Conversion Rate Optimisation (CRO): A Practical Guide for NZ Businesses

CRO is a structured way of improving the share of visitors who take a valuable action on your website. It combines analytics (what people do), research (why they do it) and changes to the site (design, copy, speed, forms, checkout), then measures whether those changes helped.

It is not the same as SEO or advertising. SEO and ads bring people to the site. CRO decides how many of them turn into customers once they arrive. The two work together: a page that converts well makes every dollar of ad spend and every hard-won ranking worth more.

A conversion is whatever action moves a visitor closer to paying you. It depends on the kind of site:

Pick one main conversion and a few smaller ones. Tracking the smaller steps shows you where people drop out, not just that they did. If you are not sure your tracking is counting these correctly, fix that first; optimising against bad data is guesswork with extra steps.

Divide conversions by sessions and multiply by 100.

If your site had 4,000 sessions last month and 60 enquiries, your conversion rate is 60 ÷ 4,000 × 100 = 1.5%.

A few things trip people up:

• Sessions, not users. Most platforms, including Shopify, calculate conversion rate per session. One person visiting three times is three sessions.
• Bots and internal traffic inflate sessions and make your rate look worse than it is. Filter them out.
• Double-counted conversions (a thank-you page that fires twice, for example) make it look better than it is.
• Look at segments, not just the total. Mobile vs desktop, new vs returning, and each traffic source often tell very different stories.

There is no single good number, but global ecommerce averages give you a rough frame. Shopify's NZ benchmark guide (updated August 2026) cites Statista data showing about 1.4% of global ecommerce visits converted in early 2026, and Dynamic Yield's benchmark of around 2.66% across more than 400 brands. Shopify's own summary is that averages "tend to sit around 2% to 3%".

Two cautions. First, the averages vary a lot by industry, price point and device, so a luxury furniture store and a supplement brand should not expect the same rate. Second, there is no reliable published benchmark for New Zealand alone. Your own trend over time is a better yardstick than anyone's average. Our ecommerce conversion rate benchmarks for NZ goes deeper on this.

Most lost conversions come from a handful of predictable places. These are the ones with the best evidence behind them.

Slow pages, especially on mobile

Speed changes behaviour. In Deloitte's Milliseconds Make Millions study for Google (2020), a 0.1 second improvement in mobile site speed was associated with an 8.4% increase in retail conversions and a 9.2% increase in average order value, across 37 brands. That is an observational study, not a controlled experiment, but the direction is consistent with Google's earlier research: their 2017 analysis found the probability of a mobile visitor bouncing rose 123% as page load time went from one second to ten.

In our own work, page speed is often the first thing worth fixing. On one client store we took load time from 11.3 seconds to 1.9 seconds within a month, before touching anything else.

Checkout and form friction

Baymard Institute puts the average documented cart abandonment rate at 70.22%, averaged across 50 studies. Many of those people were only browsing, but plenty leave for fixable reasons. In Baymard's 2026 survey of US online shoppers, the top reasons included:

Baymard's benchmark also found the average US checkout has 23.48 form elements, against an achievable 12 to 14. The same logic applies to enquiry forms on service sites: every field you ask for is a reason to stop.

Pages that don't answer the visitor's question

People land with a question and leave when the page doesn't answer it quickly. Common culprits: a headline that describes the company instead of the problem it solves, no price guidance at all, proof (reviews, examples) buried at the bottom, and a call to action that asks for too much too soon.

Trust gaps

Visitors look for signs you are real and reliable before they hand over money or details. Real reviews, clear contact details, a visible returns or guarantee policy, recognisable payment options and a site that looks maintained all reduce that hesitation.

Our website conversion audit checklist walks through 30 checks across these areas that you can run yourself.

Use a framework so you are not guessing. One of the most widely used is the LIFT model from Conversion (formerly WiderFunnel). It scores a page on six factors:

Work through your highest-traffic pages first, note the problems under each factor, then rank the fixes by likely impact and effort. A slow product page that gets half your traffic beats a perfect redesign of your About page every time.

1. Check your tracking. Make sure conversions are counted once, correctly, on every device.
2. Find the leaks. Use analytics to see which pages and steps lose the most people.
3. Find out why. Heatmaps, session recordings, on-site surveys, customer interviews and a heuristic review (an expert walking the site with a framework like LIFT).
4. Form a hypothesis. "Showing delivery costs on the product page will reduce checkout drop-off, because surprise costs are the top reason people abandon."
5. Make the change. Test it with an A/B test if you have the traffic, or ship it and monitor if you don't (more on that below).
6. Measure and repeat. Keep what works, roll back what doesn't, and move to the next item on the list.

If you want to see what a professional version of steps 2 and 3 looks like, see what's included in a CRO audit.

A/B testing is the most reliable way to prove a change worked, but it needs more traffic than most small sites have.

An A/B test splits visitors between two versions of a page and compares the results. To trust the answer you need enough conversions in each version. Evan Miller's widely used sample size calculator shows that even a large effect (a 20% baseline rate moving by 5 percentage points) needs around 1,030 visitors per version. Smaller, more realistic improvements on lower baseline rates need many thousands. Seer Interactive's rule of thumb is that a test becomes viable with at least a few thousand monthly visits and around 100 conversions per version.

Two mistakes to avoid if you do test:

• Don't stop the test early because it looks like a winner. Evan Miller showed that checking results repeatedly and stopping at the first "significant" moment can produce false winners around a quarter of the time in his example. Decide the sample size in advance.
• Don't test tiny changes on small traffic. Test bold changes, fewer versions at once, and earlier steps in the funnel (like add to cart) where there are more events to count.

CRO without A/B testing

If your site is too small to test, you can still do CRO. Base changes on solid research (analytics, recordings, customer feedback and established usability evidence like Baymard's), make one meaningful change at a time, and compare performance over comparable periods. It is less certain than a controlled test, so be honest with yourself about seasonality and other changes, but it is far better than leaving known problems in place.

CRO is usually sold as a one-off audit or a monthly retainer. We have broken down typical NZ pricing models and published ranges in how much CRO costs in NZ.

Timelines depend on how much there is to fix and how much traffic you have. Fixing obvious problems like a slow site or a broken mobile checkout can show up in your numbers within weeks. A proper testing programme is an ongoing loop measured in months, because each test needs time to collect enough data.

Start in-house if you have the time and someone comfortable with analytics. The checklist above and a free tool like Microsoft Clarity for recordings and heatmaps will take you a long way.

Bring in help when you have done the obvious fixes and progress has stalled, when you don't trust your tracking, or when development work is the bottleneck. A specialist will usually spot issues faster and can make the changes as well as recommend them. If you want an outside view first, our free conversion audit shows the biggest opportunities on your site, and our conversion rate optimisation service covers the full loop for Shopify and WooCommerce stores.