A website that cannot be found is not an asset.

Our builds run $2,500 to $12,000 as a one-time project. An Essential build for a single-location service business runs $2,500 to $4,500. A Standard build with location pages and written content runs $4,500 to $8,000. A Complex build with booking integration, multiple locations or e-commerce runs $8,000 to $12,000. That is in line with national data: Clutch's 2026 survey found 61 percent of small business buyers spent under $10,000 on their most recent website project, down from 68 percent in 2023.

Most sites can be fixed, and we would rather do a $3,000 technical project than a $9,000 rebuild if the technical project solves it. The clearest signals that a rebuild is genuinely necessary are structural: a page builder generating markup nobody controls, a theme that has to be fought on every template, or a structure where separating your services means rebuilding the navigation, URLs and internal linking anyway. At that point the repair costs more than the replacement and still leaves you with a compromised foundation. The triage tool on this page gives you a starting read.

Because it has a directly measured revenue effect. A joint Deloitte and Google study found a 0.1-second load time improvement lifted retail conversion by 8.4 percent and average order value by 9.2 percent, with travel conversions up 10.1 percent. In a Google case study, Rakuten 24 ran an A/B test with Core Web Vitals optimization as the only variable and recorded a 53.37 percent increase in revenue per visitor and a 33.13 percent conversion rate increase. Vodafone Italy improved LCP by 31 percent and saw 8 percent more sales.

Three metrics Google measures on real visitors. Largest Contentful Paint should be 2.5 seconds or less — how long until the biggest element appears. Interaction to Next Paint should be 200 milliseconds or less — how fast the page responds to a tap. Cumulative Layout Shift should be under 0.1 — how much the page jumps while loading. Google assesses these at the 75th percentile of real visits over the previous 28 days, so three out of four visitors have to hit the threshold. Roughly 42 percent of mobile sites currently pass all three, against about 63 percent on desktop.

Because Google is not measuring your machine. It scores the 75th percentile of real visits, which in Hawaii means someone in a rental car in Hana on one bar, a visitor on hotel wifi shared with four hundred guests, or a resident on a four-year-old Android. Your site can feel instant on office wifi and be comprehensively failing the people who were about to call you. Check the Core Web Vitals report in Google Search Console, which uses real field data rather than a lab score.

It is a technical problem, not a reason to use fewer photos. Your imagery is the product, particularly in tourism, and cutting it would cost you more than the speed gain is worth. The resolution is modern formats, correctly sized variants served by device, explicit width and height on every image so nothing shifts, lazy loading below the fold, and preloading the hero image that determines your LCP. Cloudflare found image optimization alone can improve LCP by 30 to 50 percent, and responsive modern-format images cut mobile page weight by around 34 percent.

Almost certainly, if it loads the default way. Third-party widgets ship JavaScript, and JavaScript is the leading performance bottleneck on the web — HTTP Archive puts the average site at around 489KB, up 23 percent since 2022. That is why INP scores collapsed when it replaced First Input Delay: FID only measured the first interaction, INP measures all of them. The fix is sequencing rather than removal. Load the booking widget on scroll or tap rather than on page load, defer chat until the page is interactive, and render reviews as static text instead of a carousel.

We build fast, hand-coded static sites by default because they consistently outperform on Core Web Vitals and have almost no maintenance surface. Framework-based sites pass Core Web Vitals at around 58 percent against roughly 38 percent for WordPress, largely due to architectural advantages in code splitting, image handling and rendering. That is not a claim that WordPress cannot be fast, but the default path leads somewhere slow. If you need a specific WordPress integration we will talk it through honestly rather than force our preference on you.

Text and image updates, yes, and we show you how at handover. Restructuring layouts in a visual drag-and-drop editor, no — that is exactly the capability that produces the bloated markup and JavaScript weight causing most INP failures. It is a real trade-off and worth being upfront about: you get a considerably faster site and give up the ability to rearrange page layouts yourself. If self-editing layouts matters more to you than performance, we are probably not the right shop and we would rather say that now.

You do, completely. The domain, hosting, analytics and every account are registered in your name with us granted access, and handover includes every credential. There is no monthly fee to keep the site online and no licence that expires. If you want to take the whole thing to another provider next month you can, without asking us. Plenty of Hawaii businesses are currently in the opposite arrangement, and if that is you, moving away from it is worth doing regardless of who you hire next.

Four to eight weeks for most projects. The bottleneck is almost always content and photography rather than the build itself, which is why we write the content rather than sending you a questionnaire and waiting. The stages are audit and architecture, content, design and build, a technical pass covering schema and performance verification, then launch with a 30-day window to fix anything that surfaces.

We write it, working from an interview with you rather than a form. The half-finished site sitting with placeholder text on three pages, waiting for content the owner never has time to write, is the most common failure mode in this industry and we designed the process specifically to avoid it. If your budget genuinely cannot cover written content we will scope a smaller site rather than launch an incomplete one.

Only if it is done carelessly, and this is the single most common way redesigns go wrong. Rankings drop after a rebuild when URLs change without redirects, when content that was ranking gets cut in the name of a cleaner design, or when schema and internal linking are not carried across. We map redirects during the architecture stage before anything is designed, preserve URLs that are already earning, and verify indexing after launch. If you have had an unexplained ranking drop after a previous redesign, that is usually what happened.

Structured data on every page rather than as an afterthought, question-phrased headings with complete self-contained answers underneath, business facts stated as text rather than baked into images, an llms.txt index at the root, and robots.txt explicitly permitting the AI crawlers instead of blocking them by accident through a plugin default. None of it is expensive during the build and all of it is expensive to retrofit onto a structure that was not designed for it.

A well-built site gets you a foundation and often some quick wins, particularly if what you had before was slow or thin. It does not get you ongoing content, review generation, local prominence, or the compounding work that wins competitive categories. For a thin neighbor-island category a good site plus a complete Google Business Profile is sometimes genuinely enough, and we will say so. For a contested Oahu category it will not be. There is no requirement to buy both, and the build is not priced on the assumption that you will.

Yes — Hawaii Island, Oahu, Maui, Kauai, Molokai and Lanai, plus mainland businesses targeting Hawaii customers. Web design work is delivered remotely regardless of island, with calls and screen shares rather than in-person meetings, which keeps the cost down and the timeline tight. Multi-island businesses need per-location page structure so locations support each other rather than competing, and that is built into the architecture stage.