Getting Started With Personalization: 5 Tests to Try for a Quick Conversion Lift.
5 Minute Read
Getting started with personalization can be daunting, with seemingly endless tests to try for that all important conversion uplift.
Getting to grips with the machine learning technology required to implement personalization can also feel like an intimidating task.
But it shouldn’t.
Firstly, it’s important to establish your goals. What do you want to achieve with personalization? Is it an increase in conversion rate? Or improved average order value?
Once you have a tool that clearly reports real time data back to you, then you can make decisions on who you want to target, and with what content, to achieve those goals. Are you struggling to convert new visitors, for example, or increase orders from traffic coming from a certain region?
At this early stage, it’s always helpful to look at tried and tested experiments that have worked for others. Here are 5 quick win tests to give you some inspiration.
1. Increase first time visitor sales with exit intent messages
First time visitors are, unsurprisingly, the hardest to convert.
They have no loyalty to your brand. They have no past purchases to benchmark for quality. They’re hesitant to buy from you, and are likely to shop around. An exit can often signal the last time you’ll ever see them.
How can you make an impression on these visitors?
A great way to engage first timers and give a little nudge to stay, is to trigger an exit-prevention light box. This message will trigger when visitors attempt to leave the store with an incentive of your choice.
As an example, WallPops.com engaged their first time visitors with an incentive-to-stay pop-up. It increased orders by an impressive 12%.
How to do this
With a tool like Bunting, it’s really easy to set up this test. ‘Increase first time visitor’ is a pre-made template, that can be clicked and installed straight away. You can choose what kind of incentive to use to pique your visitors’ interests, such as free delivery, or a % off discount (NB: your store must have a coupon code to do this). Or even better, split test both, to see which incentive works most effectively.
Don’t want to give out discounts or free delivery? Worry not. An alternative here could be an ‘email me my cart’ message, or ‘sign up for our newsletter before you go’. This gives you opportunities to re-market to these visitors with personalized emails based on their browsing history.
2. Raise average order value with triggered messaging
A common way to raise order values is to offer an incentive to raise the value of a customer’s cart. One that works well is to offer something when a customer exceeds a certain spend. This could be a free gift, or free delivery.
This can be targeted to specific users – for example, you might only want to target visitors who have purchased from you several times, to reward them for their loyalty.
In addition to displaying this offer on your homepage, you can also trigger a reminder message to your visitors when their cart reaches a certain amount. It will then ask them to increase their order value in order to get free shipping. Saltrock surfwear did this well. They triggered a message when the value of a visitor’s cart exceeded £20, letting them know that they would be entitled to free delivery if they increased their order to £30.
How to do this
With Bunting, you can easily set up this test by creating a pop-up light box. Once the content is created, you can set the trigger to target visitors that match a certain criteria. You can select the condition type ‘value of cart’ under ‘shopping cart metrics’. Then, specify the order value you would like to target, such as $20. Finally we recommend you run a split test, and then you can measure precisely how well the experiment performed.
3. Product recommendations on product pages
Recommendations are an indispensable part of any e-commerce store, and rightly so – machine learning algorithmic recommendations undoubtedly raise your conversion rate and average order value.
Of course, they help customers find what they really want. But they also help your team by freeing up precious time spent manually creating recommendations. A good tool will let you set up recommendations that require zero coding, so they can easily be set up by anyone on your team.
How to do this
With Bunting, it’s a case of drag and drop (and thankfully, no coding).Simply paste the URL of any product page, and the tool will generate accurate recommendations on every product page. Of course other pages such as the homepage or cart page can also be selected. Choose carefully where on the page you’d like the recommendations to be. We recommend above the fold, and in the case of product pages, we suggest the closer to the main image the better.
4. Raise visitor sales based on geolocation
Research frequently shows that many retailers are missing out on opportunities to attract and convert visitors from certain regions.
Your international visitors are less likely to trust your store than a local website, so it’s essential to earn their trust through your site’s offering. Have a look through your site – are any areas displaying messaging that is not relevant to your international visitors?
Looking at the Adventure Direct site, we can see the messaging is aimed at the UK market, and is presented to every visitor regardless of location. This may actually damage overseas orders. With Bunting you can dynamically change the website text to show different messaging to your overseas visitors.
How to do this
Any part of your website can be changed to display messaging and imagery that is relevant to that particular audience. With Bunting, you can change any part of the page for this segment. For example, it will let you edit website text and even include a merge tag, which will display the visitor’s country/region name, wherever they are.
Welcome pop up for overseas visitors Another great little feature to impress your overseas visitors is to trigger a welcome pop-up. This can assure your visitors of your custom, that you deliver to them, or can display your international contact details.
5. Cross sells on the cart page
The recommendation best known to raise order values is the mighty cross-sell. These are algorithmic recommendations for products that complement the product/s in the cart.
Cross-sells are based on real time order data, meaning they’re the products statistically most likely to be purchased in conjunction with the item/s in the cart. They help visitors find useful or desirable products and, in doing so, not only raise orders but help you build better relationships with your customers.
How to do this
To start with, it’s worth trying cross sells out on your cart page. A good tool will allow you to drag and drop your recommendations into place. Importantly, we recommend for your first attempt at cross selling to target visitors who have converted at least once before. This is because sometimes cross-sells can take first time buyers out the flow of purchase and serve as a distraction.
Measure your results
We cannot stress enough – when getting started with personalization, test everything.
Many tools don’t allow for split testing, meaning you’re left without the important insights you need to build on and improve your personalization campaigns. So, run an A/B test with each experiment, and look at the impact on conversion rate, but also revenue increase, abandoned cart rate and the number of page views.
We hope some of the tests in this article have given you some ideas for your own campaigns. Good luck!
Want to know more about personalization, and how to get started? Contact Bunting for a free review of your website, and to chat with an expert.