Enterprise B2B SEO is mostly an organisational problem

At a company with 40 employees, fixing a canonical tag takes an afternoon. At a company with 4,000, it takes a quarter, because the CMS is owned by a platform team with its own roadmap, the change needs approval from brand and legal, and the person who understands the template left in 2023.

Nobody at the enterprise lacks the knowledge to fix it. The constraint is entirely organisational, and pretending otherwise is why so many enterprise SEO engagements produce a strategy document and no change.

The five recurring constraints

Nobody owns the whole site. Marketing owns the top-level pages, product marketing owns solution pages, a regional team owns the country sites, a documentation team owns docs, a demand gen team owns landing pages, and someone in HR owns careers. Six roadmaps, six sets of priorities, one domain sharing the consequences.

Legal review is a fixed cost per page. In financial services, healthcare, and defence-adjacent categories, every public claim needs review. Five weeks is common. This does not make content impossible; it makes high-volume publishing impossible, which changes the entire strategy.

The CMS is a decade old and heavily customised. Changes require a release cycle. The templates contain assumptions nobody documented.

Regional autonomy produces duplicate content. Eleven country sites, each with a locally written version of the same product page, competing with each other and with the .com. Nobody agreed to this; it accumulated.

Attribution is owned by a team that does not report to marketing. Which means the SEO program is being judged by a dashboard it cannot influence.

What to do about the ownership problem

The intervention that has worked for us is dull and effective: a written map of who owns which URL patterns, agreed by all parties, with a named individual per pattern.

That document does not exist at most enterprises. Producing it takes about three weeks of meetings and immediately surfaces the pages nobody owns, which are usually the ones with the worst problems.

Then a standing forum — monthly, an hour, with a decision-making representative from each owning team. Not a status update. A decision meeting with a short agenda of cross-cutting changes that need agreement.

The alternative, which is a central SEO team issuing recommendations into six different backlogs and hoping, does not work and I have watched it fail at four large organisations.

Working with legal rather than around it

Two things reduce the review burden substantially.

First, a pre-approved claims library. Legal reviews a set of standard statements once — the compliance certifications you hold, the performance numbers you can cite, the customer results you have permission to use — and writers draw from it without triggering a new review. On one client this cut average review time from five weeks to nine days for pieces that stayed within the library.

Second, batching. Submitting eleven pieces on the same theme together gets reviewed faster than eleven submissions across three months, because the reviewer builds context once.

And accept the constraint in the strategy. If your realistic throughput is four pages a quarter, the plan needs to be four excellent pages a quarter aimed at the highest-value clusters, not a content calendar that will jam.

Consolidation, which is usually the biggest single win

Large organisations accumulate duplicate and near-duplicate pages. Two product pages for the same product from a rebrand. Regional variants of identical English content. Campaign landing pages that were never taken down. Acquired companies’ sites redirected badly or not at all.

On a 12,000-page enterprise site we audited, 3,100 pages were near-duplicates of another page on the same domain. Consolidating them was a nine-month project across four teams, and it produced a larger traffic increase than any content work that year.

The pattern is consistent enough that on any enterprise engagement, consolidation is the first thing we scope. It is unglamorous, it requires no new content budget, and it is politically hard because every page belongs to someone who will defend it.

International, properly

The most common enterprise mistake is eleven country sites all serving English content to different markets with no hreflang, competing with each other.

The correct setup is not complicated but it has to be complete. Hreflang annotations that are reciprocal and include a self-reference, an x-default for the fallback, and consistency between what hreflang claims and what the page actually serves. Partial hreflang is worse than none because it produces inconsistent signals.

The harder question is which markets deserve genuinely localised content rather than a translated copy. Our rule of thumb is that a market gets a real content program if it represents more than about 8% of pipeline or is a named strategic priority. Everything else gets the English page with correct annotations, which performs better than a bad translation.

Reporting upward at scale

Enterprise stakeholders need a different report from a Series B marketing lead. The things that land, in our experience:

Influenced pipeline by business unit, because that is how budget is allocated. Share of voice against a fixed competitor set, because executives think competitively. Progress against the agreed roadmap, with blockers named and attributed, because half the report’s job is to make organisational friction visible to people who can remove it.

That last one is the one that changes outcomes. A slide saying “these four initiatives are blocked on the platform team’s Q3 roadmap” gets things unblocked in a way that no amount of explaining the value of SEO ever will.

The realistic timeline

At enterprise scale, expect the first two quarters to produce mostly diagnosis, governance, and consolidation, with limited visible traffic movement. Real gains typically start appearing in quarter three and compound heavily through years two and three, because the changes are structural rather than incremental.

If your sponsor needs a win in the first six months, find one small, self-contained property they own outright and produce it there while the larger work proceeds.