Talk to us

Let's talk websites

Ready to create a website, or improve your current? We are!
We just need to know a few things so we can provide the best advice.

We'll be in touch within 24 hours!
Oops! Something went wrong while submitting the form.
We reply within one business day.

One Size Fits None: How Personalisation Changes Conversion Rates

Research lab
Founder of skyrocket, samantha thornton-jones
October 9, 2026

Written by Samantha Thornton-Jones, founder of Skyrocket. We’re a certified Shopify and Webflow studio in Auckland and Queenstown, building for established New Zealand and Australian businesses.

Personalisation lifts conversion when it starts with a few segments that already behave differently, such as location, traffic source, behaviour and purchase history, and each change is measured against a control group. This covers how it works on Shopify and Webflow and where it goes wrong.

A first-time visitor lands on your site and sees the same homepage, the same featured products and the same offer as a customer who has bought from you six times. Neither of them is served especially well. The new visitor needs proof and a reason to trust you, while the returning customer needs to get back to what they buy without wading through an introduction they have already read.

Personalisation means showing different content to different visitors based on what you know about them: where they are, how they arrived, what they have looked at and what they have bought. Done well, it shortens the path to purchase for the people most likely to buy. Done badly, it shows irrelevant recommendations, feels intrusive and adds a layer of complexity the team cannot maintain. The difference comes down to three decisions: which visitors to treat differently, what to change for them and how to prove the change earned its keep.

Personalise by segment before you personalise by person

Most of the commercial value in personalisation sits at the segment level, well before any one-to-one recommendation engine. A handful of well-chosen segments, each with a small number of changes, will outperform an elaborate system that tries to adjust everything for every visitor and has too little data to do it accurately.

The segments worth starting with are the ones that already behave differently in your analytics. Compare new visitors with returning ones, mobile with desktop and New Zealand traffic with Australian. Then compare visitors from paid search against those from email. Where two groups convert at very different rates, or buy very different products, there is a case for showing them different things. Where they behave the same, personalising for them adds work and changes nothing.

Four kinds of personalisation, in order of effort

Location

Location is the simplest and often the most valuable. A store selling into New Zealand and Australia should show each visitor the right currency, the right shipping information and products that can be sent to their country. Shopify Markets handles currency and country-specific pricing natively. On a service site, location can mean showing the right office, phone number or service area. If you sell across the Tasman, our piece on the website setup NZ brands get wrong when going global covers the structural decisions that sit underneath this.

Traffic source

A visitor arriving from an ad for a specific product range has told you what they are interested in before the page loads. Sending them to a landing page that continues that message, with the same product and the same offer, is personalisation by another name and one of the cheapest forms to run. On Webflow, this is built through campaign landing pages. On Shopify, it can be done with dedicated collection pages or landing templates.

Behaviour

Recently viewed products, related products and complementary products all respond to what the visitor is doing in the current session. Shopify's Search and Discovery app provides product recommendations that the merchant can adjust by hand, which matters because automatic recommendations on a small catalogue will sometimes pair products that make no sense together. Behavioural personalisation earns most on stores with a broad range, where a visitor can easily lose the thread of what they were looking at.

Purchase history

Returning customers who are logged in, or who arrive from an email, can be shown content based on what they have bought. Replenishment reminders for consumables, accessories for a product they already own and early access for repeat buyers all fit here. Most of this value lives in email rather than on the site itself, where the customer is identified and the timing can be controlled. Our guide to email marketing on Shopify covers how those flows drive repeat orders.

What changes for each segment

Once you know who you are personalising for, decide what changes. The best candidates are the elements with the most influence on the decision: the homepage hero, the featured products or services, the primary call to action, the proof shown beside it and the offer.

A returning customer might see the products they have bought before and a quicker route to reorder. A first-time visitor from Australia might see shipping times to Australia and reviews from Australian customers. A visitor from a campaign for commercial clients might see commercial case studies instead of residential ones. Each of those is one change to one area of a page, made for one reason. Keep the list short. Every variant is something the team has to write, maintain and keep accurate when prices, products or stock change.

How Shopify and Webflow handle it

On Shopify, the building blocks are native or close to it. Shopify Markets covers location and currency, while the Search and Discovery app covers recommendations. Customer segments can be used for targeted discounts and email. Apps add more advanced options such as content that changes for logged-in customers or by traffic source, and each one adds weight to the storefront, so it needs to earn its place.

On Webflow, the platform's own Optimize add-on offers testing and personalisation by audience, including location, traffic source and behaviour, inside the Webflow environment. It is a paid add-on, and it makes most sense for sites with enough traffic to measure the results. Third-party personalisation tools also integrate with Webflow through custom code. For many Webflow sites, well-built campaign landing pages and location-aware contact details will deliver most of the value before any tool is needed.

Where personalisation lowers conversion

Recommendations that miss

A customer who looked at running shoes and is then shown gardening supplies has learned that the site does not understand them. Poor recommendations are worse than none, because they take up space that could have shown your best sellers and they make the store look automated in the wrong way. Review what the recommendation blocks show for your top twenty products and correct the pairings by hand where needed.

Personalising with too little data

A new visitor has no history. A store with a few hundred orders a month has too little data to produce reliable individual recommendations across a large catalogue. In both cases, the right content is your best-selling and best-reviewed products, chosen by a person. Personalisation based on thin data produces noise and calls it relevance.

Making the tracking visible

Recommending relevant products is helpful. Telling the visitor, in words, that you have been watching what they browsed across several visits feels like surveillance. Keep personalisation in the content itself. A section titled "Recently viewed" is expected and useful, while a pop-up that names the products someone looked at last week on another device reads as intrusive.

Privacy and consent in New Zealand

New Zealand's Privacy Act 2020 sets out how businesses collect, use and store personal information, including what you tell people about why you are collecting it. If you sell into Australia, Europe or the UK, those jurisdictions have their own rules on cookies and consent. Personalisation built on cookies and tracking needs a clear privacy policy, a consent mechanism suited to the markets you sell into and a setup that still works when a visitor declines. Get proper advice on the legal side. On the website itself, personalisation should improve the experience for visitors who consent and leave the site fully usable for those who do not.

Measure against a control group

Every personalisation should be tested against the version it replaced. The cleanest way is to hold back a share of the segment, show them the standard content and compare their conversion rate and revenue per session against the group that saw the personalised version. Without that control, any lift will be credited to personalisation when it may have come from a promotion, a season or a change in traffic.

Run each test long enough to cover a full buying cycle, and judge it on revenue per visitor or qualified enquiries. Clicks on the recommendation block tell you very little on their own. A recommendation that gets clicked and then leads nowhere has added a step without adding a sale. When a change holds up, keep it and move to the next segment. When it does not, remove it and keep the page simpler.

If you are not sure which segments behave differently on your site, a website audit will show you where the meaningful differences in your data sit before any personalisation tool is chosen.

Common questions

What is website personalisation?

Website personalisation means showing different content to different visitors based on information such as their location, how they arrived, what they have browsed and what they have bought. Common examples include local currency and shipping details, recently viewed products, related product recommendations and landing pages that continue the message of the ad a visitor clicked. The aim is to shorten the path to purchase for each group.

Does personalisation increase conversion rates?

It can, when it is aimed at segments that already behave differently and when each change is measured against a control group. The lift depends on your traffic, your catalogue and how well the personalised content is chosen. Poor recommendations and personalisation based on too little data can lower conversion, so test each change before rolling it out across the site.

Can you personalise a Webflow website?

Yes, Webflow sites can be personalised. Webflow offers an Optimize add-on for testing and personalising content by audience, including location, traffic source and behaviour. Third-party tools can also be added through custom code. For many Webflow sites, campaign-specific landing pages and location-aware contact details provide much of the benefit at lower cost, and they are a sensible place to start before adding a paid tool.

How does Shopify handle product recommendations?

Shopify's Search and Discovery app provides related and complementary product recommendations, which merchants can adjust manually to correct poor pairings. Many themes display these on product pages and in the cart. Apps can add more advanced recommendation logic, but each adds code to the storefront, so weigh the extra revenue against the effect on page speed.

Is website personalisation legal under the NZ Privacy Act?

Personalisation is allowed, but the Privacy Act 2020 governs how you collect and use personal information, including telling people what you collect and why. If you sell overseas, other rules on cookies and consent may apply. Use a clear privacy policy, a consent mechanism suited to your markets and a site that works fully for visitors who decline. Take legal advice on the specifics.

Summary

Website personalisation shows different content to different visitors based on their browsing behaviour, purchase history or location, such as showing a returning customer the products they looked at last time. Done well, it lifts conversion rates by 10 to 30 percent. Done badly, with irrelevant recommendations or tracking that feels intrusive, it lowers them.

Start with simple behaviour-based recommendations on Shopify or Webflow, measure the effect and expand only where the data supports it.

Yellow cooler filled with ice and assorted Slight Twist Brewing Co. cocktail cans including Pineapple Mango Lime Cherry, Cherry, and Citrus Spritz flavours.
Launched into supermarkets NZ wide, first page on Google
View of Earth’s horizon from space showing clouds and atmosphere below a starry black sky, with the Outlier logo and tagline 'Access the Impossible' above and a button labeled 'Unlock Microgravity'.
Raised US$7.35m at pre-seed, one of NZ's largest raises
Cliffside with trees and houses above overlooking calm water at sunset, with text 'Heritage Saunas' centered near the bottom.
2 x the standard industry enquiry rate, with no product ready
A gourmet dish on a grey plate includes a rolled meat portion with crispy skin, charred vegetables, a dollop of smooth brown sauce, garnished with small purple microgreens, and a small black sauce jug in the background.
10.4% CTR on google, Queenstown Top Restuarant

The kind of site you wish you'd launched sooner

Want more enquiries from the traffic you already have? Talk to us about optimisation
Clean Collective
103% conversion increase
Heritage Saunas
4.6% enquiry rate vs 2.7% benchmark
Outlier Space
raises US$7.35m FOR microgravity return platform
Deadly Ponies
Robomate
SHOPIFY MIGRATION
Slight Twist
CUSTOM SHOPIFY BUILD
The Supreme Group
The Chiller
Talk Recruitment
Grandpa's
See more work
arrow pointing right
Our trophy cabinet
5 CSSDA Award badgesAwwwards logoNewmarket Business Finalist Logo