Drupal Migrations: Why Small Businesses Underestimate the Real Cost
November 19, 2026A client came to me last year evaluating a Drupal migration with a competitive proposal in hand. Three vendors, three scopes, all clustering around the same number. “We migrate your WordPress site to Drupal 11. Content included.” He signed with the lowest bidder.
Three months into the project, the invoice was already at twice the original estimate. By go-live, it had doubled again.
I’ve seen this pattern more than a dozen times. The number on the proposal isn’t wrong exactly — it just isn’t scoping the same project you think it is. There are four places where migration cost reliably goes invisible on a quote, and every small business owner evaluating a Drupal migration should know them before signing anything.
Bottom Line for Stakeholders
| Hidden Cost Driver | What Gets Missed | Migration Type Most Affected | Typical Risk |
|---|---|---|---|
| Content model translation | Schema audit, field mapping, custom process plugins | D7 / WordPress → D11 | High — can double the labor estimate |
| SEO continuity | Pre-migration crawl, GSC URL inventory, Day 1 redirect map | All types | Medium-High — organic traffic loss at launch |
| Feature parity scope | “While we’re at it” additions multiplying the SOW | All types | High — scope doubles silently |
| Migration complexity tier | In-place upgrade vs. architectural rebuild | D7/WP vs. D10/D9 sources | Very High — these are different projects |
The Content Model Trap
The first question I ask on any migration scoping call is: what does your content look like in the database? Not “how many pages do you have” — what does the actual data model look like under the surface?
Most small business sites that come to me for migration have been built and modified by multiple people over several years. WordPress custom post types were added for a campaign and never cleaned up. Legacy Drupal 7 CCK field definitions accumulated across half a dozen content types with no consistent schema. When you move that into Drupal 11’s typed entity and field architecture, you’re not exporting rows and importing them somewhere else. You’re translating disparate schemas into typed entity bundles, field definitions, and media entities — and that translation requires a schema audit, field mapping decisions, and in many cases custom source plugins and process plugins to handle relational dependencies [1].
Drupal’s Migrate API addresses this with an ETL pipeline: source plugins extract data, process plugins like migration_lookup and entity_lookup resolve relational lookups across the entity graph, and destination plugins write the result into Drupal’s data layer. The map tables it generates track source IDs to destination IDs and support rollbacks via Drush [2]. This is powerful. It is also not a CSV export and import. It requires engineering time that most single-line “content migration included” SOW items don’t price.
Do that math against your quote.
SEO Continuity Isn’t an Afterthought
Search rankings belong to your public URLs. If your URL structure changes — and in a WordPress-to-Drupal migration, it almost always does — every inbound link, indexed page, and backlink points at a URL that no longer exists unless you redirect it. Google’s own site-move guidance is to map every old URL to its new one and serve permanent 301 redirects [3].
A pre-migration URL inventory built from two sources is non-negotiable: a live crawl for currently indexed pages, and a Google Search Console performance export to catch long-tail landing pages that crawlers miss because they aren’t linked internally but drive real traffic. The redirect plan built from that inventory needs to be active on Day 1 of cutover, not retrofitted two weeks after launch when the organic traffic drop shows up in analytics. In Drupal, the Redirect module manages those 301s, and its 404 logging shows you what the map missed [4].
Proposals that treat redirect mapping as a post-launch cleanup task are under-scoped by definition. That work belongs in the launch plan, not the post-launch punch list.
Feature Parity vs. Feature Creep
“While we’re at it, we could also redesign the homepage. And update the brand colors. And build that event calendar we always wanted.”
I’ve watched a 15-page business site scope into a $60,000 rebuild within eight weeks of kickoff because nobody pushed back on the “while we’re at its.” The migration became the design project became the product development project. All three are real work. None of them are the same project.
The goal of Phase 1 is feature parity: reproduce existing business utility on modern, maintainable architecture. If it wasn’t on the live site, it doesn’t go in the Phase 1 SOW. Enhancements go in Phase 2 — with their own scope, their own budget, and their own timeline, written after Phase 1 is complete and you understand the new platform.
Pick the phase you’re in. Then do the other one.
Two Drupal Migration Budgets Nobody Compares Side by Side
Here’s the distinction most migration proposals bury: an in-place major version upgrade and a replatform from a legacy CMS are not the same kind of work. They share a word — “migration” — and nothing else.
A well-maintained Drupal 10 to Drupal 11 upgrade preserves your content models, views, and configuration. The work is dependency auditing, Composer updates, and resolving deprecated APIs before they become hard blockers. One UK agency puts that at two to four weeks and roughly £6,000–£15,000 for a well-maintained site, and four to eight weeks and £15,000–£35,000 when abandoned modules and removed APIs have to be worked around [5]. Rates vary by market, but the shape holds: weeks, not months. If you’re still on Drupal 9, the path is sequential — Drupal 9 to 10, then 10 to 11. No skipping major versions [6].
A replatform from Drupal 7 or WordPress is a rebuild. Drupal 7 reached end of life on January 5, 2025 [7]. There are no community security advisories and no supported upgrade path in the traditional sense. You’re building a new Drupal 11 application from scratch: a new theme in Twig and Single Directory Components, custom module logic re-architected for Drupal’s service container, and a full ETL content migration via the Migrate API. The same agency puts a Drupal 7 rebuild at three to six months and commonly £40,000–£120,000 or more, and says plainly that the range is wide because it’s really a rebuild budget [5].
Same word. Different project.
If a proposal gives you roughly the same quote for both, ask which one they’re actually building.
What to Ask Before You Sign a Migration Proposal
You don’t need to be a Drupal engineer to vet a migration proposal. You need four questions.
Is schema audit and field mapping in the SOW? “Content migration included” as a single line item is not an answer. Push for explicit scope on schema discovery, field mapping, and custom plugin development for complex fields.
What is the redirect deliverable, and when does it deploy? You want a redirect map built from a site crawl and GSC export, delivered before cutover — not “we’ll handle redirects post-launch.”
Are migrations run via Drush or the web UI? Production migrations run from the command line, using the Drush commands the Migrate Tools module provides [2]. If a vendor doesn’t know what drush migrate:import is, that’s information.
Is Phase 1 strictly feature parity? Get this in writing. Any SOW that bundles redesign, new features, or significant UX work into the initial migration phase is setting up the scope problem I described above.
A well-scoped migration isn’t cheap. But a poorly-scoped one costs more — just on the other side of the contract, when the invoices start arriving and there’s no clean exit.
If you’re evaluating a migration proposal or trying to sanity-check a quote against what the work actually requires, let’s talk.
Mapping out that scope before you sign is exactly what I do.