What a redesign actually costs

The quote says forty thousand. The invoice, by the time the site is live and stable, says closer to sixty-five. Nobody lied on the way there — every change order was approved, every extra was requested by someone on the client side, and the agency's own line items matched what was delivered for that price. The gap wasn't dishonesty. It was scope that was never priced in the first place, because it was never the agency's job to price it.

That is worth conceding before anything else, because it is the fair version of the agency's side of this: a design-and-build quote prices design and build. It does not price the things the client controls — the content that has to be written, the hours internal stakeholders spend reviewing drafts, the weeks after launch when the new site underperforms the old one while search engines and visitors recalibrate. Those costs are real, they are large, and they belong to the client's budget whether or not anyone put them in a spreadsheet. The problem isn't the quote. It's that most people read a design-and-build quote as a project budget, when it's actually about sixty per cent of one.

What the line items actually cover, versus what the project needs

A typical mid-market redesign quote breaks down into something like: discovery and strategy, information architecture, visual design across a defined number of templates, front-end and back-end build, QA, and a launch window. That is a real, complete description of the agency's work. It is not a complete description of the project.

Missing from it, almost every time: writing or rewriting the actual page copy beyond a placeholder pass; sourcing or producing new photography and graphics; migrating existing content into the new structure and mapping old URLs to new ones with redirects; the internal hours spent in review cycles, feedback rounds and sign-off meetings across however many stakeholders the organisation has; and anything that happens to performance in the weeks after go-live. Some agencies will quote a content package as an add-on line; most will quote "content migration" as a checkbox that means "we will move what you give us," not "we will write what you need."

The itemised comparison, for a mid-sized brochure site — roughly twenty to thirty pages, one content type beyond the basics:

Line item	In the typical quote	Actually required
Design and build	Yes, in full	Same
Content strategy document	Sometimes	Yes
Page copy, written or rewritten	Rarely, or one placeholder pass	Yes, every page
Photography / imagery	Stock only, usually	Often custom
URL mapping and redirects	Partially	Fully, or search rankings break
Internal review hours	Not priced	Real, and large
Post-launch monitoring and fixes	First 30 days, sometimes	60–90 days realistically
Performance dip	Never	Should be, as a named risk

None of the items in the right column are exotic. They are the ordinary cost of the same project the quote was written for. They just don't show up on an agency invoice because an agency invoice only ever bills for what the agency does.

Content is the line item everyone underprices, including themselves

Ask a marketing lead how much the new site's content will cost and the honest answer, most of the time, is "nothing, I'll write it." That answer is wrong in the way that "nothing, I'll do it on the weekend" is wrong for a home renovation — the work still takes the hours it takes, the hours just move off the invoice and onto a calendar that was already full.

Cost it properly and the number changes. Take the loaded hourly cost of the person who will actually write the copy — not a junior copywriter's day rate, the real cost of the marketing director's or product manager's time, because that is usually who ends up doing it — and multiply by a realistic hour count per page: research the audience and the competitive positioning once, then something on the order of two to four hours per page for a first draft plus revisions, more for pages that need input from sales or product. Twenty-five pages at three hours each is seventy-five hours. At almost any loaded internal rate, that is a five-figure cost sitting inside a project that was budgeted as a design-and-build number with a content line that said "TBD."

The reason this matters beyond the arithmetic: content that is costed at zero gets treated as free, which means it gets deprioritised against everything else on that person's calendar, which is the single most common reason a redesign timeline slips. The fix isn't clever. It's pricing the writer's time the same way the agency's time was priced, before the schedule is set rather than after it's already broken.

Internal hours are a cost even though no invoice arrives for them

Every stage of a redesign generates a review cycle: someone from marketing, someone from sales, sometimes legal, sometimes an executive with opinions about the hero image, looking at a deliverable and either approving it or sending it back. Multiply the number of people in each review by the number of review stages — discovery sign-off, IA sign-off, wireframe review, two or three rounds of visual design, content review page by page, and a pre-launch QA pass — and the internal hour count on a mid-sized project routinely runs into the hundreds, spread across people whose primary job is something else.

None of that appears on the agency's invoice, because none of it is the agency's cost to bear. It is nonetheless a real cost to the business: those are hours that weren't spent on whatever those people's actual roles are, and if the organisation is disciplined enough to track opportunity cost, it belongs in the same total the design-and-build quote sits inside.

The post-launch dip is the cost nobody plans for because nobody expects it

The redesign decision covers why this happens: a site that has been live for years has accumulated search authority and behavioural data that a rebuild resets, even when the migration is technically clean. What that piece doesn't put a number on, and what belongs in a budget rather than a warning, is that the dip has a duration, and treating it as an open-ended risk rather than a bounded one is what causes the panic response — a second emergency project to "fix" a redesign that was actually recovering on schedule.

The budgeting move that actually works is to hold a specific number of weeks — not a vague caution — as a named contingency line, sized to how much of the URL structure and content is changing. A visual refresh with the same URLs and largely the same copy carries a small version of this risk. A full rebuild with a new information architecture and renamed pages carries a much larger one, because every redirect is a chance for a page that used to rank to lose the signal it built up. Budgeting for the dip means two things: holding cash or headcount in reserve for the weeks after launch rather than assuming the project ends at go-live, and agreeing in advance what "recovering on schedule" looks like so a normal dip doesn't get mistaken for a failed project.

Three worked totals

None of these are quotes — they're the arithmetic of adding the missing line items to a plausible agency quote at three sizes, with a contingency line stated rather than implied.

Small site, ten to fifteen pages, single content type. Agency quote: $15,000. Content (internal hours, costed): $6,000. Internal review hours across the project: $3,000. Contingency for the post-launch dip: $2,500 (roughly 17% of the agency line). Total: $26,500 — about 77% over the quoted figure.

Mid-sized site, twenty-five to thirty pages, one added content type. Agency quote: $45,000. Content: $16,000. Internal review hours: $8,000. Contingency: $9,000 (20%, because more pages means more redirect risk). Total: $78,000 — about 73% over the quoted figure.

Larger site, sixty-plus pages, new information architecture, migrated CMS. Agency quote: $120,000. Content: $40,000. Internal review hours: $22,000. Contingency: $30,000 (25%, because a full IA change carries the largest dip risk of the three). Total: $212,000 — about 77% over the quoted figure.

The pattern across all three sizes is consistent: the real total lands somewhere between seventy and one hundred per cent over the agency's own number, and the contingency line grows as a share of the total the more structural the change is. That consistency is the actual planning takeaway — not the dollar figures, which will move with your market and your team's rates, but the multiplier. A design-and-build quote is roughly sixty per cent of the project.

What this changes about how the brief gets written

None of this argues against redesigning. It argues against approving a redesign against a number that was never meant to describe the whole project. The three triggers that justify a rebuild in the first place don't get weaker because the true cost is higher — if the content model has genuinely outgrown the business, that's still true at $78,000 and it was still true at $45,000. What changes is who signs the number, and whether the person signing it knows the agency's line is the floor, not the total. A company that wants the fuller comparison — what a site costs not just to build but to run for the years after — has that laid out separately in what a company website costs over three years, which is the number that actually determines whether the redesign was worth it.

The agencies quoting these projects are not the ones getting this wrong. They are pricing exactly what they were asked to price. The error happens one step earlier, when a design-and-build number gets treated as a project budget instead of as the line item it actually is.