Turn an old site into something that works.

What renewal conversations usually include

“It’s slow” arrives with almost every brief. The cause is not one thing. Images may be heavy, or the first byte (TTFB) may be late. Rebuilding without looking leaves the same wait.

Broken phone layouts usually mean the original build was checked on a desktop, or later edits added only padding and images. Fixing the desktop view does not fix a handset.

“We cannot update it ourselves” is an operations problem. Every change goes back to the vendor, only a former employee has the login, or the admin UI is still in a language nobody reads.

When applications do not arrive, the jobs page is often buried, the form is only an email address, and there are no faces. After extra language versions, weight often comes from loading the same assets twice per language, not from the translations themselves.

We measure first

The first deliverable is not a design. We measure display speed (TTFB and LCP), the page map and duplicate URLs, language tags (hreflang), the sitemap, and how search engines see the site. Starting from measurement, not from a feeling, is what sets this page apart.

Felt slowness is often a cache setting, not page count. A site with a dozen pages can still rebuild HTML on every visit, leave images uncompressed, and never tell the browser it may reuse them. Visitors download the site from scratch each time. Fixing cache and images can help more than deleting pages.

We return the measurements as a list of pages to keep, merge, or redirect. Taste in visuals comes after that list.

Rebuild, or keep and repair?

Whether to stay on WordPress turns on who edits. If someone in-house still posts news and case studies in the admin UI, repairing the theme and the speed can be enough. If updates are a few times a year and each one is a brief to a vendor, a static setup often means fewer outages and fewer hands.

Page count matters. Dozens of pages on one template are safer to migrate than a site where every page invented its own heading style. Function also matters. Login, booking, or payments treated as a “looks-only” renewal will break what already works. If function is the center, we split the brief from new production and from business systems before we cut scope.

How a move runs

We inventory the live site: URLs, titles, pages that still get search traffic, pages nobody opens. The new map is drawn from that table. We migrate copy and photos that still earn their place, and we write only what is missing. Before launch we put redirects from old URLs and check internal links. After launch we check that old addresses in search still land, and that speed on the main pages has not gone backwards.

Failed moves are less about color preference than about dropping old addresses. Bookmarks and search visits are kept with redirects.

Recruiting sites

A single “Careers” page at the end of the company site is often a weak path. Roles, working style, the selection steps, and the form should follow the order a candidate actually reads.

A separate recruiting site versus a path under the current domain depends on how job boards send traffic. If those links land on the company home, extra clicks sit between the ad and the apply button. If we add a recruiting site, company facts stay consistent with the corporate site so the same facts are not maintained twice.

Timeline and cost

Four to eight weeks, from ¥500,000. Page count, how many URLs move, and whether hiring is split off change the number. All listed prices exclude tax. A URL list and the screens that currently hurt help us bound the measurement.

Questions we hear

Renewal starts with a measurement.

We split “old” and “slow” into numbers. First consult and quote are free.