I have sat with a set of monthly agency reports that were all, individually, fine. Each agency had done what its contract said. The work was competent, the reports arrived on time and the commentary was accurate. Read side by side, they told me almost nothing about whether organic search was any closer to the commercial targets the business had agreed at the start of the year.

At Iron Mountain I was responsible for organic search across more than 60 country websites in 35 languages, with agency partners and in-house specialists spread across time zones, and that experience of contracted, competent work adding up to less than the sum of its parts is the problem I want to examine.

None of this is an argument against agencies. Most of the agencies I have worked with were good at what they were hired for, and the best of them understood their market's search behaviour better than anyone in head office ever would. The problem is structural. Markets, agencies and internal teams work to different priorities, different definitions of success and different levels of authority, and an agency can honour every line of its scope while contributing little to the objective the scope was written to serve. The World Federation of Advertisers published research in September 2026 covering 69 multinationals, and 35% of them said managing agency relationships had become more difficult, against 12% who said it had become easier. The reasons they gave were more stakeholders, more complex scopes, faster timelines and higher expectations around data. That is a fair description of international SEO.

Strategy dilutes between global and local

A global SEO strategy can be well written and still mean seven different things in seven markets. Local marketing teams have their own budgets, campaigns and product priorities, and an agency that reports day to day to a local marketing director will, quite reasonably, serve that director. Nobody is being difficult. They are responding to the authority structure they actually sit in rather than the one on the organisation chart, and the global strategy document is usually the weakest voice in that room.

The distinction I came to rely on is between decisions that have to be consistent everywhere and decisions that should be made as close to the searcher as possible. URL architecture, hreflang, canonicalisation, page templates and the definitions used in measurement are global. Google's guidance on multi-regional and multilingual sites treats them as one technical system, and a market that fixes its own version of a duplicate content problem will usually hand a new problem to a neighbour. Keyword targeting, content, local digital PR and the search features that matter in a given market are local, and a central team that overrides local judgement on those is wasting the expertise it is paying for. The mistake most organisations make is leaving the line between the two unstated. Too much central control and the agencies stop thinking. Too little and each market drifts into a strategy of its own while believing it is following the global one. A piece in California Management Review this summer on how multinationals are restructuring found the same thing about marketing in general: when nobody has said which decisions are global, regional and national, decisions slow down and accountability weakens. In SEO it also breaks things that are expensive to repair.

The north star has to decide things

Agencies are busy organisations, and busyness fills whatever space the brief leaves open. Rankings, content calendars, audits and monthly deliverables are all legitimate work. They are also the work an agency can control from its own desk, which is why they tend to crowd out the work that matters to the business. The agreed commercial objective was clear in the original briefing. By month six it is competing with a local stakeholder's product launch, an agency recommendation for a content hub that is sound in isolation, and an urgent technical issue that may or may not be urgent.

The north star only has value if it changes those decisions. That means being prepared to tell an agency that a recommendation is good and will not be done this quarter, and explaining why in terms the account lead can repeat to their own team without embarrassment. It means telling a local stakeholder that their request sits behind something less visible and more valuable. I did not enjoy all of those conversations and I did not win all of them. But an agency that has never heard no from the centre has no way of knowing what the centre values, and it will fill that gap with its own definition of good work. The WFA research found that brands rate the quality of briefing well above financial incentives as a driver of agency performance, which matches what I saw. A brief that is written once and never revisited is a formality, and agencies treat it as one.

Communication is harder than the meeting schedule suggests

A weekly call across three time zones with an agency contact who changed last quarter is a calendar entry, and it is easy to mistake it for communication. Time zones mean decisions wait a day, and a question asked on Friday afternoon in one market is answered on Tuesday in another. Language matters more than people admit. A recommendation written in careful English by a non-native speaker reads as less confident than it is, and I was slower to notice that than I should have been. Agency account staff move on, and the context that lived in their heads leaves with them. Inside the business, ownership is often unclear: the agency knows what to fix but has no idea who in engineering decides when it ships, and nobody has thought to tell them.

The distinction that matters is between an agency reporting what it did and an agency telling you what is not working, what needs a decision and where it is stuck. The second kind of conversation only happens when the agency believes it is safe to have it and that something will come of it, and both of those are things the client controls. Most of the global SEO lead's job here is plumbing. It is connecting agencies to the local marketing team, to the developers, to the content owners and to whoever can actually say yes. The other half is moving knowledge sideways. A technical fix or a content approach that worked in one market belongs to the whole estate, and if it stays inside one agency relationship because nobody in the centre carried it across, that is a management failure.

Reporting that proves activity

This is where misalignment becomes visible, and where it is easiest to misread as an agency problem. Different agencies arrive with different KPIs, different attribution models, different dashboards and different reporting periods. One market reports ranking improvements. Another reports organic sessions. A third has built a dashboard nobody outside that market can log into. None of them is wrong, and none of them explains whether the business is better off than it was last quarter.

Some of the disagreement is built into the tools, and it is worth knowing that before spending a meeting on it. Google's own help documentation says Search Console and Analytics will not reconcile: Search Console filters data for privacy, processes it differently, delays it by two to three days and, in every view except the 24-hour one, labels each day in Pacific Time. A market reporting in its own time zone will never match on the day. Search Console also aggregates under canonical URLs while Analytics uses the landing page the visitor actually hit, and a linked Analytics property shows only one domain. On an estate of 60 country sites that alone produces arguments that go nowhere. Knowing it does not fix the reporting problem, but it does stop the global team from treating reconciliation as a goal.

What useful reporting has to establish is simpler and harder. What changed, why it changed, whether the work contributed to the agreed outcome and what should happen next. Common definitions for the handful of measures that matter are a global decision, like templates. What counts as a good result in a mature market and an emerging one is a local one, and a central team that applies the same bar to both is being lazy. The WFA found that 89% of brands believe they get value for money from their agencies, but only 48% feel they have enough transparency into how that value is produced. Value and visibility are different things, and a report designed to prove activity is designed to keep them apart.

Accountability runs both ways

Agencies are often judged on results they cannot control. A technical recommendation depends on a global development team with its own backlog, a CMS that cannot do what the recommendation asks, an approval process designed for press releases and a legal team in another region with other things to do. I have watched sound recommendations age in a ticket queue for months and then heard the agency asked why performance had not moved. The business owns removing those barriers, making the decisions and setting the priorities, and if it will not do that it should stop measuring the agency on outcomes it has blocked.

The obligation runs the other way as well. An agency should say when an agreed piece of work is unlikely to deliver what the client expects, should challenge assumptions it believes are wrong and should explain the risk before the work starts. Most are reluctant to, and the contract is usually the reason. The WFA research shows labour-based agency fees have fallen from 54% of multinationals in 2011 to 19% now, with fixed-fee and output-based models taking their place. That is a sensible shift, and I would not argue for a return to timesheets. It is also a structure in which a delivered audit counts as a success whether or not anyone acts on it. If the scope measures the volume of deliverables, the agency will deliver volume, and it is reading the contract when it does, not being cynical. The scope is the client's document.

Alignment is continuous leadership work

The uncomfortable conclusion from all of this is that most of the failure sits with the client. Strategy has to be revisited when the business changes its priorities, and businesses change their priorities far more often than they update their agency briefs. Agencies need enough context to make good decisions without asking, access to the people who implement, and the standing to challenge the centre when the centre is wrong. Regional expertise should shape the global strategy. It should not be allowed to replace it with six local ones, and keeping that balance is a decision the global SEO lead makes every month, usually in small conversations nobody else notices.

Looking back at the estate I ran, the agency relationships that worked were rarely the ones with the strongest agency. They were the ones where somebody on our side had decided what the agency was for, had said so, and kept saying so as the priorities moved. Owning that decision was the part of the job I could not hand to anyone else.

Sources