WEBSITE TECHNOLOGY & CYBERSECURITY

Website Security Checklist for Indian MSMEs: Backups, MFA, Access Control and Incident Readiness

A practical security baseline for Indian small-business websites and digital operations. Secure domain access, hosting, CMS, forms, APIs, backups and third-party tools, then prepare a realistic incident-response plan.

A small-business website is connected to more than a hosting account. Its security depends on the domain registrar, DNS, hosting or cloud platform, CMS or application code, administrator accounts, forms, databases, payment and email integrations, backups and the people who can access them. Start by listing those assets and owners, enable multi-factor authentication on privileged accounts, patch supported software, keep tested backups outside the production environment, restrict access and decide who responds if something goes wrong. This guide adapts CERT-In's MSME cyber-defence baseline, OWASP's current application-risk guidance and NIST's small-business framework into a practical website-focused programme. It reduces avoidable exposure; it does not guarantee that a site cannot be compromised.

1. What website security actually covers

Website security is the protection of the website, its administration and the connected systems that make it work. A brochure site may have a domain registrar, DNS provider, hosting account, CMS, contact form and business email. An e-commerce store or customer portal may also include a product database, payment gateway, order-management system, CRM, cloud storage, API credentials and deployment pipeline.

A control on one layer does not secure every other layer. HTTPS encrypts traffic in transit, but it does not make a vulnerable plugin safe. A hosting provider may secure the underlying server while the website owner remains responsible for CMS accounts, application permissions, form handling and customer data. WordPress's own hardening guidance describes security as risk reduction rather than a promise of perfect protection and explains that host and application responsibilities differ.

Use this guide as a management checklist. For a live incident, a payment system, a customer portal or a site storing sensitive information, involve a qualified security professional. Do not describe a checklist review as a penetration test, certification or legal compliance audit.

  • Confidentiality: only authorised people and systems can view business or customer information.
  • Integrity: pages, prices, orders, records and configurations cannot be changed without permission.
  • Availability: the site and its essential services can be restored after a failure or attack.
  • Accountability: important access, administrative changes and incidents leave usable records.

2. Build an inventory of the website attack surface

Begin with a simple asset register. Record every account, service, application and integration that can change the website, route traffic, access customer information or restore the service. A common failure is to secure the CMS while leaving the domain registrar, shared administrator mailbox, deployment account or third-party form service outside the review.

For each asset, record its business purpose, owner, administrator, recovery method, data handled, renewal date and the process for removing access. Keep the register in a restricted location. Do not store passwords, recovery codes or API secrets in the inventory itself; use an approved password manager or secret store.

AssetWhat to recordResponsible owner
Domain registrarAccount owner, MFA, recovery email, renewal and transfer-lock statusBusiness owner or authorised domain administrator
DNS and CDNZone administrators, nameservers, DNS records, TLS and security settingsTechnical owner
Hosting or cloudProject/account owner, billing contact, support route, backups and access rolesTechnical owner or hosting provider
CMS or applicationVersion, plugins/dependencies, admin users, deployment source and update processDeveloper or site administrator
Forms, APIs and databaseData collected, endpoints, permissions, rate limits, secrets and retentionApplication owner
Email, CRM and paymentsIntegration owner, data access, webhook credentials and vendor support routeBusiness process owner
Backups and logsLocation, encryption, retention, restore owner and retrieval testOperations or security owner

3. Secure the domain registrar, DNS and recovery accounts

The domain account is a high-impact control point. Someone who can change nameservers or DNS records may redirect visitors, interfere with business email or disrupt the website even if the hosting account itself is secure. Use a named business-controlled account, enable MFA where available, keep recovery details current and restrict who can change nameservers, DNS records or domain ownership.

Avoid sharing one registrar password among an agency, developer and owner. Give each authorised person an individual account where the provider supports it. Keep a dated record of approved DNS changes and know how to contact the registrar if access is lost. Review domain renewal notices through the registrar's authenticated dashboard rather than trusting an unexpected email link.

For DNS, document the purpose of important A, AAAA, CNAME, MX, TXT and verification records before editing them. A website change should not accidentally remove mail-routing or email-authentication records. Where supported, use DNSSEC after confirming the registrar and DNS provider configuration; do not enable it without understanding the required DS-record coordination.

  • Turn on MFA for the registrar, DNS provider and the mailbox used for account recovery.
  • Use unique named users and remove former agencies, contractors and employees promptly.
  • Restrict nameserver, transfer and high-impact DNS changes to a small authorised group.
  • Keep a current DNS export or change record and a known support contact.
  • Test recovery access without exposing recovery codes in shared documents.

4. Protect administrator identities with MFA and least privilege

Use unique accounts for the hosting platform, CMS, cloud console, source-control organisation, analytics, email, payment dashboard and customer-management tools. Shared administrator logins make it difficult to identify who performed an action and make safe offboarding harder.

Enable multi-factor authentication on every privileged account that supports it. Prefer passkeys or phishing-resistant security keys where available; otherwise use a reputable authenticator method. Store recovery codes in a controlled password manager or other protected recovery process. MFA on the CMS alone is not enough if the mailbox that resets the CMS password is unprotected.

Apply least privilege: a content editor should not automatically receive billing, DNS, production deployment or database-owner access. Use separate roles for content, development, support and finance. Review access after role changes and immediately when a person or vendor leaves. Revoke sessions, API tokens, SSH keys and integration credentials—not just the visible user account.

  • Maintain a named owner account and at least one controlled recovery route.
  • Use MFA for domain, hosting, source control, business email and payment administration.
  • Avoid routine use of owner or super-admin accounts for everyday content changes.
  • Review privileged users and connected applications on a scheduled basis.
  • Include account, device, token and data-return checks in contractor offboarding.

5. Patch the CMS, themes, plugins, frameworks and dependencies

Outdated software creates avoidable exposure. Maintain an inventory of the CMS core, themes, plugins, server runtime, framework packages, database engine, operating system and any managed integrations. Subscribe to vendor security advisories and assign someone to decide whether an update is routine or urgent.

For WordPress, remove unused plugins and themes, use supported releases from trusted sources and test important updates before applying them to a production store or custom integration. The official WordPress hardening handbook recommends keeping software current and explains that plugins, themes, server configuration, database access, backups and logging all form part of the security picture.

For a custom Next.js or other application, review dependency advisories, lockfiles, build actions and deployment permissions. A clean build does not prove that dependencies are free of vulnerabilities. Do not run an automatic major-version upgrade on production without checking compatibility and rollback options.

  • Assign an owner and review cadence for every software component.
  • Prioritise security fixes according to vendor guidance, exposure and business impact.
  • Test updates on a staging environment where practical, then deploy with a rollback plan.
  • Remove abandoned plugins, unused packages, test accounts and old integrations.
  • Keep a record of the update, test result, approver and any accepted exception.

6. Keep secrets out of source code and the browser

API keys, database passwords, signing secrets, private tokens and webhook credentials must not be placed in public repositories, screenshots, support tickets or client-side JavaScript. Anything delivered to a browser can be inspected by the visitor. A variable prefixed for public exposure should contain only information intended to be public.

Use the hosting platform's environment-variable or secret-management facility for server-side credentials. Separate development, preview and production values. Grant each integration only the permissions it needs. If a secret is exposed, treat it as compromised: revoke or rotate it at the provider, update the deployment configuration and review logs for use of the old credential.

A developer leaving the project should trigger a review of personal access tokens, deploy keys, repository membership, cloud roles, database accounts and shared support access. Deleting the developer's laptop account does not revoke credentials already issued to the project.

  • Keep .env files out of Git and confirm secret-scanning or repository review where available.
  • Do not put service-role keys, database credentials or private API tokens in browser bundles.
  • Use different credentials for production and non-production environments.
  • Rotate exposed or no-longer-needed secrets and revoke unused tokens.
  • Limit who can read or change production environment variables.

7. Configure HTTPS and TLS without creating a false sense of security

Serve the website over HTTPS and redirect ordinary HTTP requests to the secure version. Confirm that the certificate covers the intended hostnames and renews correctly. Check the origin-to-platform connection as well as the visitor-to-edge connection when a CDN or reverse proxy is used.

Cloudflare's Always Use HTTPS setting redirects visitor requests from HTTP to HTTPS, but Cloudflare notes that this does not fix mixed-content resources. The origin encryption mode must also be selected deliberately. A redirect alone does not prove that the origin connection is encrypted or that the certificate chain is valid.

Consider HSTS only after all required hostnames and subdomains are confirmed to work over HTTPS. A premature HSTS or preload configuration can make an incorrectly configured hostname inaccessible. Do not copy a TLS setting from another website without understanding the hosting architecture.

  • Check the certificate and renewal status for the apex domain and every active subdomain.
  • Redirect HTTP to HTTPS once all required endpoints support secure connections.
  • Use an appropriate origin TLS mode and validate the origin certificate where supported.
  • Check images, scripts, forms and embedded resources for mixed-content warnings.
  • Document DNS/CDN changes so a future migration does not silently bypass security.

8. Harden forms, APIs, uploads and database access

A contact form is an application endpoint, not just a visual box. Validate inputs on the server, enforce sensible size limits, reject unexpected data types and avoid trusting values supplied by the browser. Apply rate limits or abuse controls appropriate to the endpoint. Do not expose internal error details, database identifiers or private customer information in the response.

If the site accepts file uploads, define allowed file types and sizes, store uploads outside executable application paths where possible, generate safe filenames and restrict who can retrieve them. A file extension alone does not prove that a file is safe. Business-critical uploads may need malware scanning or a review step before staff open them.

Restrict database accounts to the tables and operations required by the application. Keep production data out of public test environments. Avoid placing personal data in URL query strings, analytics events or debug logs. Review exports and webhook payloads because they can move customer information into systems that the website owner does not routinely inspect.

Entry pointBaseline controlEvidence to retain
Contact or enquiry formServer-side validation, abuse limits, minimal fields and safe error messagesTest cases, endpoint owner and alert route
Login or admin APIMFA where supported, session controls, rate limits and role checksAccess review and authentication logs
File uploadType/size limits, safe storage path, access checks and scanning where appropriateUpload policy and sample test results
DatabaseLeast-privilege service account, protected credentials and restricted network accessAccount inventory and backup/restore record
Payment or CRM webhookVerify provider signatures where available, reject unexpected events and prevent duplicate processingWebhook owner, secret rotation record and failure alerts

9. Make backups that can actually restore the business

A backup is useful only if it is recent enough, complete enough and recoverable. Identify the items required to rebuild the website: application files or source repository, database, uploads, product and order records, configuration, DNS documentation, encryption keys or recovery instructions, and any business data held outside the website.

Keep backup copies separate from the production account or environment so that a compromised administrator cannot simply delete every copy. Encrypt sensitive backups, restrict who can restore or download them, and define how long each type is retained. CERT-In's MSME baseline recommends regular encrypted backups, secure copies at other network locations and offline copies, plus periodic restoration tests.

Choose a backup frequency based on how much data the business can afford to lose and how quickly it needs to recover. A brochure site may need a different schedule from a high-volume store or booking portal. Record the recovery point objective (how much recent data may be lost) and recovery time objective (how long restoration may take), then test whether the current process can meet them.

  • Back up the database and user-uploaded files, not just the website code.
  • Keep an off-site or isolated copy and protect it with separate access controls.
  • Encrypt backups containing personal, financial or confidential business information.
  • Test restoration to a separate environment before relying on the backup.
  • Record the restore date, duration, missing items and corrective actions.

10. Keep logs and alerts that help you investigate

Decide which events are important enough to record: administrator logins, failed authentication, permission changes, new users, DNS changes, production deployments, secret rotation, form/API errors, payment-webhook failures and database administration. Collect enough context to investigate without copying unnecessary personal data into every log.

Restrict log access and protect logs from unauthorised alteration or deletion. Set alerts for high-impact events such as a new owner account, a disabled MFA setting, an unexpected deployment, a sudden change to DNS or repeated failed access. A log that nobody reviews may help later, but it is not an alerting process.

CERT-In's 2022 Directions require specified covered entities to enable ICT-system logs and maintain them securely for a rolling 180 days within Indian jurisdiction. CERT-In's 2025 MSME baseline also recommends minimum 180-day system and application log retention for MSMEs. Determine the legal applicability and technical scope for your own entity and systems; do not assume every third-party platform exposes every log or that a website access log alone satisfies all record requirements.

  • List the systems that generate security-relevant logs and who can retrieve them.
  • Synchronise system time and keep timestamps consistent across services.
  • Protect log access and document retention, export and deletion controls.
  • Alert a named owner when a critical account, setting or production route changes.
  • Test that logs can be retrieved when an incident occurs.

11. Review hosting, plugins and other third-party vendors

Your website may rely on a registrar, DNS provider, hosting platform, CMS vendor, payment gateway, form provider, CRM, analytics service, email provider, backup service and one or more development agencies. Each connection introduces an account, credential, support process or data flow that should have an owner.

Before relying on a vendor, understand what it hosts, what data it can access, how incidents are communicated, how backups are handled, how accounts are recovered and what happens when the contract ends. For critical services, keep an export or migration plan where feasible. A service-level promise is not the same as a tested recovery process.

Do not assume that a cloud platform's security controls automatically cover application code, CMS permissions, customer-data exports or your own staff accounts. Document the shared-responsibility boundary in plain language and review it whenever the stack changes.

  • Keep a vendor register with service owner, support contact and renewal date.
  • Record which vendors process customer data and what permissions they receive.
  • Review plugin and integration permissions before installing or connecting them.
  • Remove unused integrations and rotate credentials when a vendor relationship ends.
  • Know how to export essential data and restore service if a critical vendor is unavailable.

12. Prepare a simple incident-response runbook before an incident

A suspected compromise should trigger a controlled response, not an improvised sequence of password changes and server deletions. Name the person who coordinates the response, the technical contact who can isolate systems, the business owner who decides customer communication and the legal or privacy adviser who reviews reporting duties.

The first priorities are to protect people and customers, limit further access, preserve useful evidence and restore trustworthy service. Use a clean device for credential changes. Revoke suspicious sessions and exposed tokens, restrict affected accounts or endpoints, capture relevant logs and record the time of each action. Do not immediately wipe the only affected server or overwrite logs if doing so would destroy evidence needed for investigation.

Restore from a known-good backup only after the likely entry point has been addressed. Verify administrator accounts, application versions, database integrity, DNS records and integrations before reopening the service. Monitor closely after recovery and document what changed.

StagePractical actionAvoid
RecogniseRecord who noticed the issue, when, what systems and what evidence is visibleAssuming a strange email or alert is harmless without checking
ContainRestrict compromised accounts, sessions, endpoints or affected servicesDestroying logs or making undocumented changes
PreserveSecure relevant logs, timestamps, backups and provider communicationsSharing credentials or evidence through public channels
AssessIdentify affected data, systems, users, vendors and business servicesDeclaring the incident resolved before scope is understood
RecoverPatch the entry point, rotate credentials and restore from a verified clean backupRestoring the same vulnerable configuration
ReviewDocument decisions, notifications, root cause and control improvementsTreating recovery as complete without a follow-up review

13. What Indian businesses should know about CERT-In

CERT-In's 15 Elemental Cyber Defence Controls for MSMEs, version 1.0 dated 1 September 2025, is an official baseline document for micro, small and medium enterprises. It presents 15 control areas and 45 implementation recommendations, including asset management, network and email security, patch management, incident management, logging, third-party risk, backup and recovery, governance, access control and vulnerability assessment. CERT-In describes the document as a minimum starting point and says MSMEs should go beyond it according to their own risk.

Keep that baseline separate from the CERT-In Directions issued on 28 April 2022 under section 70B(6) of the Information Technology Act. The Directions identify classes of entities and specific duties, including reporting listed cyber incidents within six hours of noticing them or being brought to notice, designating a point of contact and maintaining specified ICT logs for a rolling 180 days. The official FAQ explains that the Directions apply to service providers, intermediaries, data centres, body corporates and other listed categories as applicable; it also explains the meaning of body corporate and the reporting scope. A business should not assume that a website checklist alone determines its legal status or reporting obligations.

The FAQ says information available at the time may be reported within the stipulated period, with additional information supplied later within a reasonable time. It is an explanatory document and expressly says it does not replace or amend the Act, Rules or Directions. Where an incident may trigger reporting, obtain prompt advice from qualified counsel or the responsible incident-response professional and consult CERT-In's current official instructions. Do not wait for a root-cause report if a statutory reporting clock may already be running.

This article is general operational guidance, not a legal opinion. The business should map its entity type, services, systems, incident category, applicable sector rules and contractual duties with an appropriately qualified professional.

  • Use the CERT-In MSME baseline as a structured self-assessment, not as a certificate of compliance.
  • Read the original Directions and current official FAQs rather than relying on a vendor summary.
  • Identify who can make incident decisions and who can contact CERT-In if required.
  • Confirm log coverage, retention, time synchronisation and retrieval capability for systems in scope.
  • Keep a record of legal applicability decisions and review them when the business or service changes.

14. Tailor the controls to the type of website

A low-complexity brochure website and an online store should not be assigned identical security priorities. Use the table to decide where to spend the first hour of work. Add controls according to the data, transactions and operational impact involved.

Website typePriority controlsAdditional questions
Brochure or information siteDomain MFA, HTTPS, software updates, restricted CMS roles, backups and uptime alertsWho can edit the site and recover the domain?
Lead-generation siteAll brochure controls plus server-side form validation, abuse controls, data minimisation and restricted lead exportsWhere do enquiries go, who can export them and how long are they retained?
E-commerce storeAll lead-site controls plus payment-provider security, order permissions, webhook verification, refund controls and tested database recoveryCan an attacker alter orders, payment instructions or customer records?
Customer portal or custom applicationAll relevant controls plus API authorisation, session security, dependency review, secrets management, logging and vulnerability assessmentWhich user can access which record or action, and how is that tested?

15. A practical 30-day website-security rollout

A small team does not need to buy a large security platform before improving its basic controls. Start with ownership, access and recovery. Then work through the exposed application paths and document the decisions that require specialist review.

PeriodWork to completeEvidence of completion
Days 1–3Inventory domain, DNS, hosting, CMS, repository, forms, database, vendors and privileged accountsNamed asset register and owner list
Days 4–7Enable MFA, remove unused accounts, review roles and secure recovery emailAccess review and offboarding checklist
Days 8–14Patch supported software, review dependencies, HTTPS/TLS, form validation and exposed secretsUpdate log, test notes and credential review
Days 15–21Configure isolated backups, restore test, logging and critical alertsSuccessful restore record and alert test
Days 22–30Write incident runbook, review vendor access, assess CERT-In applicability and close high-risk gapsApproved runbook, vendor register and risk-action list

16. Monthly website-security review checklist

  • Confirm the domain, DNS, hosting and business-email accounts still have named owners and MFA.
  • Review new administrators, API tokens, connected applications and former contractor access.
  • Check CMS, themes, plugins, framework packages and server components for supported security updates.
  • Review certificate status, HTTPS redirects, DNS changes and important security settings.
  • Test key forms, login flows, payment callbacks and file-upload restrictions.
  • Confirm the latest backup completed and perform a restore test on a safe environment according to the recovery plan.
  • Check critical authentication, administrator, deployment and application logs and confirm alerts reach a monitored contact.
  • Review vendor access, contract changes, data exports and integration permissions.
  • Review open vulnerabilities, accepted risks, incident actions and the next remediation owner.
  • Revisit legal and contractual incident-reporting duties when the entity, service or data handled changes.

17. Frequently asked questions

What is the first security step for a small business website? Start with an inventory of the domain, DNS, hosting, CMS, administrator accounts, forms, database, integrations and backups. Secure the accounts that can change or recover those systems before buying more tools.

Does HTTPS make a website secure? No. HTTPS protects traffic in transit when configured correctly. It does not fix weak passwords, vulnerable plugins, insecure application code, excessive permissions, exposed secrets or missing backups.

Is the CERT-In MSME baseline mandatory certification? No. The 15 Elemental Cyber Defence Controls document is a baseline of recommended controls and self-assessment guidance. It is not a certificate or a substitute for a risk assessment. Separate statutory directions may apply to an entity depending on its status and activities.

Does every Indian website have to report every incident to CERT-In within six hours? Do not reduce the Directions to that blanket statement. The Directions and official FAQ identify entity categories and incident-reporting requirements. Applicability and the incident type must be assessed on the facts. Obtain prompt professional advice if an incident may be reportable.

How often should a website be backed up? Set frequency according to how much recent data the business can afford to lose and how quickly it must recover. A transactional store generally needs a more frequent and tested recovery process than a static brochure site.

Should a small business use WordPress or a custom website for security? Neither platform is automatically secure. WordPress needs disciplined core, theme and plugin maintenance; custom applications need secure development, dependency, identity, secret and deployment controls. Choose based on capability and the business's ability to maintain it.

Can a free CDN or HTTPS setting replace a security review? No. Free platform features can help with specific controls, but they do not replace secure application code, account governance, backups, incident readiness or a professional assessment where risk warrants it.

When should a business hire a security professional? Use specialist help for customer portals, payment systems, sensitive personal data, repeated incidents, unclear regulatory scope, suspected compromise or any system whose failure could materially disrupt the business.

Key takeaway

Treat website security as an owned operating process: know every account and integration, protect privileged access, patch the application, test isolated backups, monitor important events and prepare an incident runbook. Use CERT-In's MSME baseline as a starting point, then confirm the legal and technical requirements that apply to your own business.

← Back to all guidesExplore related support ↗
Diwali Business Growth • Wishing you prosperity and new beginnings
Chat on WhatsApp