digi.ai.in
← digi.ai.in / blog
2026-08-298 min readSecurity

Why We Treat the CMS as a Liability, Not a Shortcut

A CMS doesn't remove risk from a website project, it just moves the risk into plugins, patches and page weight.

·

Most businesses pick a CMS because it looks like the fast, cheap option. It isn't automatically either. A CMS is a large, shared piece of software maintained by thousands of third-party plugin authors, and every plugin you install is code you didn't write, don't control, and are responsible for anyway. We don't refuse to build on CMS platforms. We just size the ongoing liability before we recommend one, instead of treating "we'll use WordPress" as a decision that closes the risk conversation.

·

This post lays out why, with the numbers behind it.

+

The ecosystem is bigger than the platform

·

CMS platforms dominate the web because they promise speed-to-launch. One market analysis citing W3Techs puts CMS usage at roughly 70.2% of all websites, with WordPress alone holding 62.2% share among CMS-powered sites as of November 2024 source. HTTP Archive's Chrome UX Report-based tracking tells a different story: WordPress at 33.21% of "measurable web origins" as of April 2026, down from a 35.76% peak in mid-2022 source.

·

These numbers aren't a typo on someone's part. W3Techs counts subdomains as single entries; HTTP Archive counts them separately and only includes sites with enough traffic to appear in real user data, which quietly drops every low-traffic site from the sample source. We mention this not to nitpick statistics, but because it's the first proof point for our argument: even the most commonly cited CMS numbers depend on how you draw the boundary of "the web." If the market-share figures are contested, you should be skeptical of any vendor who tells you a CMS decision is simple.

·

What isn't contested is the scale of the plugin ecosystem sitting on top of the core software. WordPress's official directory alone lists more than 63,000 free plugins, plus roughly 1,500 premium ones — around 64,500 add-ons in total source. Each one is a dependency you inherit. Each one is also a potential entry point for an attacker, and a potential drag on page speed. That's the liability, in one sentence: a CMS isn't really one piece of software, it's a marketplace of loosely reviewed extensions bolted onto a core.

+

Security: the core isn't the problem, the plugins are

·

The WPScan vulnerability database, built specifically to track WordPress security issues, now holds close to 55,000 known vulnerabilities, and the count keeps growing source. WPScan's 2024 report breaks that down further:

·
Severity / condition Share of reported vulnerabilities
High or critical severity Just over 20%
Medium severity 67%
Exploitable with subscriber-level access or none at all About 22%
·

That last row is the one that should worry a business owner most. Nearly a quarter of tracked vulnerabilities don't require a privileged account to exploit — meaning an attacker doesn't need to steal an admin password first. The same report identifies weak passwords and nulled (pirated) plugins carrying hidden backdoors as common infection causes source. Both of those are decisions made after launch, by whoever is left maintaining the site — which is usually nobody, once the agency that built it has moved on.

·

Patchstack's 2025 whitepaper adds the patch-management angle that matters most for ongoing risk: in 2024, 33% of vulnerabilities were not fixed in time for public disclosure, and many were found in plugins that have been abandoned by their developers and will likely never be patched source. Sucuri, a security firm, observed over 500,000 infected websites in 2024, and blamed WordPress core, plugin, and theme vulnerabilities for nearly half of the malware infections it tracked source.

·

Even the vendors tracking this disagree on whether things are improving. Patchstack's own data shows vulnerabilities reported through WPScan trending down year on year, while disclosures coordinated through Wordfence are rising — and Patchstack itself now coordinates 52% of all new disclosures source. Patchstack also flags a scoring problem worth knowing about if you ever get a security report full of red flags: WordPress vulnerabilities rarely receive a low severity score even when they need high privileges and are unlikely to be exploited in practice, which creates what they call a "when everything is important, nothing is" effect source. In plain terms — a vulnerability scan on a CMS site will almost always look alarming. Knowing which alarms are real requires someone who understands the codebase, not just someone who can read a report.

+

Performance: the CMS doesn't decide this, the implementation does

·

Security is only half the liability. The other half is speed, and here the CMS-selection advice gets even murkier.

·

HTTP Archive's 2025 Web Almanac puts the average page weight at roughly 2.67 MB on desktop and 2.28 MB on mobile — both well above the commonly recommended 1–1.5 MB range, with most of the growth coming from images and JavaScript rather than markup source. Across five popular CMS platforms measured on mobile:

·
  • Squarespace: median 3,974 KB (heaviest)
  • Joomla: median 2,900 KB
  • WordPress: median 2,894 KB
·

The Almanac is explicit that this isn't cosmetic. Pages that take longer than three seconds to load see much higher bounce rates, and additional delay is strongly linked to lower conversion source.

·

Here's the finding that undercuts the "just pick a lighter CMS" advice some agencies give clients: the Almanac's own analysis found that vertically integrated, hosted platforms deliver more consistent Core Web Vitals, especially on mobile, while self-hosted CMS platforms like WordPress show much wider variance driven by themes, plugins, and third-party integrations source. Translation — two WordPress sites can perform wildly differently depending entirely on who built them and what they installed on top. The CMS name on the box tells you almost nothing about the outcome you'll get.

·

And the outcome is measurable in revenue, not just in a Lighthouse score. Google's own web.dev case studies report Rakuten 24 saw a 33.13% increase in conversion rate and a 53.37% increase in revenue per visitor after optimizing Core Web Vitals, and Vodafone Italy improved Largest Contentful Paint by 31% and saw 8% more sales source. A Google/Deloitte collaborative study reportedly found that a 0.1-second improvement in load time produced a 10.1% increase in conversions for travel sites and 8.4% for e-commerce source. If a tenth of a second moves conversions by nearly a tenth, then the megabytes of unused plugin JavaScript sitting on a typical CMS install are not a minor technical detail — they're a line item against revenue.

+

Where the rules get formal: government and public-sector sites

·

For any Indian business building on behalf of, or adjacent to, government bodies, CMS choice stops being a private decision entirely. The Guidelines for Indian Government Websites and Apps (GIGW), prepared by the National Informatics Centre and adopted by the Department of Administrative Reforms and Public Grievances, were first released in 2009, updated to GIGW 2.0 in 2018 to cover mobile apps, and revised again as GIGW 3.0 in 2023 in coordination with MeitY's STQC Directorate and CERT-In source.

·

GIGW 3.0 explicitly directs government organisations toward intuitive page loading and use of a "state-of-the-art content management system," backed by centralised monitoring dashboards that flag non-conformity source. In practice this means government and PSU sites go through a formal Certified Quality Website (CQW) audit run through STQC source. That's a useful benchmark even for private-sector clients who will never face a GIGW audit: it's an official acknowledgment, from India's own IT ministry, that CMS choice and configuration are compliance-grade decisions, not a checkbox you tick once at launch and forget.

+

A checklist before you commit to a CMS

·

Given the above, here's what we actually walk through with a client before recommending a platform:

·
  1. Who patches it, and on what schedule. A CMS with no assigned maintenance owner will accumulate the exact abandoned-plugin risk Patchstack describes source.
  2. How many plugins does the site actually need, not how many the template ships with. Every plugin is inherited code and inherited risk.
  3. What's the real page weight, not the theoretical one. A CMS demo site loads fast; a CMS site with six plugins, a page builder, and a marketing pixel usually doesn't, per the Almanac's own variance findings source.
  4. Is anyone reading the vulnerability scans, and do they understand which alerts are noise. Patchstack's severity-scoring critique is a warning that raw scan output isn't self-explanatory source.
  5. Does this site face any regulatory audit, GIGW or otherwise, that turns CMS configuration into a compliance requirement rather than a preference source.
·

None of this means "don't use a CMS." It means the CMS is the start of a maintenance relationship, not the end of a project. Treating it as a shortcut is how sites end up slow, unpatched, and quietly bleeding conversions a year after launch.

+

Questions people also ask

+

Is WordPress inherently less secure than a custom-built site?

·

Not inherently — the core software isn't the main source of risk. WPScan's data shows the vast majority of tracked vulnerabilities come from the plugin ecosystem, and about 22% require no authentication or only subscriber-level access to exploit source. A custom build has its own risks; the difference is that a CMS site's risk is spread across thousands of third-party plugin authors you have no relationship with.

+

Does choosing a "lighter" CMS guarantee a faster website?

·

No. HTTP Archive's Web Almanac found that self-hosted CMS platforms like WordPress show wide performance variance depending on themes, plugins, and integrations, while some platforms named as comparatively light in one measurement (like WordPress at a median 2,894 KB on mobile) can still be built into a heavy, slow site depending on implementation choices source.

+

Do Indian government-adjacent projects have specific CMS requirements?

·

Yes, if the site falls under GIGW. GIGW 3.0, developed with MeitY's STQC Directorate and CERT-In, requires use of a "state-of-the-art content management system" and subjects sites to a formal Certified Quality Website audit process source source. Private-sector sites aren't bound by GIGW, but it's a useful signal of how seriously CMS decisions should be taken even outside government work.

Written by the studio behind digi.ai.in. Questions? hi@digi.ai.in or WhatsApp +91 98605 96905
WhatsApp us