
Start with the business job of the website
Write down what the site must help the business do.
Not the page count. Not the animation style. Not the stack. The business job.
For most B2B SaaS teams, the website needs to support some mix of:
- qualified demo requests
- consultation calls
- enterprise sales conversations
- self-serve signups
- partner inquiries
- pricing evaluation
- product education
- hiring confidence
- investor or analyst trust
Each outcome needs a path. A demo request may need a product page, use-case page, proof section, pricing explanation, form, routing logic, and thank-you state. A partner inquiry may need different content, a different form, and a different CRM workflow.
If every path points to the same generic CTA, you do not have a conversion strategy. You have a button.
Define buyer paths before page types
A sitemap helps, but it should come after buyer-path work.
Start with the main audiences and the questions each one brings to the site. A CFO does not read the same way a product lead reads. A founder comparing agencies does not need the same evidence as a marketing director trying to fix an old CMS.
Map each buyer path in plain language:
| Buyer path | Key question | Page support | Conversion action |
|---|---|---|---|
| New category visitor | What does this product do? | homepage, product overview, use cases | explore product or book consult |
| Active evaluator | Can this solve my specific problem? | use-case pages, comparison pages, proof, FAQ | request demo |
| Technical reviewer | Will this fit our stack? | integrations, security notes, docs links | talk to team |
| Budget holder | Is this worth the spend? | pricing logic, implementation scope, risk reduction | book sales call |
| Existing lead | Can I trust this company? | about, process, resources, case proof when approved | choose next step |
This map gives an agency something useful. It tells the team which components, templates, forms, and CMS fields need to exist because the business needs them, not because a competitor had them.
Decide what marketing must control in the CMS
A CMS requirement is not “marketing can edit the website.” That sentence hides the real decisions.
Define what editors can create, change, reuse, and publish without developer help.
For a B2B SaaS site, the CMS may need to control:
- landing pages
- product pages
- feature cards
- use cases
- industries
- integrations
- pricing content
- comparison tables
- resource posts
- author profiles
- FAQ entries
- CTAs
- form destinations
- SEO fields
- redirect notes
The agency should know which fields need guardrails. Editors may need to change a headline, proof block, CTA, and FAQ. They may not need to change spacing, component order, or schema rules.
Good CMS modeling protects speed and consistency at the same time. Marketing gets room to publish. Design keeps the system intact. Developers stop receiving tickets for small copy edits.
Write conversion requirements like product requirements
Conversion points need more detail than “add a contact form.”
Define each conversion action and the rules around it:
| Conversion point | Required decisions |
|---|---|
| Demo form | fields, required fields, spam protection, CRM routing, calendar handoff |
| Pricing inquiry | qualification fields, source tracking, thank-you state, sales notification |
| Newsletter signup | consent language, list destination, double opt-in needs |
| Content download | gated or ungated, nurture workflow, attribution fields |
| Consultation CTA | destination, booking rules, fallback contact route |
Also define the content around each action. A high-intent CTA on a pricing page may need less explanation than a CTA on a technical integration page. Treat the form, copy, tracking, and follow-up as one system.
That work matters because a prettier form will not fix a vague offer, poor routing, or missing context.
Build the analytics plan before implementation starts
Analytics work after launch turns into cleanup work.
Before development starts, define what the team needs to measure:
- key page views
- CTA clicks
- form starts
- form submissions
- booking clicks
- resource downloads
- pricing interactions
- scroll or engagement events only where useful
- error states
- source and campaign parameters
Name events before developers build the components. Decide which events fire in the browser, which fire after a successful form submission, and which need CRM or marketing automation context.
For most teams, the goal is not more dashboards. The goal is cleaner decisions. You want to know which pages help qualified visitors move forward and which paths create friction.
List required integrations and failure states
Most website scopes mention integrations. Few scopes define what happens when they fail.
List every system the site touches:
- CRM
- marketing automation
- calendar booking
- email platform
- analytics
- tag manager
- chat or support tool
- authentication
- product docs
- payment or billing links
- search
- consent management
Then define the failure states. If the CRM API fails, does the form store the lead somewhere else? If the calendar tool is down, where does the CTA send visitors? If consent settings block analytics, what still needs to work?
This keeps the site from depending on a chain of silent assumptions.
Set performance and accessibility requirements early
Performance and accessibility cannot sit in the QA column at the end.
Set requirements before design and development choices harden:
- image strategy
- font loading
- animation rules
- script budget
- third-party tools
- Core Web Vitals targets
- keyboard navigation
- color contrast
- focus states
- semantic headings
- form labels and errors
- reduced motion needs
web.dev frames site health around loading, responsiveness, and visual stability. Google Search Central frames SEO around making pages crawlable, useful, and clear. Those ideas affect design systems, content structure, and engineering choices from the start.
A SaaS site can look premium and still feel slow, confusing, or fragile. Requirements prevent that tradeoff from sneaking into the project.
Define SEO migration needs before the build
If you already have a live site, the rebuild needs migration requirements.
Create a pre-build list:
- current URLs
- target URLs
- redirects
- pages to merge
- pages to remove
- metadata updates
- canonical rules
- XML sitemap needs
- robots.txt rules
- structured data needs
- internal links
- image alt text
- launch-day crawl checks
Do this before development starts. URL structure affects routing, CMS fields, page templates, and QA. If the agency discovers the migration plan during launch week, the team has already lost time.
Choose the stack after the requirements are clear
The stack should answer the requirements. It should not replace them.
A custom full-stack build may fit when the site needs a strong component system, structured content, fast performance, flexible integrations, and long-term control. A hosted website platform may fit when the team needs speed, simpler governance, and fewer technical decisions. A headless CMS may fit when marketing needs reusable content across many page types. A traditional CMS may fit when the team values familiar editing and plugin availability.
The right choice depends on operations, not taste.
Ask:
- Who edits the site each week?
- Which pages change most often?
- Which content needs reuse?
- Which integrations does the site need?
- How much engineering support exists after launch?
- What performance standard does the site need to meet?
- What will break if the marketing team moves fast?
Virdis starts here because maintainability decides whether the site keeps improving after launch. A strong build should make future changes safer, not make the agency necessary for every small update.
Ask agencies for a requirements response, not a sales deck
Once you write the requirements, use them in the agency selection process.
Ask each agency to respond with:
- assumptions they would challenge
- risks they see in the scope
- CMS model recommendations
- analytics implementation approach
- migration plan
- QA process
- launch and post-launch handoff
- what they would not build first
The last point matters. A strong agency will protect the project from low-value complexity. If every requested feature gets a yes, you may be buying agreement instead of judgment.
The short requirements checklist
Before you hire, document these items:
- Business outcomes the website must support
- Primary buyer paths and conversion actions
- Page types and reusable components
- CMS roles, fields, and publishing rules
- Form logic, CRM routing, and thank-you states
- Analytics events and naming rules
- Third-party integrations and failure states
- Performance budget and script limits
- Accessibility requirements
- SEO migration plan
- Security and privacy expectations
- QA process and launch checklist
- Post-launch ownership model
If this feels like too much work before hiring, that is the point. The requirements do not make the project heavier. They expose the weight that was already there.
When Virdis is a fit
Virdis is a fit when the website needs more than a visual refresh.
That means the site has to support clearer positioning, stronger buyer paths, cleaner content operations, reliable tracking, and a maintainable technical foundation. The work sits across strategy, design, development, CMS structure, and launch discipline.
If you need a custom website that your team can keep using after launch, requirements are where the project should start.
Frequently asked questions
What should B2B SaaS teams define before hiring a web development agency?
Define buyer paths, page types, CMS needs, conversion actions, analytics events, integrations, performance requirements, accessibility standards, SEO migration needs, and post-launch ownership before hiring.
Should a SaaS team choose a platform before hiring an agency?
Choose the platform after the requirements are clear. The stack should fit content operations, integrations, performance needs, team ownership, and long-term maintenance.
Why does CMS structure matter in web development scope?
CMS structure controls how the marketing team updates the site after launch. Good structure gives editors useful control while protecting layout, SEO, content relationships, and conversion paths.
What makes a web development scope risky?
A risky scope hides decisions about forms, tracking, redirects, CMS fields, integrations, accessibility, performance, and ownership. Those gaps turn into delays and change orders.
