Not everyone’s got piles of cash to spend on improving their website, so in this post I go through what I would do if I only had £1,500 to make quantifiable improvements to my (fictitious) e-commerce website.
First, a word about my £1,500 budget. In this scenario, I am lucky I don’t have to pay for analysing the data nor using well-known ‘rules of thumb’ (often called heuristics).
But I do have to stump up hard cash for services like website development, design/copy changes, usability testers and setting up split tests. You may have these functions in-house or can beg, borrow or steal to reduce their cost.
So, I need to spend my £1,500 wisely. It’s going to be a mix of expenditure on finding stuff out and implementing changes to the website. Since I am cautious man, I am going hold some of the budget back, at least 60%, to make sure I have enough for funds left for implementation.
Plus, if I have any cash left over, I can always spend it on celebrating our great results with the team
So here are the three main stages of my approach:-
- Going for Eureka
- Taking action
- Validating the results
I’ll go into a bit more detail into each one below, but suffice to say I’m not going to blow the budget on discovery, only to find I have no cash for taking action.
1. Going for Eureka
Perhaps, I have made decisions on gut instinct rather than looking into the numbers. I may have even turned into a HiPPO (someone whose opinion is listened to only because s/he is the highest paid person).
So, I need to emotionally get ready to be wrong. I am going to turn the oven on, because I may be eating some humble pie. This means I am going to be looking to find out how exactly my website visitors behave on the site, and what’s the experience.
Thus, far I have committed to looking at taking both a quantitative and qualitative approach to making discoveries.
This discovery phase has three stages:-
- Tracking the right numbers
- Collecting information
- Developing hypotheses
Tracking the right numbers
Before I begin my journey of discovery, I need to make sure that I am tracking the right numbers in my analytics tool (We use Google Analytics, but if you don’t the main points are still valid).
To configure your web analytics tool, you need to:
- Get rid of the data that’s of no use to you – this means applying filters (e.g. eliminating website visitors from your own office) and excluding query parameters – so that you’re left with data that you can make sense of
- Measure the right goals for your type of business, and remember to investigate this data by segmenting it – e.g. new vs returning, branded searches vs. non-branded, short visits vs long visits.
Since this is an e-commerce website, I want to set up extra goals to ensure we measure the following:-
- % of visitors who view a category page
- % of visitors who view a product page
- % of visitors who view their basket
- % of visitors who view the pages in the checkout process
- % of visitors who complete a failed search e.g. those search terms resulting in ‘no matches found’
At this stage I am setting up my tool so it’s easy to see where people are dropping out, and looking at how these drop-out rates vary across different segments. To do this I would create a horizontal checkout funnel report (first devised by Justin Cutroni).
Segmenting your checkout funnel means you can see where new visitors, who won’t have got used to the idiosyncracies of your checkout process, are abandoning.
Once I have got these goals in place I can then apply values to the completion of the goals. This helps me identify which step in the process might be best to optimise.
For example, depending on my e-commerce data, I might give these goals the following values (Your mileage will vary).
|Goal||Value of ‘goal’|
|Visit to a category page||£1.58|
|Visit to a product page||£3.45|
|Visit to a key landing page||£2.21|
|Visit to a basket||£8.01|
|Visit to the payment page||£15.54|
|A failed search||-£4.28 (this is a ‘loss’ of revenue)|
Once I have sensibly configured my analytics tool and set up financial measures, I can now start to collect information.
By analysing the ‘behaviour’ of my website visitors, I can start to answer these types of questions:-
- Which key landing pages have the highest bounce rates, and how much revenue do I lose?
- Which search terms have the highest rates of ‘no matches found’, and again how much revenue am I losing?
- Which category pages are great at converting visitors into viewing a product, but have relatively low levels of traffic?
- Which products have higher-than-average add-to-basket ratios, but low levels of product views?
- Which product has the highest level of exits?
- Which stages in the checkout process has the highest level of drop out for new visitors?
Cost of getting answer to these questions: £0, remaining £1,500
To help me understand the experience of my website visitors, here are 3 things I would do:
- Conduct a home page review
- Perform an eye tracking study on my key landing pages
- Carry out usability testing
Home page review
First of all, I want to be convinced that new visitors understand what our website offers. New website visitors typically make their decision as to whether or not they ‘get’ what a website is offering in less than 8 seconds.
That’s why I would use a service from Feedback Army and ask 50 testers to tell me:-
- What does the site do?
- What stood out on the home page?
- What aspect of the home page confused you?
- What benefits does the website offer?
Here's an example of the type of feedback that I would get.
Eye tracking studies
Inevitable truth: you cannot convert a website visitor into a customer if they have left your website (at least that day). So I am interested in why website visitors are leaving my key landing pages at such an alarming rate.
One way to get a handle on this question, is to use eye-tracking studies to see what people do and do not notice about our landing pages. Are they seeing the call-to-actions, or are they distracted by an internal promotion banner?
I’ll get answers to these questions by looking at heat maps and replays of the tester’s eye movements.
Services like Gazehawk are great for eye-tracking studies, and turn around results in less than 10 days.
Home page reviews and eye tracking studies are insightful ways of understanding visitors’ reactions to specific pages. However, to really understand how easy it is for website visitors to complete a process, nothing beats usability testing.
As you can see from the YouTube video, usability testing involves testers who have screen capture software and a microphone, completing a specified task on your website.
I would write a script for the usability testers to follow, that includes:
- A scenario (It’s your father’s birthday..)
- A budget
- A requirement to change add and remove items from the basket
- An instruction to change their delivery address
- An directive to complete the checkout process (stopping before they reveal their credit card)
Remote usability services such as WhatUsersDo and UserTesting.com simplify the process of recruit usability testers and producing videos of the testers’ experiences. Plus, they already have thousands of testers on their panel. WhatUsersDo.com also allows you to tag up significant point in the usability videos, so I could share these with my team.
Using the experiences of my website visitors, I can start to answer another set of questions, such as:-
- What benefits do visitors believe we offer them?
- Why do visitors bounce off the home page?
- What do visitors notice, and not notice, on key landing pages?
- Why do visitors have difficulty in finding certain products?
- What obstacles do we put in visitors’ way who want to checkout?
The cost of answering these questions is at total of £485 (see below), so I have £1,015 remaining.
- Home page test (cost: approx. £30)
- Eye tracking study on my key landing pages (cost: approx. £300)
- Usability testing with 5 testers (cost: approx. £125)
2. Taking action
From answering these key questions, I have developed some great insights into what I can to improve my website.
To prioritize the action I want to take, I need to first of all understand the financial impact of improving various aspects of the site.
Let’s assume that we believe, with our remaining budget, we could achieve a 5% improvement in the following measures. What’s the revenue increase of making that 5% improvement, all other things equal.
Just to be clear – that 5% of the measure, not 5 percentage points. So, our aim would be reduce the bounce rate on the home page from 60% to 57%, rather than to 55%.
A 5% improvement feels an achievable goal, and will give me a clear target against which I can judge our results. Otherwise, I won’t know when to crack open the champagne!
The numbers below are for illustration only.
|Website area||Measure||Revenue increase of 5% improvement in measure|
|Home pages||Bounce rates||£12,500|
|Landing page||Bounce rate||£30,500|
|On-site search||Failed searches||£96,650|
From the table above, I know I need to prioritise fixes to the product page, followed by on-site search. Improving these two areas will give me the biggest bang for my buck.
Within the product page and site search areas of the site, I would scan my usability videos to identify the issues my websites visitors face. Armed with a list of improvements to these two areas, I would go through this list with my website developer, designer and copywriters to get a cost for each improvement.
Here I am looking for high-impact, low cost improvements within my remaining £1,015 budget.
Let’s say I cap my budget to £800, to give myself some budget for running some tests, what might that get me?
- Essential delivery charges and returns information on my product page
- Changes to the number of images that can be displayed
- A redesign of the Add to Basket button
- Better copy that promotes their benefits of my bestsellers
- A changes to the on-site search results page so that I can at least promote my bestsellers when ‘no matches found’
It all depends on what insight you get from your discovery phase, and what extra sales you can generate from a 5% improvement, as well as what your website developers and creative folks charge. As I said before, your mileage may vary;
So, assuming I have kept to my implementation budget of £800, I still have £215 left.
3. Validating the results
Once I have chosen what changes I am going to make to the site, I want to be 100% sure that these improvements have made a real difference to sales.
Using Google Website Optimizer, a free testing tool, allows the web server to serve pages that have been improved as well as a version of the same page that has not changed at all.
The difference in conversion rates from the two version – new and old – tells me whether the new version outperformed the existing page, and by what margin.
I have got a web developer friend who’s agreed to set up and run a split test for me for just £200.
So with my remaining £15, I have bought a bottle of Cava, (no funds left for champagne) to celebrate once the results from the split test come in.
To recap, this is how I spent my £1,500 budget:-
|Home page review||£30|
|Eye tracking study on key landing pages||£330|
|Website development / images / copy||£800|
|Set up of split test||£200|
Conversion Rate Optimisation – or CRO – may seem like a dark art if you’ve never experienced it first-hand. Ecommerce companies who embrace CRO claim to get stellar results and ever increasing online sales.
If you’d like to replicate their achievements, then the first step is to understand the CRO process as a whole. To help with that, we’ve created this infographic which gives an overview of all the key elements involved in successful CRO, designed like the well-known periodic table for chemical elements.