created a system for prioritising site improvements using Google Analytics.

1. Setting goals

When optimising a site, an important place to begin is by clarifying what purpose optimisation will serve. The key question here is what the site is being optimised for – conversions? Improved bounce rate? Promotion of a new service?

This article is written on the assumption that conversion optimisation takes top priority. Feel free to contribute ideas of how to optimise for other objectives.

2. Creating a quick conversion path map

A conversion path includes all the stages customers go through before hitting the designated conversion point – be this a sale, getting in touch via contact form, etc.

There are two quick methods I like to use to rapidly map out conversion paths: a creative method and a more technical one:

The creative method involves sitting down with a big piece of paper and brainstorming journeys potential buyers might take throughout a site. For this exercise, I tend to include all parts of their journey – even the offsite ones – to properly get into the user mindset. As such, a typical journey might be something like “Social media account > Blog > About Us > Contact page”, or “Search Engine > Home > Product page > Shopping Cart”.

The technical method requires a Google Analytics account with conversion tracking set up. Open Analytics, go to Conversions > Multi-Channel Funnels > Top Conversion Paths, and set the primary dimension as “Landing Page URL Path” (see screenshot). This shows all paths that have been taken to reach a conversion, and puts them in order of which is most common.
conversion path
Both methods have their limitations. Sitting down and brainstorming doesn’t incorporate any kind of external input. Analytics tells a limited story, as it only measures currently popular paths instead of potential ones which it might be wise to optimise. The ideal strategy, of course, is to use a mix of the two, but let’s not forget that our aim here is to keep moving and not get hung up on the details! So if conversion tracking isn’t set up or creative inspiration is lacking, don’t hesitate to just use the most readily available method.

For more information about more thorough ways to map out conversion paths, see this article.

3. Organising into an Excel spreadsheet

To undertake the next step, first make a list of all types of page which could potentially do with improvement. This might include: home page, blog pages, portfolio pages, product pages, “about us” page, contact page, error pages and anything else featured. The more page types, the more complicated the next step will be, so aim to keep things simple and group pages together where possible.

Then, put these pages into groups according to where they’re up to in the conversion path.

Level 1 – external pages

Includes pages which, arguably, require the lowest engagement levels of all from users: the different external pages they may come from. This includes social, AdWords, search engine results pages, referrals, campaigns… It might be more difficult to control these pages than on-site ones, but it’s worth keeping them on the map anyway as extensions of your site.

Level 2 – low-level engagement internal pages

This category includes pages which are mostly read by users still relatively low on in the sales cycle and a couple of steps away from conversion: blogs, press releases and the like. A homepage might also feature here, depending on how far along it features on the average conversion path.

Level 3 – medium-level engagement internal pages

This category includes pages frequented by users who seem to have a clear interest in the products or services on offer. It might include solutions/product pages, and perhaps a portfolio page if this often leads directly to conversions.

Level 4 – conversion pages

This category includes all pages which contain the designated conversion point: contact pages and the like.

Groups don’t have to be too closely defined here – we’re just aiming to give things a little structure. So don’t worry too much about putting things in the “wrong” group.

marketing funnel
Next, create an Excel tab for each level, and within each tab create a heading for each page. Under the heading, write the next higher-level steps consumers are likely to take from the current heading, as dictated by the conversion path maps created in step 2 (see screenshot).
marketing excel

This should create a record of any pending ideas and proposed changes that have been hanging around in the air (every company has them!), organised according to the different stages in the conversion path.

4.       Prioritise

Got all those niggling tasks down on paper? Great! Now it’s time to decide where to begin optimising.

Set a priority page. We’ll use Google Analytics to find the optimum page to begin with. Go to Google Analytics > Behaviour > Site Content > Landing Pages on the left-hand menu to find out which is the most common landing page. This page, which is where the majority of users will begin their journey, takes priority over other pages, and is where site optimisation should begin.

Now, let’s look at where exactly to start improving this page. If conversion tracking has been set up, it’s possible to see where the most common conversion path from this page normally lead. As above, open Analytics, go to Conversions > Multi-Channel Funnels > Top Conversion Paths, and primary dimension as “Landing Page URL Path”. It’s then possible to search for paths including the priority landing page using the search box to see where they commonly lead. Herding users from the priority page to the next logical stage in the journey should be nº 1 at the top of the task list.

If conversion tracking isn’t set up, or a site receives too many conversions for this to be practical, it’s also worth thinking about where the most optimal next place to drive traffic from the priority landing page is.

Organise other pages. Once optimisations for the priority page are well underway, there’s several choices for how to prioritise other pages, depending on which effect is desired:

  1. High-traffic site with lots of interesting content that draws users in, but few conversions? Prioritise with level 1 pages and work up to push users through the final stages of the marketing funnel.
  2. Small site with just one or two key conversion paths? Work through each path in turn, starting with the one with the most frequently visited landing page.
  3. Looking for general site improvement? Steadily work through the list of highest traffic pages (find this information in Google Analytics > Behaviour > Site Content > All Pages) for a general site-wide boost.

In any case, there’s a good chance that just brainstorming ideas for the priority page will give plenty to get started on!


Next steps

Time to get cracking on those site improvements! Hopefully the way to move through them is now a little clearer.

It’s important to note, however, that the point here isn’t to follow this article to the letter. More than anything, this exercise is designed to promote thinking about site improvements in a structured way, and by doing so help break out of paralysed indecision and spur on action. So don’t hesitate to get distracted and go out and start improving things – that’s the idea!

I’d like to point out as well that this is not an exact science, and is simply designed to share my own system for organising things. If you have another way to categorise your site improvements, I’d be delighted to hear what you recommend.


Add a Comment

Your email address will not be published. Required fields are marked *