Rebuilds fail less from weak design than from scope that keeps expanding after kickoff
Enterprise website rebuilds often look like one project and behave like ten: brand refresh, CMS swap, locales, forms and CRM, SEO salvage, auth-area tweaks, compliance rewrites—all jammed into one launch date. KSX requires written scope in Discover: success metrics, page types, migration duties, integration list, and explicit non-goals. Design wireframes only inside those bounds; Build ships a URL tree and content model; Launch accepts on redirects, indexation, and conversion paths—not on “prettier.” Below is an enterprise website rebuild scope checklist you can drop into an RFP or internal charter, ordered by decisions rather than fashion stacks. Built for group brand sites, manufacturing and B2B service sites, and teams leaving WordPress or legacy stacks for a modern front end.
Write success metrics first: three observable numbers the rebuild must move
Rebuilds without numbers fill the calendar with taste arguments. Put 2–4 metrics at the top of the scope doc: qualified leads, conversion on key service pages, recovery of organic entry URLs, form completion, sales citations of the site—pick ones someone will actually read, not “brand love.” Note baseline source (GA, Search Console, CRM) and windows (30/90 days post-launch). If the only goal is compliance domain or host change, still write “do-not-regress” metrics such as indexed URL count and 404 ceiling. Metrics decide cuts later: if leads are core, hero theater and long homepage motion go to later phases.
Content and URL inventory: the rebuild’s non-optional legal duty list
Export the full legacy URL set: anything with traffic, backlinks, indexation, or citations in sales decks must have a fate (keep, merge with 301, or written retire). List PDF cases, newsroom, careers, privacy, and ICP/compliance pages separately. Mark assets that must become HTML versus noise that can archive. The inventory is a scope appendix, not an SEO side hobby—sales and legal co-sign the “safe to drop” list. Quoting without inventory is signing a blind box.
Page types and template caps: stop scope exploding via template count
Lock scope with a template list: home, service hub, sub-service, case, insight/news, about, contact, legal, careers, landing—fields and CMS needs per type. Ban “unique interaction system per division” unless separately funded. List dynamic widgets (filters, maps, calculators) as build-now or phase-two one by one. Template caps force merges of similar pages—the main cost lever on rebuilds. KSX accepts template count in design wireframes before full visual production discovers seven missing layouts.
Integration bounds: CRM, forms, SSO, analytics, and marketing pixels
Integrations are the number-one rebuild slip. Scope must state form destinations (email/CRM/ticket), field dictionary, spam strategy, who owns API credentials, and whether sandboxes exist. If auth only deep-links a legacy system, write “preserve external link,” not “unified experience.” Freeze analytics events (lead, download, locale switch) and consent mode. Pixels owned by marketing need a launch window and named accepter. Anything “later” must sit in non-goals or it becomes mandatory on UAT night.
SEO and locales: migration duties in the contract, not in “optimization tips”
Scope should include: 301 map deliverable, canonical/hreflang strategy, title and primary-content migration rules, structured-data choices, and who owns post-launch Search Console verification. For CN/EN pairs, state nav isomorphism, default locale, and x-default. “All-new paths, old links die naturally” is not an acceptable default. If Core Web Vitals are acceptance criteria, use field thresholds—not lab screenshot slogans. SEO is not a paint color on the rebuild; it is a hard chapter of the enterprise website rebuild scope checklist.
Content production and governance: who writes, who approves, who hits publish
Rebuilds chronically underestimate copy. Scope must name first-wave URLs that need real body text, translation ownership, legal review cycles, and whether cases may be anonymized. CMS roles and workflow (draft→review→publish) freeze before build. If client content will lag, contract “placeholder launch / phased publish” so engineering does not idle for words. Manufacturing clients resemble Weaverine-style work: product and capability narrative is client-sourced; the agency owns structure and presentation, not invented specs.
Non-goals and phase-two pool: move political asks off the first ship
Written non-goals beat wish lists. Typical first-ship exclusions: full multi-surface design system, personalization, member center, programmatic city pages, whole-site motion system, legacy comment migration. Keep a numbered phase-two pool with rough sizing so “no” is not heard as “never.” Executive late pages go through change control: if launch date, budget, or SEO map moves, re-estimate. KSX’s Discover→Design gate is mutual signature on non-goals before high-fidelity work.
Acceptance and launch gates: from “feels ready” to a checkable list
Sample gates: critical path click-through, form-to-CRM sample success, 301 samples without chains, correct locale switch targets, legal pages live, baseline security headers and backup policy confirmed, readable rollback. Name business and technical accepters—no “everyone in chat said OK.” Staff a 72-hour post-launch watch on 404s and indexation anomalies. Tie gates to milestones and payment in the scope doc. If you are chartering or RFP-ing, send sitemap, metric baselines, and integration list to hi@keshangxian.com (Shanghai); we compress a signable scope through discover→design→build→launch.
Checklist
- 12–4 success or do-not-regress metrics have baselines and 30/90-day windows
- 2Full URL inventory done; high-value pages have keep/301/retire decisions
- 3Template types capped; dynamic widgets labeled first-ship or phase-two
- 4Forms/CRM/analytics/pixels and owners written into scope appendices
- 5Non-goals, 301 deliverable, and launch gates mutually signed
Key takeaways
- An enterprise website rebuild scope checklist centers metrics, inventory, template caps, and non-goals—not framework shopping.
- SEO migration and integrations are contractual duties; parking them under “recommendations” is post-launch blame theater.
- Checkable launch gates beat aesthetic consensus if you want the rebuild to converge on a date.
FAQ
- If we only reskin without changing IA, do we still need a full scope checklist?
- Yes, thinner. Still lock templates, content ownership, performance, and launch gates. If URLs and nav truly stay, the 301 chapter shrinks to a “no path change” confirmation. Many “reskin only” projects quietly change nav—and full migration duties return the moment they do.
- Should marketing or IT own rebuild scope?
- Marketing/brand owns business goals and content; IT/security owns integrations, domains, and hosting; both co-sign non-goals and gates. Single-department scope that the other vetoes at launch is a classic delay path.
- How does KSX participate in rebuild scoping?
- In Discover we facilitate inventory and priority workshops and ship a URL-bearing scope note, template list, and migration table before design and build. We do not bury scope debate inside visual critique. Contact hi@keshangxian.com.