How to tell if your website design is costing you conversions

Work out the number before you change anything

Conversion rate is the count of people who did the thing you wanted, divided by the number of sessions, over the same period. For a service business the thing you wanted is almost always a phone call or a submitted form.

The part most sites get wrong is the calls. In Google Analytics 4, a tap on a phone number link fires as a click event, and you have to mark it as a key event yourself before it appears in any report. Until you do, a business that gets ninety percent of its work by phone will look like it converts at nearly zero. If you want the calls attributed to specific pages and campaigns, use tracking numbers that swap in dynamically rather than a static number in the header.

What a normal conversion rate looks like

Local service sites usually land between two and five percent. Apparel and other retail sits lower, often one to three. Anything above eight is worth checking rather than celebrating, because it usually means most of your traffic is people typing your business name into search, and returning customers convert several times better than strangers.

That is the honest limitation of the benchmark. Traffic mix moves the number more than design does, so compare your site to itself over time rather than to somebody else's figure.

Speed is a design decision, not a hosting one

The slow pages we inherit are almost never slow because of the server. They are slow because of an autoplaying hero video, product photos exported at four megabytes, three separate icon fonts, and a chat widget that loads a second and a half of third-party script before it draws a bubble nobody clicks.

Run one page through PageSpeed Insights and look only at Largest Contentful Paint, which is roughly when the main image or headline finishes drawing. On a phone on a weak signal, that is the moment your visitor decides whether to wait. Every one of the items above is something a designer chose and can remove.

Every form field you add costs you submissions

A quote form with nine fields will collect fewer leads than the same form with three, and the leads it does collect are not meaningfully better qualified. Name, phone, and a free text box gets you further than a required dropdown of eleven service categories that the customer has to guess between.

On mobile, check that the phone field brings up the number keypad and the email field brings up the keyboard with the at sign. That is one attribute in the markup, it takes a minute, and it is missing on most sites we audit. Also check what happens after submission. A page that quietly reloads with no confirmation makes people submit twice or leave assuming it failed.

What people look for before they call a stranger

License and insurance numbers where a homeowner can see them. Photographs of your actual crew and trucks rather than stock images of models in clean coveralls. The towns you cover, named. Some indication of price, even a range, because the alternative is that the visitor calls three competitors to find out.

None of this is decoration. It is the evidence a cautious buyer needs, and leaving it out is a design decision as much as choosing a typeface is.

You probably cannot A/B test this

Split testing needs a few hundred conversions per variant before the result means anything. A local site producing forty leads a month will never reach that inside a useful timeframe, and the tools will happily show you a confident winner anyway.

So change things sequentially instead. Make one substantial change, leave it a full month, and compare booked jobs rather than clicks. It is slower and less impressive than a dashboard, but it is the only reading you can trust at that volume. When we rebuild pages at Digital Branding Blocks, that is the pace we hold clients to.

What does work at low volume is watching people use the site. A session recording tool will show you twenty real visits inside an hour, and the pattern is usually obvious by the tenth. Everybody scrolls straight past the section you are proudest of. Three people tap an image that is not a link. One person fills in half the quote form and abandons at the field asking for a street address. That is not statistically valid either, but it points at specific problems rather than at percentages.

Where the gains usually are

Almost all of it is on the phone. Check your analytics for the mobile share of sessions, which for most local trades in New Jersey now runs well past two thirds, then look at your own site on an actual phone rather than a narrowed browser window.

The recurring wins are the same few things: a tappable number in the header, one clear action per page repeated near the bottom, a shorter form, and removing whatever competes with the button you care about. That is usually where the first point or two of conversion rate comes from, before anything more interesting gets a chance.