Case Study 1 — Marks & Spencer's 2014 Relaunch: When a Well-Planned Migration Still Hurt

Type: Real, public case analysis. The factual claims about the relaunch come from Marks & Spencer's own public reporting and the contemporaneous financial and trade press. The SEO-specific reading is labeled as practitioner analysis (Tier 2), because clean SEO attribution is not possible from the public record. No traffic figure is invented, and where the cause of the decline is genuinely mixed, that is said plainly. The lesson survives the uncertainty — in fact, the uncertainty is part of the lesson.

Background

Marks & Spencer (M&S) is one of Britain's largest and oldest retailers. For roughly seven years, its e-commerce site, M&S.com, had run on Amazon's e-commerce platform — a common arrangement in that era, in which a big retailer effectively rented Amazon's technology to run its online store. By the early 2010s, M&S wanted its own platform: its own technology, its own design, its own control over the customer experience and the data. So it built a completely new M&S.com in-house and, in February 2014, launched it — replacing the Amazon-powered site with a fresh, redesigned, self-owned platform in a single major relaunch.

In the vocabulary of this chapter, this was one of the most ambitious migrations a large business can attempt: a platform change (off Amazon's systems onto M&S's own) combined with a full redesign. Two of our five migration types, stacked into one launch. M&S was not careless about it — this was a heavily resourced, long-planned, strategically central project for the company. And it still hurt. That is precisely why it is the right case to open this chapter: it shows that migrations are dangerous even when a sophisticated, well-funded team does them on purpose, with plenty of notice.

The issue: the disruption the company itself had flagged

In the quarter that followed the launch — the thirteen weeks to the end of June 2014 — M&S reported a notable decline in its online sales compared with the same period a year earlier. This was widely covered in the mainstream and financial press at the time (the reported drop was in the high single digits of percent for that quarter's online general-merchandise sales). Crucially, M&S did not blame a mystery; it publicly linked the decline to the disruption of the relaunch itself. The company had gone in expecting a short-term hit while customers adjusted to the new site, and it got one.

Why would a brand-new, better-resourced site lose sales? Several documented factors, and this is where the case becomes a teaching tool:

  • Customers were required to re-register. The most-reported friction point: existing M&S.com account holders could not simply log in to the new site with their old credentials; they were asked to create new accounts. Every re-registration is a moment a loyal customer can abandon — and many did, at least temporarily. This is a conversion wound, not strictly an SEO one, but it illustrates the chapter's deeper point: a migration is judged by the business it protects, and a technically "successful" cutover that adds friction to the path to purchase has still failed the number that matters.
  • A wholesale platform-and-design change all at once. The new site changed the technology, the templates, the navigation, and the experience simultaneously. When everything changes together (§21.1's "change one variable at a time" caution), a dip is hard to attribute and easy to trigger from any of a dozen directions.
  • The general reprocessing every migration incurs. A move of this scale hands search engines an enormous reprocessing job — new URLs, new templates, redirects to follow, signals to reconsolidate — during which some volatility is expected regardless of execution quality (§21.7).

📄 Read the Report

text FIGURE C21.1 — "A giant's migration, read honestly" [after public reporting; SEO read = Tier 2] THE SOURCE M&S public results reporting + contemporaneous press on the Feb 2014 M&S.com relaunch. WHAT'S THERE A combined platform-change + redesign; a required customer re-registration; a reported drop in online sales the following quarter, which the company linked to relaunch disruption; recovery over subsequent quarters as the site settled. WHAT IT SHOWS Migrations are business-critical and risky even for a sophisticated, well-funded team with long notice. The biggest self-inflicted wound sat at the SEO/UX seam (forcing re-registration), and stacking two migration types raised the stakes. WHAT IT DOESN'T The public record does NOT isolate an "SEO traffic loss" number; the decline blends UX friction, replatform disruption, and search reprocessing. Anyone citing a precise "M&S lost X% of organic traffic to the migration" is inventing the attribution. THE MOVE For your own migration: don't stack changes you don't have to; preserve continuity of the customer path (logins, carts, key URLs); expect and plan for a dip even at scale. THE LESSON A migration is graded on the business it protects and how fast it recovers — not on how good the new site looks on launch day.

What it shows

Three lessons fall out of this case, and they anchor the chapter.

  1. Size and resources do not exempt you. M&S had money, time, and expertise, and still took a real hit. The discipline in this chapter — inventory, redirect map, checklist, monitoring — is not a small-business consolation for those who can't afford to "do it properly." It is what everyone needs, because the risk is structural: you are asking search engines and customers to relearn your site. The giants dip too.
  2. The dangerous wounds cluster at the SEO/UX seam. The single most-cited problem was re-registration — a decision that made sense for the new architecture and quietly punished the existing customer base. Migrations tempt teams to optimize for the clean new system over continuity for the people and the signals that were already there. Continuity is the whole job: keep the URLs, keep the links, keep the logins, keep the path to purchase.
  3. Attribution is genuinely hard, and honesty about that is the professional posture. The public data cannot cleanly separate "SEO" from "UX" from "the market that quarter." A good practitioner says so — "the relaunch cost us, the causes were mixed, here is what we can measure and what we can't" — rather than inventing a tidy single-cause story. This is the same evidence honesty the whole book insists on, applied under real commercial pressure.

Outcome

The story does not end at the dip, and that matters. M&S's online sales recovered and returned to growth over the following quarters as the new site stabilized, customers re-registered, and the platform matured. The relaunch was a genuine setback in the near term and, over the longer arc, delivered the strategic control M&S had wanted — its own platform to build on. Read through §21.7, this is the normal migration curve playing out at corporate scale: a real dip at launch, a period of turbulence, and recovery over a horizon of quarters rather than weeks (large, complex sites take longer). The mistake would have been to judge the whole project by that first painful quarter.

The lesson

Even a well-planned migration by a sophisticated team can hurt — so plan for the dip, protect continuity above novelty, and judge the move on recovery, not on the launch. M&S did some things right (it expected the hit and communicated it) and paid for some things dearly (re-registration; changing everything at once). For your own migrations, the takeaways are concrete: preserve the customer's path and the search engine's map with equal care; resist stacking migration types when you can sequence them; expect turbulence even if you do everything well; and hold your nerve through a normal recovery instead of panicking at the first bad quarter. The biggest brand in the case had a year to prepare and still dipped. Respect the operation accordingly.

Discussion questions

  1. M&S combined a platform change and a redesign in one launch. Using §21.1's "change one variable at a time" principle, describe how you might have sequenced this migration instead — and name the business reasons M&S might have had for doing it all at once anyway.
  2. The most-cited problem was forcing customers to re-register. Explain why this is best understood as a wound at the "SEO/UX seam" rather than a pure SEO or pure UX problem — and what "preserving continuity" would have meant in practice.
  3. The public record cannot cleanly separate SEO, UX, and market causes for the sales drop. Draft two sentences you would put in a client report that are honest about the uncertainty while still being useful to a decision-maker.
  4. M&S's online sales recovered over subsequent quarters. How does this map onto §21.7's "normal migration curve," and why is it a mistake to grade a migration of this scale by its first post-launch quarter?
  5. This is a real, public case, yet the chapter refuses to attach a precise "organic traffic lost" figure to it. Explain why inventing that number — even a plausible-sounding one — would violate the book's evidence standards, and what you can legitimately say instead.