The comparison is usually framed as which platform is better, which has no answer. Both will publish a one-person publication competently for a decade. The useful framing is narrower: the two platforms are strong on different dimensions of the same rubric, and the choice is really a decision about which dimension you would rather buy and which one you are willing to build.
The short version. Ghost wins on time to publish and on the reading experience you get without doing anything. WordPress wins on everything you might want to add in year three. Neither advantage is permanent, and both can be bought off with work.
Who should pick which
If you want the answer without the reasoning, this is it. Everything below is why.
| If this describes you | Pick | Because |
|---|---|---|
| The publication is the product, funded by paid subscriptions | Ghost | Subscriptions native, zero share, publishing in an afternoon |
| You cannot yet name how it will make money | WordPress | Optionality is the thing you are buying |
| Large free list monetised by sponsorship | WordPress | Member-count pricing charges you for the free list |
| You would rather write than configure | Ghost | Craft is handled by default rather than by you |
| The archive will be structured, not chronological | WordPress | Native taxonomies and archive templates |
| The publication is one surface of something larger | WordPress | Commercial surfaces other than subscriptions |
| You will not maintain performance work | Ghost | Its defaults are good; WordPress defaults are not |
| You enjoy running infrastructure | Self-hosted Ghost | Removes the pricing model entirely |
Two of these rows contradict each other for most publishers, which is the honest state of the decision. If you would rather write than configure and you cannot name the business model, you are being asked to choose between a lower ceiling and more work, and nothing in a comparison table resolves that for you.
What each one actually is
Ghost is a publishing application. It ships a CMS, a newsletter, member accounts, tiers and a paywall as one product, with a deliberately small surface. It is open source under the MIT licence, so it can be self-hosted, and Ghost(Pro) is the managed hosting run by the same organisation.
WordPress is a general content management system with a plugin architecture. Newsletter, membership and commerce are not in the box; they are things you assemble. The trade in both directions follows from that single difference.
Dimension by dimension
| Dimension | Ghost | WordPress |
|---|---|---|
| Owned hub strength | Strong. Self-hostable, full export, portable archive | Strong. Self-hosted by default, full database access |
| Discovery | Sitemap and feed by default. Hub layer needs routes.yaml |
Sitemap and feed by default. Taxonomies and archive templates native |
| Craft | Strong out of the box. Fast default themes, controlled typography | Varies entirely with theme and plugin load |
| Monetization design | Subscriptions native, 0% platform fee. Other surfaces are yours to build | Nothing native. Any surface is buildable |
| Distribution loop | Newsletter and capture native. Sending bundled on Ghost(Pro) | Plugin plus external sending platform |
| Compounding | Good. Portable archive, stable URLs | Stronger. Deeper taxonomy, full control of URL structure |
Where Ghost wins, and why it is not a small thing
Craft, on day one. Ghost’s default themes are fast, the type scale is deliberate, and the reading experience on mobile is not an afterthought. This is the single largest practical advantage, because Craft is the dimension most publishers score worst on and the one they are least equipped to fix. A default Ghost install lands around the middle of the band range without intervention. A default WordPress install lands wherever the theme puts it, and most themes are built to demonstrate features rather than to be read.
Subscriptions, without a stack. Tiers, a paywall, member accounts and Stripe are one product. Ghost takes zero percent of subscription revenue on every plan; you pay Stripe’s processing and nothing else. Assembling the equivalent on WordPress means a membership plugin, a payment integration, and the maintenance surface both create.
Delivery, handled. On Ghost(Pro), sending is bundled. That matters more than it sounds, because the alternative is configuring a sending domain, authentication records and a reputation you own. The requirements involved are set out on the email authentication reference page, and they are the most common place a small publication quietly breaks.
Where WordPress wins, and why it takes three years to notice
The hub layer. The rubric’s Discovery band 4 requires a maintained hub layer covering the subjects that recur most often, and band 5 adds structured data on those hubs. WordPress does this natively: categories, tags, custom taxonomies, custom post types, and an archive template for each. Ghost can do it, through tags and dynamic routing in routes.yaml, but it is a configuration file rather than an interface, and the ceiling on how many distinct content structures you can model is lower.
This is invisible with forty posts and decisive with four hundred. It is also the mechanism by which an archive compounds: a growing body of work needs somewhere for the structure to live, or it stays a reverse-chronological list forever.
Commercial surfaces other than subscriptions. Ghost is excellent at one monetization model. Monetization design band 5 asks for something packaged rather than sold per placement, with disclosure and a stated policy. If what you eventually package is a course, a data product, a directory, a sponsor inventory page with published rates, or anything that is not a recurring content subscription, WordPress is where that gets built without fighting the platform.
No member-count pricing. This is the cost that surprises people, so it is worth stating precisely. Ghost(Pro) prices by total member count, and free subscribers count the same as paying ones. Eight hundred free signups and two hundred paying members is a thousand members. Starter is capped at a thousand and, since the mid-2025 pricing change, does not support paid subscriptions at all, so a publication that wants to charge starts at Publisher regardless of size. Above the entry brackets, Publisher and Business scale continuously with audience.
The structural point is not the price. It is that a large free list is a cost centre on Ghost(Pro) and a rounding error on self-hosted WordPress. If your model is a big free audience monetized by sponsorship rather than a small paid one, the pricing model is working against your strategy.
What each actually costs to run
The bill is the least interesting cost on either side, and the shapes are different enough that comparing entry rates is misleading.
| Ghost(Pro) | WordPress, self-hosted | |
|---|---|---|
| Pricing basis | Total member count | Hosting, independent of audience |
| Free subscribers | Billed the same as paying ones | Cost nothing |
| Sending | Bundled | Separate provider, separate bill |
| Subscriptions | Native, 0% platform fee | Plugin plus payment integration |
| Theme and design | Included, good by default | Free to expensive, quality varies wildly |
| Maintenance | Handled | Yours: updates, backups, security, performance |
| Scales badly when | The free list grows | The plugin count grows |
The row that decides it for many publications is the second one. A member-count model is efficient for a small paid list and actively hostile to a large free one, and most publishers do not know which they are building until it is already built. The full comparison of what each pricing model does to a publication that succeeds is in email platforms ranked by what they cost the distribution loop.
The WordPress column has a cost that does not appear as a line item and is larger than the hosting bill: attention. Every plugin is a dependency on someone else’s release schedule and weight on the page. The default outcome of an unmanaged install is a slow site, which is why performance is measured rather than assumed in every teardown, against the published thresholds.
What each choice costs you
Ghost’s cost is ceiling. Everything it does, it does well, and the set is fixed. The escape hatch is self-hosting, which trades the subscription for a server, a mail provider, updates, backups and deliverability that are now your responsibility. That is a real job, and the people who do it well were going to enjoy doing it.
WordPress’s cost is that Craft is now your problem. Every plugin is weight on the page and a dependency on someone else’s release schedule. The default outcome of an unmanaged WordPress install is a slow site, which is why performance is measured rather than assumed in every teardown. The thresholds are on the Core Web Vitals reference page. A WordPress publication can score at the top of the Craft band; it will not get there by accident.
Self-hosted Ghost, which is the third option
The comparison is usually framed as two choices and there are three, because Ghost is open source under the MIT licence and can be run on your own server.
This removes the objection that does the most damage to the Ghost case: member-count pricing. A self-hosted install has no member brackets, so a large free list costs nothing, and the sponsorship-funded model that bracket pricing punishes becomes viable on a platform with excellent defaults.
What it costs is the thing Ghost(Pro) was selling. Sending is now a mail provider you configure and a sender reputation you own, which means authentication and alignment are your responsibility and the most common place a small publication quietly breaks. Updates, backups and uptime are yours. So is the recovery when something fails at an inconvenient hour.
The honest reading is that self-hosted Ghost has the best position on paper and a real operational cost that is easy to underestimate before you have carried it for a year. It suits publishers who would have enjoyed running the server anyway, and it is a poor recommendation for anyone choosing a platform in order to stop thinking about platforms.
A decision rule
Choose Ghost if the publication is the product, the model is paid subscriptions, and you would rather write than configure. You will be publishing in an afternoon and the result will look better than what most people build in a month.
Choose WordPress if the publication is one surface of something larger, if the archive is going to be structured rather than chronological, or if you cannot yet name how it will make money. Optionality is the thing you are buying, and it costs you the first month and some ongoing attention to performance.
The question that decides it is not about features. It is whether you know what the publication will be doing in three years. If you do, buy the platform that is good at that. If you do not, buy the one that can become anything.
Switching later, in both directions
Neither choice is permanent, and they are not equally reversible, which is worth knowing before treating this as a decision you can defer by picking either.
Ghost to WordPress is the easier direction. Ghost’s export is complete: posts, members and settings in a JSON file, with media available from the install. Nothing is being held back, so the work is import mapping and rebuilding templates rather than reconstruction.
WordPress to Ghost is harder in proportion to how much WordPress you used. A publication that used posts, categories and tags moves cleanly. One that used custom post types, custom fields, shortcodes and a page builder is moving content that has no equivalent on the other side, and the migration becomes a rewrite of the content model rather than a transfer.
In both directions the decisive factor is the same one that decides every migration, and it is not the platform: whether the URLs survive. An archive that keeps its addresses through a platform change keeps everything those addresses have accumulated. The mechanics, including how to build and verify a redirect map, are on the redirects and URL stability reference page, and a worked example of a real migration is in the Substack to Ghost migration.
Both platforms score well on portability, which is the reason this comparison is a genuine choice rather than a warning. Neither of them is the platform that traps you.
How this is scored
No teardown on this site scores a platform. Platforms set ceilings and defaults; publishers decide where inside them they land. A Ghost publication with no capture unit scores badly on Distribution loop despite the newsletter being built in. A WordPress publication with a maintained taxonomy, measured performance and published sponsor rates can score at the top of every band.
What the platform choice genuinely determines is which failures are the default and which take effort. Ghost’s defaults are good and its ceiling is lower. WordPress’s ceiling is higher and its defaults will not get you there. Both statements are in the benchmarks as the teardown set accumulates.
Questions
Is Ghost better for SEO than WordPress, or the reverse?
Neither, and the question misdescribes what is happening. Both emit sitemaps and feeds by default, both allow full control of URL structure, and both can produce a technically excellent site. WordPress has an advantage above the level of individual posts, because native taxonomies make a maintained hub layer straightforward, and the hub layer is what Discovery bands 4 and 5 are about. That is an architecture advantage rather than an SEO one, and it is invisible until the archive is large.
Does Ghost really take nothing from subscription revenue?
Ghost takes no platform fee on subscriptions on any plan; you pay payment processing and the plan itself. The cost arrives through member-count pricing instead, which is a different shape rather than a hidden version of the same thing: a share scales with revenue, a bracket scales with audience, and they diverge sharply for a publication with a large free list.
Can WordPress reach the top band on Craft?
Yes, and not by accident. A WordPress publication with a deliberate type scale, a controlled measure and good measured performance on every template can score at the top. The difference from Ghost is not the ceiling, it is where the default lands and how much sustained attention it takes to hold a position above it.
What about the other platforms?
This comparison covers the two most one-person publications actually weigh against each other. Substack, beehiiv, Kit and the rest sit in a different frame, because the question there is about what the platform keeps rather than about what it can build. That comparison is in email platforms ranked by what they cost the distribution loop.
I have already chosen. Is this worth reading?
The useful part is the failure list rather than the recommendation. Each platform makes a specific set of failures the default, and knowing which ones you have inherited is more actionable than knowing whether you chose correctly. A Ghost publication should check its hub layer and its commercial surface. A WordPress publication should check its performance and its type scale, on the article template, on a real post.
Sources
Plan structure and pricing behaviour verified August 9, 2026. Ghost has revised its plan structure and member brackets more than once, and Publisher and Business are priced on a sliding scale rather than a flat rate, so treat any figure here as a floor and read the current pricing page before budgeting.