Do you feel completely overwhelmed when it comes to improving your website? Most MSPs I speak to feel exactly the same.
You know how important a great website is for your business. But knowing what to include, how it should be written or how it should look seems… overly difficult 🙄
I bring good news.
Google Optimize can demystify all of this for you. Stop guessing, use Google Optimize and let data guide you instead.
What is Google Optimize?
It’s a free optimisation tool that helps you test different variants on your website and see how they perform against a specific objective.
This is A/B split testing, my friend. It’s what virtually all big web companies use to change their web pages and make them better. Ever wondered why the Amazon sales page is SO cluttered and full of info? Because the testing has shown them over many years the format that results in the greatest number of sales.
Yes, an MSP can do this (and really should). You have a powerful and free tool in Google Optimize. It runs the experiment, measures the results and tells you which variant is the leader. Let me put it another way – you can change your website based on the behaviour of the people using it.
Cool, huh? Every MSP should always be running an Optimize experiment on their website.
Here’s how to get started
- Follow these steps to get your account set up and connected with Google Analytics: https://support.google.com/optimize/answer/6211921?hl=en
- Then, follow these steps to get the Optimize code on your website: https://support.google.com/optimize/answer/7513085
(We used Tag Manager to do ours)
Once that’s done, you’re then ready to create an account:
- Name your account, change to your country and tick the boxes.
- Add your site container (your website’s URL)
Now you’re ready to create your first experience:
Before you continue, you need to think about what you would actually like to achieve.
Maybe you would like to:
- Decrease bounce rates
- Get more views of your home page video
- Spur more people into requesting a 15 minute video call
Desired actions like these become your objectives. You can play around with variants such as page design, headlines and placements to see which variant helps you get more objective conversions.
Here’s a practical example. Let me talk you through our case study first, then I’ll explain the different options that are available to you for your MSP:
We’ve been using Google Optimize. What was our objective?
To increase the amount of people searching for their postcode on the UK version of www.mspmarketingedge.com.
Here’s what the site looked like above the fold* before our experiment:
* NOTE: Above the fold means what you see on the screen without having the scroll down. Obviously this will vary hugely depending on screen size, resolution and device. There are tools that can help you.
The term originally comes from newspapers. When the paper was folded to go on a shop counter, the only thing that could help to sell the paper to the reader was content “above the fold”. It’s why you must always put the most important content at the top of your web page.
To measure our objective, we used a Google Analytics goal set up on the website’s “search” button. So, every time someone searched for their postcode, a goal was fired in Google Analytics.
What type of experiment did we want to run?
I’ll talk you through the different experiments you can run later on, but for this experiment we wanted to run an A/B split experiment.
What variants did we want to test?
We wanted to experiment with our main headline to see if changing it could improve postcode conversion rates. Yes, this is exactly the kind of thing that Google Optimize will figure out for you.
Existing headline: “Every marketing tool you need to get new clients for your MSP”
New headline to test: “Check now to see if one of your competitors has locked your area”
Optimize calls these experiments “experiences”. What’s beautiful is that it creates the test page for you. No mucking about creating duplicate pages in WordPress 🥳
Here’s a step by step guide how we did it:
- Go to your account -> container -> create experience
- Create Name and pick what experience you would like to create
3. Add varient -> Name Varient
4. Click Edit and it will bring up your website
5. Highlight the element that you would like to change, in our case the headline and amend
6. Click save and done at top right hand corner
7. Make sure page targeting is the right URL
8. Define audience targeting (typically leave it for all visitors)
9. Pick your objective
10. Check settings are all okay
11. Scroll back up to the top and click START
The experiment will run and you’ll be notified when it has completed and when the winner will be revealed.
As you can see the new headline is the clear winner, with a 10.01% conversion rate over a 7.22% conversion rate of the original headline.
Interestingly, when we ran the same experiment on the US version of our website, we saw different results.
For our US site, the original headline had a 14.12% conversion rate, and the new headline a 10.34% conversion rate. This is why you must run experiments on every website you own (especially if you have a general MSP website; and then a version aimed at a specific vertical. Different audiences behave in different ways).
How did we interpret these results?
The first headline, “Every marketing tool you need to get new clients for your MSP” is a descriptive sentence outlining the benefits of the service.
Whereas the second headline, “Check now to see if one of your competitors has locked your area”, is a suggested action highlighting scarcity of our service.
We believe that the second headline works for the UK, because our UK audience has more awareness of who we are and what our service provides. So a more direct headline increases our conversion rates because many visitors don’t need to explore what our service is all about first.
Whereas, in the USA we get a lot of new traffic to the site (via Google and Facebook ads). These visitors might have not necessarily heard of us before. Our descriptive headline is more appropriate to them as it gives them a quick summary of what our product is – once they know this, they are more inclined to search their zip code straight away. Whereas the more direct headline leaves them trying to figure out what our service is about first.
The takeaway for this for your MSP is to know your audience and how they are feeling when landing on your website.
This was a really interesting experiment for us. We now have a huge number of other experiments underway and planned.
Here’s what you want:
Let me help you understand how you can use Google Optimize for your MSP
What are the different type of experiments you can create?
- A/B tests
This type of experiment measures two or more variants of the same web page against a specific objective. For example, what’s the impact of a different coloured call-to-action button, or a different headline? If you can optimise dozens of different elements on your website and get just one more enquiry that becomes a client – the power of LTV (LifeTime Revenue) makes that a big win, right?
- Redirect tests
This experiment (aka split URL test) allows you to test separate web pages against each other. Variants are identified by URL or path instead of page elements. You’d use this to test different landing pages, or a complete redesign of a page.
- Multivariate tests (MVT)
This experiment tests two or more elements simultaneously to see which combination creates the best outcome. Unless you’re getting really really good traffic, I’d avoid these. Stick to the first two tests.
Unlike experiments, personalizations can run forever and don’t have variants. A personalization is a single set of website changes that are shown to all website visitors that meet the targeting conditions.
- Banner template
Again, this isn’t an experiment. It’s a measure Google has put in place to help with important updates for your customers.
What objectives can you measure?
There are 3 types of objectives you can measure:
- System objectives:
2. Google Analytic Goals:
3. Custom Objectives:
What type of variants could you test?
This is completely down to you. And depends on how your website is designed and what objectives you would like to measure.
Try starting at the very beginning. So, when someone lands on your home page, what’s the next action you would like them to take?
Maybe it’s to watch your home page client testimonials video. Or click through to your about us page. Or book a 15 minute video call with you.
Use this as your objective. I’d start by changing your main page headline as the first test. You can test every element of the page over time.
Once you’ve done this, you can then move on to more advanced experiments.