Publishers rebuild the front page every three years and break fifteen years of archive doing it. That archive is most of what still earns.
Media and publishing website design in New York costs $15,000 to $80,000. Two things separate this work from ordinary content sites: the revenue model is subscriptions, advertising or both and each one shapes the build differently, and the archive is usually the single largest source of traffic, which makes URL continuity more important than any design decision.
Years of articles carry links and rankings that no new content will replace. Every URL maps forward before a rebuild starts. Publishers who skip this discover the cost in the quarter after launch, and it is not recoverable.
A metered paywall, a hard wall, a membership or an ad-funded free site are four different products. Registration walls, metering rules and what search engines are allowed to see all need deciding before design, not after.
Drafting, editing, scheduling, embargoes, corrections and rights. If publishing something takes eleven clicks, editors work around the system and the system stops reflecting reality within a month.
If ads fund the site they have to be designed in: fixed slots that do not shift the layout, lazy loading below the fold, and a hard limit on what third parties may do. Unmanaged ad tech is the usual reason publishing sites feel broken.
Article pages carry images, embeds, video and ad calls. Getting a heavy page to render quickly on a phone is the core engineering, and it affects both reading time and ad viewability.
Articles, authors, topics, series and issues as data rather than as a folder of pages. That is what makes archives navigable, newsletters automatable and syndication possible.
| Project | Range (USD) | Timeline | What you get |
|---|---|---|---|
| Independent publication | $15,000 to $30,000 | 8 to 16 weeks | Article and index templates, editorial workflow, newsletter, archive migration |
| Subscription publication | $30,000 to $55,000 | 14 to 24 weeks | Above plus paywall, metering, accounts, billing integration, retention flows |
| Larger publisher | $55,000 to $90,000 | 20 to 36 weeks | Multi-title structure, ad stack, rights handling, syndication, phased migration |
| Archive migration alone | $5,000 to $25,000 | 4 to 12 weeks | URL inventory, mapping, content transfer, verification, monitoring |
| Care and development | $1,500 to $8,000 per month | Ongoing | Releases, editorial requests, performance, ad and paywall tuning |
Archive migration is the line most publishing quotes underprice. On a fifteen-year archive it is frequently the largest single piece of the project.
| The question | A good answer | Walk away if |
|---|---|---|
| How do you migrate the archive? | URL inventory first, mapped before build | At the end, with redirects |
| Have you built a paywall before? | Named projects and what the metering rules were | We can add one |
| What can ad tech do to the page? | A stated limit, enforced in the build | That is the ad team's decision |
| How many clicks to publish? | A number, demonstrated on staging | Editors will get used to it |
Ask about the archive first. A supplier who treats fifteen years of URLs as a launch-week task has never watched a publisher lose them.
| Problem | Consequence |
|---|---|
| Archive URLs changed without mapping | Years of rankings and links, gone in one night |
| Paywall rules decided after design | Registration flows retrofitted badly into finished screens |
| Ad slots that shift the layout | Readers tapping the wrong thing, and a poor layout score |
| Editorial workflow nobody consulted on | Editors working around the system within a month |
| Articles built as pages rather than data | No series, no topic pages, no automation |
| Heavy embeds with no budget | Article pages that take eight seconds on a phone |
| Problem | Consequence |
|---|---|
| Archive URLs changed without mapping | Years of rankings and links, gone in one night |
| Paywall rules decided after design | Registration flows retrofitted badly into finished screens |
| Ad slots that shift the layout | Readers tapping the wrong thing, and a poor layout score |
| Editorial workflow nobody consulted on | Editors working around the system within a month |
| Articles built as pages rather than data | No series, no topic pages, no automation |
| Heavy embeds with no budget | Article pages that take eight seconds on a phone |
An independent publication runs $15,000 to $30,000. A subscription publication with paywall and billing runs $30,000 to $55,000. Larger publishers run $55,000 to $90,000.
It is the first thing scoped, not the last. Every URL is inventoried and mapped before the build starts. Publishers who leave this to launch week lose years of rankings and links, and that loss is not recoverable.
Yes, and the rules need deciding before design. Metered, hard, or membership, what registration requires, and what search engines are permitted to see. Those decisions shape screens, so retrofitting them produces a worse product.
By treating ad calls, embeds and images as a budget rather than as whatever the page ends up with. Fixed ad slots that do not shift the layout, lazy loading below the fold, and a hard limit on third party scripts.
Only if they were asked. Drafting, editing, scheduling, embargoes and corrections have to match how the newsroom actually works. If publishing takes eleven clicks, editors route around the system and it stops reflecting reality.
Because articles, authors, topics and series as data are what make archives navigable, newsletters automatable and syndication possible. As a folder of pages you get none of that and rebuild it manually every time.
Yes, with limits agreed in the build. Unmanaged ad tech is the most common reason publishing sites feel broken, and the fix is a stated boundary on what third parties may insert rather than a conversation after launch.
No. We work with New York publishers remotely on Eastern hours and would rather say that than imply otherwise. Deadlines and release windows are agreed at the start regardless.
Send us the site and we will tell you how many indexed URLs are at risk and what a mapping exercise would cover.