Every agency says design lifts conversions. Few will tell you which parts of design do it. The honest answer is that a small number of changes carry almost all the effect, and most of them are unglamorous: the number of fields in a form, whether a price is visible, how fast the first meaningful pixel arrives on a mid-range Android phone. This guide covers what we would change first, the patterns we reach for, what US agencies charge, and how to tell whether any of it worked.
Good UX raises conversions by removing friction at the points where people decide: page load speed, clarity of the offer, price comparison, and the form itself. Cutting a form from nine fields to four, publishing pricing, and meeting Google's Core Web Vitals thresholds move numbers. Cosmetic redesigns usually do not.
Where conversion is actually won
Most pages do not lose buyers because they look dated. They lose buyers at four or five specific moments: the second the page starts loading, the moment someone tries to work out what you sell, the moment they compare you against an alternative, the moment they meet a form, and the moment they are asked for money or a phone number. Useful UX work targets those moments and leaves the rest alone.
That framing changes how you brief ux design agencies. Ask for "something modern" and you will get a modern site and a flat conversion rate. Ask for "our demo form converts at 1.8 percent, we think the nine fields are the problem, and we want to test a two step version", and you get work someone can actually check three weeks later.
What can anyone honestly claim about UX and conversion?
Plenty of agencies will quote you a percentage from a study nobody can find. We would rather give you the two things that are publicly verifiable and then be clear about where our own numbers come from.
First, Google publishes its Core Web Vitals thresholds. A page is considered good when Largest Contentful Paint happens in under 2.5 seconds, Interaction to Next Paint stays under 200 milliseconds, and Cumulative Layout Shift stays under 0.1. Those are documented numbers, not opinions. Second, page experience is part of how Google evaluates pages, so speed and stability are not only a conversion question, they touch acquisition too.
Everything else we say about lift comes from our own project work, and we label it that way. In our own redesigns, the changes that move the needle most reliably are form reduction, clearer pricing presentation, and fixing mobile layout shift. The changes that most often do nothing measurable are illustration refreshes, new icon sets, and animated hero sections. If an agency promises you a specific percentage before looking at your analytics, treat the number as marketing copy.
Before any design work starts, pull four numbers: conversion rate by device, form abandonment rate by field, mobile LCP on your two highest traffic templates, and scroll depth on the page you most want people to read. If you cannot produce those four, the first two weeks of the project are measurement, not design. Redesigning without a baseline means you will never know whether you helped.
Common UX problems and the conversion symptom they cause
Symptoms are easier to spot than causes. This table maps what you see in analytics back to the design decision that usually created it.
| UX problem | Conversion symptom | The fix we would make |
|---|---|---|
| Hero says what you are, not what you do for the visitor | High bounce on paid traffic, low scroll depth | Rewrite to outcome plus proof, keep it under 12 words, move the offer above the fold |
| Form asks for 8 or more fields on first contact | Form starts high, completions low, big drop at phone or company size | Cut to 3 or 4 fields, move the rest to a second step or the sales call |
| Pricing hidden behind a contact request | Long sessions, repeat visits, few conversions | Publish a range or a starting price with a comparison table |
| Layout shifts as images and fonts load (CLS above 0.1) | Mis-taps on mobile, higher exit rate in the first 5 seconds | Reserve image dimensions, preload the hero font, stop injecting banners after render |
| Primary action competes with 5 other buttons | Clicks spread thin, low click-through on the action that matters | One primary button per screen, everything else becomes a text link |
| Mobile CTA sits only at the top of the page | Mobile converts at a third of desktop despite similar traffic | Sticky bottom bar with the single next step, visible after the first scroll |
| Errors appear only after submit | Repeat submissions, users abandoning on the second attempt | Inline validation as the field loses focus, with the specific fix stated |
Design patterns that tend to earn their keep
These are patterns, not client stories. We describe them so you can picture the change and decide whether it applies to your page.
The pricing page that puts comparison above the fold
A common pricing layout opens with a headline, a paragraph of positioning, and then the plans. Move the plan comparison up so the first thing on screen is three columns with prices and the two or three differences that matter. Positioning copy drops below. People arriving on a pricing page have already decided to evaluate you, and making them scroll past reassurance they did not ask for costs you attention.
The two step form
Split a long form so step one asks for name and email, and step two asks the qualifying questions. Two things happen. The commitment on screen one is small, and you capture a partial lead even when someone leaves at step two. For sales teams that follow up on partials, this alone changes the shape of the pipeline.
Proof placed next to the decision, not on a separate page
Testimonials sit on a testimonials page that nobody visits. Take the two most specific ones, the ones with a job title and a concrete outcome, and place them beside the form and beside the pricing table. Proof works where doubt appears.
Sticky mobile action bar
On phones, a single fixed bar with one action ("Get a quote" or "Book a call") removes the need to scroll back up. Keep it 56 to 64 pixels tall, keep the label a verb, and do not stack two competing buttons in it.
Product pages that answer shipping and returns before the add to cart
On ecommerce, the questions that stop a purchase are delivery date, return window, and size or fit. Putting those three answers within one screen of the buy button removes the tab someone would otherwise open to check. Our Shopify development builds treat that block as required, not optional.
How would we run a UX project that has to pay for itself?
- Week 1, measure. Analytics review, funnel by device, Core Web Vitals from field data, form field drop-off, and 6 to 10 session recordings on the highest value template.
- Week 1, listen. Read 30 to 50 sales emails or support tickets. The objections written there become the copy on the page.
- Week 2, prioritise. List every issue found, score by traffic affected and effort required, and pick the five that sit high on both.
- Weeks 2 to 4, design. Wireframe the templates that carry the traffic, usually home, one service or category page, one product or pricing page, and the form. Not 30 screens.
- Weeks 4 to 6, build and instrument. Ship with event tracking already in place: form starts, field-level abandonment, CTA clicks, scroll milestones.
- Week 6, verify performance. Confirm LCP under 2.5 seconds on a mid-range Android over 4G, INP under 200 milliseconds, CLS under 0.1.
- Weeks 7 onward, test. One meaningful test at a time. If traffic is under roughly 10,000 sessions a month, run sequential before and after comparisons over full weeks rather than pretending you have enough volume for a split test.
That last point gets ignored a lot. Small businesses are told to A/B test everything, then run a test with 400 sessions and read noise as a result. With modest traffic, make fewer and larger changes and judge them over four full weeks against the same period before.
What does UX work cost in the US?
Ranges below reflect what we see quoted across the US market and what we charge for comparable scopes. Prices vary with team seniority and how much research is included.
| Engagement | Typical US range | Timeline | Best when |
|---|---|---|---|
| Conversion audit, no design | $1,500 to $5,000 | 1 to 2 weeks | You have traffic and want a prioritised list before committing budget |
| Single template redesign (pricing, landing, product) | $3,000 to $12,000 | 2 to 4 weeks | One page carries most of your revenue |
| Small business site redesign, 5 to 8 pages | $1,500 to $15,000 | 3 to 6 weeks | The whole site is dated and the funnel is simple |
| Full UX programme with research and testing | $25,000 to $80,000 | 3 to 6 months | Multiple funnels, real traffic, an in-house team to hand off to |
| Ongoing optimisation retainer | $2,000 to $8,000 a month | Rolling | You want continuous testing rather than a one-off project |
The low end of the redesign row is not a typo. Our current five page website offer covers design, development and copywriting at $1,700 instead of $3,000, which suits a service business whose funnel is one page and one form. If your funnel involves configurators, logins or a catalogue of 2,000 SKUs, you are in the higher rows and should plan accordingly.
If your budget only stretches to one thing this quarter, spend it on the form and the page directly before it. That pair usually sits between you and every conversion on the site, and it is the cheapest surface to change. A hero redesign looks better in a portfolio. The form change shows up in your CRM.
UX design agency or product design agency?
The labels overlap, so use scope to tell them apart. A UX design agency working on a marketing site is optimising a short, mostly linear journey where the goal is a form, a call or a purchase. A product design agency is designing software that people use repeatedly, where the questions are onboarding, information density, permissions, empty states and what happens on the four hundredth visit rather than the first.
If your problem is that visitors are not becoming leads, you want the first. If your problem is that customers sign up and then stop using the thing, you want the second. Plenty of teams, ours included, do both, but the discovery work is genuinely different and so is the price. Our conversion optimization work sits on the marketing side, and our website design and development team handles the build so the design does not degrade in implementation, which is where a surprising amount of UX quality goes to die.
Frequently asked questions
Does better UX design really increase conversions?
Yes, when it removes specific friction. Fewer form fields, clearer pricing, faster loading and one obvious next step are the changes that show up in analytics. New illustrations and animated heroes usually do not. Be sceptical of any agency quoting you a fixed percentage lift before it has looked at your funnel data, because that number cannot exist yet.
What are Google's Core Web Vitals thresholds?
Google publishes three: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, and Cumulative Layout Shift below 0.1. They are measured from real visitors, weighted heavily toward mobile. Page experience is part of how Google evaluates pages, so hitting these thresholds matters for acquisition as well as for the people already on your site.
How much do ux design agencies charge in the United States?
Audits without design work usually run $1,500 to $5,000. One redesigned template costs $3,000 to $12,000. A small business site of five to eight pages spans roughly $1,500 to $15,000 depending on how much research and copywriting is included, and a full research-led programme runs $25,000 to $80,000 over three to six months. Retainers for continuous testing commonly sit at $2,000 to $8,000 a month.
How long before a UX change shows results?
Performance fixes appear in field data in about 28 days, because Core Web Vitals reporting uses a rolling 28 day window. Conversion effects need volume, so compare four full weeks after launch against the equivalent four weeks before, matching for seasonality. If your site sees under roughly 10,000 sessions a month, skip split testing and make fewer, larger changes you can judge sequentially.
What is the difference between a UX design agency and a product design agency?
A UX design agency generally works on marketing journeys: short visits ending in a form, a call or a purchase. A product design agency designs software people use repeatedly, so onboarding, permissions, empty states and information density dominate the work. If visitors are not becoming leads, hire the first. If customers sign up and then go quiet, hire the second.
Which single UX change should a small business make first?
Fix the form and the page directly before it. Almost every conversion on a small business site passes through that pair, and it is the cheapest thing to change. Cut to three or four fields, validate inline rather than after submit, put your two most specific testimonials beside it, and add a sticky action bar on mobile so the next step never scrolls away.
The takeaway
Treat UX as a list of specific frictions with owners and deadlines, not as a look. Measure four numbers before anyone opens a design tool, change the form and the page before it first, hold the page to Google's published performance thresholds, then judge the work against four full weeks of before and after data. If a change cannot be checked, it was decoration.