Website content planning: what must be ready before design and development begin

Website content rarely arrives late because someone forgot to write a few paragraphs. It arrives late because the organisation has not agreed what exists, what the new website needs, who owns the source material, what can be removed or how approval will work.

Design then begins with tidy placeholder copy. The layouts appear settled until real service descriptions, legal qualifications, product data, biographies, resources, forms and edge cases arrive. Components change, migration effort grows and launch timing becomes dependent on reviewers who were never allocated to the project.

The answer is not to finish every sentence before opening a design file. It is to establish content truth: realistic types, structure, volume, priority, ownership and migration assumptions. Final copy can progress in parallel when those foundations and dependencies are managed.

The short answer: design needs content truth, not necessarily final copy

Before experience design hardens, the team should know the priority audience tasks, main content types, representative examples, likely volumes, structural relationships, owners, approval pathway and broad migration disposition. The team should also understand which content is regulated, duplicated, held in another system or difficult to export.

Copy production can continue alongside design and development where the format is stable and missing content does not change the solution. High-risk content, complex structured data and unusual edge cases should be brought forward. If content uncertainty materially changes architecture, component scope, data migration or governance, it belongs in paid Full Website Discovery rather than an optimistic allowance.

• Know what audiences need to understand or do
• Inventory content types and approximate volumes
• Define repeatable fields and relationships
• Name sources, authors, reviewers and approvers
• Decide what is retained, improved, combined, created, migrated or retired
• Plan copy production, entry, migration and quality assurance as separate work

Final copy can continue when these foundations and dependencies are explicit.

Start with audience tasks and business outcomes

Content planning should begin with purpose, not the current navigation. Identify the questions people need answered, decisions they need to make and actions they need to complete. Consider where they arrive from, what they already know, what may concern them and what evidence supports trust.

Connect those needs to organisational outcomes. A service page may need to qualify enquiries, explain eligibility, reduce avoidable calls or direct people to a location. A resource may support early research, existing customers or a long sales cycle. A location page may combine contact details, availability, services and local proof.

Include accessibility, plain-language, privacy and specialist review needs from the beginning. GOV.UK guidance on interface writing emphasises clear, direct language and using the language of users. W3C guidance explains how headings and meaningful page structure help people understand and navigate content. Those are design and content responsibilities, not a final editorial polish.

Build a usable content inventory

A content inventory records what exists. At minimum, capture the URL or location, title, format, content type, owner, source system, intended audience, purpose, status and relevant dependencies. Add evidence such as traffic, search visibility, enquiries, user feedback or support demand where it is available and reliable.

Look beyond ordinary web pages. Include PDFs and other documents, forms, calculators, videos, images, policies, events, people, locations, case studies, product or service data, FAQs, gated resources and content held in external tools. Identify items that are reused manually across several places or depend on data from another system.

Keep the distinction between inventory and audit. An inventory describes the estate. An audit judges content quality, value and action. A site-specific audit, strategy or page plan requires agreed criteria, evidence and professional effort. The educational checklist in this article should not be treated as a substitute for that engagement.

Choose a disposition for each content group

The new website should not automatically inherit the current estate. Apply consistent criteria to decide whether content should be retained, improved, combined, created, migrated or retired. Consider audience need, business value, accuracy, uniqueness, risk, performance evidence, ownership and the cost of keeping it current. Use the separate guide to update, consolidate, redirect or retire older content.

Treat the decision as a content group or pattern before attempting line-by-line review. A group of location pages, staff profiles or product documents may share a rule, while high-value or sensitive items need individual attention. Record uncertainty and the subject-matter evidence required.

Protect search and user journeys before content is removed. Understand existing URLs, links, backlinks, campaign destinations and external references. Decide whether a retired URL should redirect to a genuinely relevant destination, remain available for a period or return an appropriate status. Do not redirect unrelated content to the home page merely to avoid a 404 response.

Protect search, compliance and redirect dependencies before removing or merging content.

Define content models before templates

A content type describes a repeatable unit of information, not only a visual page. A service might have a name, summary, audience, benefits, process, locations, related specialists, FAQs and call to action. A team member might have credentials, locations, areas of expertise and related resources. These fields and relationships can support several views without copying the same information into disconnected pages.

Separate meaning from presentation. The CMS should know that a field is an address, credential, date, category or price where that structure matters. Design can then determine how the information appears in different contexts and devices. Unstructured rich-text fields can feel flexible but often create inconsistent content and weak reuse.

Use representative real examples during design. Include the shortest and longest credible title, an empty optional field, an expired item, a service offered only in one location, a person with several roles, a resource with a complex file and any regulated qualification. The edge cases reveal component and governance requirements that ideal placeholder copy hides.

Assign owners, sources and approvals

Every material content type needs a source of truth and an accountable owner. Name who provides the initial material, who writes or edits it, who verifies subject matter, who reviews privacy, legal or compliance matters where required, and who gives final approval. The same person may hold several roles in a smaller team, but the decision sequence should remain clear. Emote can separately scope content writing and editorial production when the project requires it.

Set review timeframes and escalation before deadlines tighten. Define whether silence means approval, rejection or delay. It is generally safer to require explicit approval for sensitive content. Use a controlled comments and version process so feedback is traceable and several conflicting copies do not circulate.

Plan ongoing ownership at the same time. Record review frequency, expiry triggers, update sources and what happens when an owner leaves. W3C accessibility guidance treats accessibility as an organisational lifecycle responsibility. Content accuracy, readability and alternative text also decay without governance.

Plan production and migration as separate workstreams

Writing and migration are not the same task. Estimate the effort to research, draft, rewrite, edit, approve, enter, transform, map, import, validate and redirect content separately. A page that needs little rewriting may still be difficult to migrate if its structure, files or relationships are poor. Protect existing visibility with an approved SEO migration checklist.

Choose manual, assisted or automated migration only after the content and data are understood. Automation can be efficient for consistent, structured material. It can reproduce inconsistency at scale when source data is unreliable. Manual entry can be appropriate for a controlled estate but still needs templates, training and quality assurance.

Include images, rights, captions, alternative text, transcripts, downloadable documents, metadata, links, redirects and forms in the migration plan. Validate not only that an item moved, but that it appears correctly, remains accurate, is accessible, links to the right destination and is owned after launch.

Govern AI-assisted drafting like any other publishing input

AI-assisted drafting may accelerate exploration or first-pass production, but it does not supply authoritative facts, usage rights or accountability. Every workflow still needs approved sources, confidentiality controls, image and quotation rights checks, a named human subject-matter reviewer, brand editing and final responsibility for published claims.

Record where generated material entered the process and keep it out of approval by default until a responsible owner has checked meaning, evidence, tone and legal or regulatory implications. Speed is useful only when the organisation can still explain and stand behind the result.

Set a minimum content-ready gate for design and build

A content-ready gate should be practical. Confirm the priority audiences and tasks, approved content types, representative real examples, approximate volumes, source systems, owners, dispositions, migration assumptions and approval plan. Identify high-risk items and dependencies that must be resolved before relevant components are approved.

Lower-risk copy can then move in parallel under a clear schedule. Design should use real content early and record field or volume assumptions. Development should not configure an import or CMS model against an unapproved spreadsheet structure. Migration should not begin before redirect and quality rules are defined.

Use Focused Website Discovery where one straightforward public corporate or lead-generation site has contained content and low complexity but still needs its scope clarified. Use Full Website Discovery where content volume, structured data, ecommerce, integrations, significant migration, multi-region delivery or complex governance materially affects the solution. The pathway should match the evidence and consequence.

Estimate content effort from patterns, not averages

A total page count is a weak production estimate because pages vary. Group the estate into archetypes such as services, locations, profiles, resources, products, policies and campaign pages. For each group, identify how many items are retained, improved, created, migrated or retired and which approval or data dependencies apply.

Estimate the activities within each pattern. A service page may require source gathering, interview, draft, subject review, compliance review, edit, entry, links, metadata and final quality assurance. A structured location record may require less writing but more data cleansing and validation. A document may need accessibility remediation or replacement rather than simple upload.

Use samples before scaling. Complete several representative items, including a difficult edge case, then update the assumptions for time, source quality and review. This produces a more credible forecast than assigning one uniform number of hours to every page. Keep any early estimate indicative until the relevant evidence is available.

Use real content to test components and workflows

Content testing should include more than pasting final text into a design. Ask whether the component supports the content hierarchy, whether editors understand each field, whether required relationships can be managed and whether the interface remains clear when optional information is absent. Preview the same content across mobile, desktop, zoom and relevant assistive-technology conditions.

Test publishing tasks with the people who will perform them. Can an editor update shared information once, create a related item, withdraw expired content, replace a document and understand validation messages? A content model that is elegant in a diagram can still create daily friction if fields, permissions and preview are unclear.

Include governance rules in acceptance criteria. If only approved components may be used for regulated content, make that constraint visible. If certain fields must be completed before publication, validate it. If a piece of content expires, decide whether the CMS warns, unpublishes or relies on an owner. Design, configuration and process need to agree.

Plan content in release waves

Where the estate is large, prioritise content by audience and launch dependency. Wave one contains material required for core journeys and legal or operational continuity. Later waves can add lower-priority resources or enhancements when that phasing does not create broken journeys, misleading gaps or search problems.

A minimum viable content release still needs to be complete for the tasks it promises. Publishing an empty service category or a partial location directory shifts internal delay onto users. Define the quality threshold for every wave and preserve the ability to link, navigate and obtain support.

Record the future backlog and owner rather than relying on memory after launch. Set dates for deferred items and review whether they remain valuable. Phasing is a governance choice, not permission to create a permanent second tier of unfinished content.

Define a freeze rule for each wave. Late factual corrections must remain possible, but structural additions should return to scope and dependency review. This prevents an approved component or migration batch from changing repeatedly while developers and content editors work from different assumptions. Record exceptions and their effect on testing, migration, training and launch readiness.

Parallel does not mean independent.

Common content-planning mistakes

• Starting design with ideal placeholder copy and no real edge cases
• Counting pages without identifying structured content, files or forms
• Treating the current sitemap as the future content strategy
• Assuming every existing page should migrate
• Combining writing, entry, migration and QA into one invisible task
• Giving review responsibility to a group without a final approver
• Removing URLs without search, campaign and redirect checks
• Leaving post-launch content ownership until handover

These mistakes do not mean content must become a long sequential phase. They mean parallel work needs evidence, boundaries and decision gates. Good content planning allows design to move faster because the team is working with the right problem.

Frequently asked questions

Should content come before website design?

Content evidence should come before detailed design. Final copy can continue in parallel once content types, hierarchy, representative examples, volumes, owners and approvals are sufficiently clear.

What is the difference between a content inventory and content audit?

An inventory records what exists and relevant facts. An audit applies agreed criteria and evidence to judge quality, value, risk and action. The second is a more interpretive professional exercise.

Can we migrate all current content and improve it later?

You can, but it often transfers duplication, inaccuracy and governance debt into the new system. At minimum, classify content, identify sensitive or obsolete material and protect URL decisions before migration.

Who should approve website content?

The approver depends on the content. A content owner may approve routine material, while product, legal, privacy or compliance specialists may be needed for particular claims. Name the final decision-maker and review timeframe.

Does the web agency write all content?

Only where copywriting or content production is expressly included. The client still needs to supply reliable source material, subject-matter access and approvals. Content strategy, writing, migration and entry should each have clear scope.

When does content complexity require Discovery?

When unknown content volume, structured data, sources, relationships, migration, languages, permissions or approvals could materially change architecture, scope, risk or cost. Contained content can use a proportionately smaller pathway.

How Emote can help

Emote can connect content planning with UX/UI design, website development, SEO migration and implementation quality assurance, with writing, entry and migration included only when expressly scoped.

A manageable content estate may need focused definition before build. Larger estates, structured data, significant migration, ecommerce or complex approvals may justify paid Full Website Discovery before design and implementation are fixed.

If content is becoming the critical dependency in your redevelopment, book an initial meeting with Emote.