Somewhere in your organisation there is a website you have never seen. A regional sales office had it built four years ago, in a language nobody at head office reads, carrying product descriptions that were accurate at the time. It ranks above your own English site for your own product names in that market, and nobody has ever mentioned it.
Manufacturers who sell through regional offices and distributors accumulate this quietly. Each site was created for a defensible reason — a local team needed something in the local language, faster than head office could supply it. The result, five years on, is a set of properties representing the same company with different specifications, different contact routes and different levels of currency, all competing for the same product queries.
The properties nobody counted
| Property | Why it was created | Who maintains it now |
|---|---|---|
| Corporate site, German | The original presence | Marketing, properly |
| Corporate site, English | Export sales channel | Marketing, less often |
| Regional office microsites | Local team needed something faster | The local team, until they left |
| Distributor product pages | Not yours at all | The distributor |
Row four is a different category and deserves separate treatment: those pages are not yours, cannot be edited, and in many markets outrank you for your own product designations. That is not automatically bad — a distributor ranking well is a distributor selling — but it becomes a problem when the specifications on their page are two revisions out of date and the enquiry never reaches you.
Something else makes it worse. New sites keep appearing and old ones are almost never shut down. When a product line is dropped its microsite stays up; when a campaign finishes its landing page stays up; when a division is renamed both names survive. Half a decade later the maintained properties are shadowed by roughly the same number of abandoned ones, which continue to load, continue to be catalogued, and continue to bid for the very product names the live sites depend on. No list anywhere contains them, since no department has ever claimed ownership.
The middle column of the table is worth reading generously: each of these decisions made sense when it was taken. There was a reason, a responsible person and a deadline. What never happened was the moment somebody looked at the resulting whole, and in an organisation growing through regional offices that moment does not arrive by itself. It has to be scheduled, and the trigger is usually the first time an outsider asks for defensible figures.
The work sits between the properties, not inside them
- Separate credentials everywhere. Often including accounts belonging to regional staff who have moved on. Nothing goes wrong until something does, and then the access you need is the one you cannot get.
- Figures that refuse to line up. One covers calendar months, another rolling weeks; the columns differ; so do the layouts. Welding them into one statement eats more time than reading the statement afterwards.
- Nothing to measure against. There is no way to say whether the English presence is performing except by setting it beside the rest, and that requires figures compiled the same way for each.
- Specifications drifting apart. The worst of the four. Three versions of the same product data, all published, none flagged as superseded.
The fourth point is where this stops being a marketing problem and becomes a commercial one. A buyer who finds an obsolete rating on a microsite, specifies against it, and discovers the discrepancy at delivery does not blame the microsite. They blame the manufacturer, and the cost is a returned shipment rather than a lost click.
There is a related effect that only appears when all the properties are viewed together: the same preparation done twice. Two people research the same product queries for two properties without knowing about each other, because intermediate results are not stored anywhere in common. Both feel productive and neither is doing anything obviously wrong. It surfaces only when both properties end up addressing the same query with the same argument and both perform worse than one properly resourced page would have.
The difference once everything reports to one place
Labels rather than separate accounts
Each property carries a marker that behaves as a filter everywhere.
- One sign-in covers all of it. More than thirty-five screens sit behind it, from position tracking through to observation of assembled answers.
- A filter holds wherever it is applied. Restrict to one property and the restriction survives into reporting, tasks and exports rather than applying only where it was set.
- Eleven services connected. Google, Analytics and the rest converge in one place instead of existing as four parallel copies.
- Google authorised once, not four times. Mail, Search Console and Analytics hang off a single permission, which disposes of the credentials problem at its root.
Submission capacity is pooled, not multiplied
The number multi-property plans consistently get wrong.
- The daily figure belongs to the account. A thousand addresses a day spreads across everything inside it. Four properties divide that; they do not each receive it.
- Ten thousand addresses per batch. Taken in at once and worked through against the daily ceiling, making a full catalogue submission a multi-week operation by arithmetic.
- Two batches active, twenty held back. Filing several property migrations at once means setting the order yourself or letting filing time decide it.
- Sitemaps three levels deep, a thousand per batch. Comfortably enough for a catalogue split by family and variant, provided the index files are genuinely linked.
The first bullet of the second block is where these plans come apart. A team assuming four thousand addresses a day because it has four properties plans a migration schedule on that basis, receives a thousand, and concludes the process is slow rather than the assumption wrong. Across a large catalogue that costs weeks, and it costs them invisibly, because the batches are running — just not at the assumed rate.
Bundling the Google permission into one gets less credit than it deserves. Twenty years of organic growth leaves access scattered across a handful of individuals' own accounts, and at least one of those individuals will have departed some time ago. Day to day this causes nobody any trouble. Then a property vanishes from the record, or a warning comes through, and the account you suddenly need turns out to be exactly the one no longer reachable. Half a day spent sorting it out closes the exposure for good, and of everything listed here it is the only item with no case for delay.
Spending one allowance across four properties
| Situation | Sensible division | Reasoning |
|---|---|---|
| English catalogue being extended | 800 there, 200 elsewhere | Only one property is changing |
| Nothing much happening | Split by how many pages each holds | Nothing argues for preferring one |
| A microsite being retired | Full capacity for two weeks | Redirects need picking up quickly |
| Ahead of a major trade fair | Priority to the English site | That is where visitors go afterwards |
Set this out on paper once and the argument that flares up whenever two teams need throughput in the same week simply stops happening. The document also makes plain that somebody is choosing how to divide the pool rather than leaving it to whatever the system does by default, and that the choice is revisited when the situation shifts. Where regional offices are involved there is a further benefit: nobody can suspect the centre of quietly looking after one market ahead of the others.
Be clear about what pulling everything together does and does not deliver. The workload stays the same size; what changes is that you can now see it and weigh one part against another. Expect the opening weeks to be an unpleasant experience, because for the first time a single screen spells out how many properties are running with nobody behind them and how thin the enquiry flow is from most of them. That unpleasantness is the point of the exercise. It is what lets somebody say which two are worth attention and which two can be left to sit without anything being lost.
What to do with the sites you inherited
Microsites duplicating the main site
Same products, same specifications, translated once and never updated. These compete with the corporate site and carry outdated data.
- Redirect page by page
- One hop, not a chain
Sites serving a genuinely local purpose
Local service capability, local certifications, local stock. Content that does not exist on the corporate site and could not sensibly live there.
- Bring under central maintenance
- Link from the corporate site
Keeping the retire-or-keep decision recorded in one place also stops a later arrival from quietly reviving a site that looked like an oversight.
The distinction is not about language. A translated copy of the corporate site is a duplicate whatever language it is in; a page describing a service workshop in a particular country is unique content that belongs somewhere. Applying that test to a list of inherited properties usually retires more than half of them, and the ones that remain become easier to justify maintaining properly.
Retiring a property is not deletion. Each page redirects permanently to its counterpart on the corporate site, in a single hop, and the accumulated standing of the old address transfers rather than being discarded. Switching a microsite off without redirects throws away whatever it had built up, which after four years is rarely nothing. Doing it properly is a day of mechanical work per property and repays itself immediately in the reduced maintenance burden. Watching the transfer take effect inside a shared overview is what turns it from an assumption into something demonstrable.
One data set, four kinds of reader
Monthly, per market
Wants to know which destinations are producing interest. Table views of fifty to two hundred rows per screen suffice, since filtering happens anyway.
- Filtered by market
- Sorted by rate, not volume
Quarterly, comparative
Wants a handful of figures beside the same quarter a year earlier. Printed files hold up to two hundred and fifty rows, well beyond what is needed.
- Properties rolled together
- No query-level detail
Monthly, their market only
Wants their own figures without a comparison against sister markets. The same data, filtered differently.
- One market each
- No cross-comparison
Rarely, under pressure
Wants a traceable history per property. Data files reach ten thousand rows and drop straight into whatever analysis is being run.
- Separable by label
- Produced against a deadline
The second and third audiences sit in tension, which any organisation with regional offices will recognise. The cross-market comparison is the most useful view for head office and the least welcome one for a regional manager. Separate outputs resolve it cleanly: each reader gets the view matching their responsibility, and both come from the same underlying figures.
How often to send them matters as much as what they contain. A monthly rhythm suits the sales-facing version, since each edition prompts somebody to do something. Anything travelling up the organisation belongs on a quarterly cycle — its purpose is different, and running it more frequently mainly manufactures arguments between the centre and the regional teams. Force both onto one timetable and you get either wasted effort or stale figures; with marketing staffed by two people, generally both.
Asking a question across four properties
Once there are several properties, most of the hour goes on getting to the right screen rather than on thinking about what it says: locating the report, setting the restriction, recalling where the previous quarter's comparison was saved. Being able to type the question in ordinary words cuts that out. Depending on what the question actually requires, anywhere from none to three blocks of data get pulled, and the twenty most recent exchanges stay on hand so follow-ups make sense.
In practice the question becomes which property lost the most positions last month, asked once rather than assembled four times. Four filters bound what the answer covers, and with several properties the property itself is the one that matters most. Running the monthly review inside one workspace removes most of the searching, and because the wording is retained, the identical review runs again four weeks later without being rebuilt.
Stopping at twenty exchanges is a sensible design choice rather than a shortcoming. Anything worth asking about four properties gets settled inside half a dozen or so back-and-forths; if it is taking far longer, the opening question was almost certainly too wide. Hitting the ceiling again and again is your cue to break the question in two, and two narrow questions get there sooner than one sprawling thread ever does.
What several properties cost
Billing attaches to the domain rather than the account. Four properties on the entry tier, with a modest network step shared across the account, come to 616 dollars a month and 7 392 across a year. A mixed arrangement — the English site on the full tier, three others on the entry tier, plus five encyclopedic slots — is 997 monthly and 11 964 annually.
| Setup | What it consists of | Per month | Across a year |
|---|---|---|---|
| Everything on the lower level | 4 × AutoSEO with 20 network slots (20 $) | 616 $ | 7 392 $ |
| Mixed | 1 × FullSEO, 3 × AutoSEO, plus 5 encyclopedic (50 $) | 997 $ | 11 964 $ |
| Once tidied up | 2 × AutoSEO | 298 $ | 3 576 $ |
Working out which properties earn their place means establishing what each one can currently be found for — a matter for keyword research and not for whoever argues hardest in the meeting. A technical review answers whether each ships and gets taken in properly. Where two of them repeat the same material word for word, the answer is on-page work and not a bigger budget. Doing all three out of a single account is the precondition for comparing the properties against each other at all.
On sequencing: nothing requires all four properties to be brought in at once. Markers and shared reporting work from the first one, and each additional property slots in without anything being rebuilt. In an organisation where regional offices control their own budgets, a staged approach is also easier politically than a decision binding every market to the same date.
Bring every property into one view
Questions from organisations selling across markets
Does each property get its own thousand a day, or do they share?
They share — the ceiling belongs to the account, and everything inside it draws from the same thousand. While a catalogue is being rolled out somebody therefore has to decide the split explicitly. Left alone, whichever batch went in first will consume the capacity, importance notwithstanding.
Our distributor outranks us for our own products. Is that a problem?
Not in itself — a distributor ranking well is a distributor selling. It becomes a problem when their specifications are out of date, because a buyer specifying against obsolete data blames the manufacturer. The practical response is to supply distributors with current data in a form they can publish, and to make sure your own page exists and is current.
Should we retire the regional microsites?
Most of them, yes. Anything that is a translated copy of the corporate site is a duplicate competing with the original and carrying stale figures. Anything describing genuinely local capability — a service workshop, a local approval, local stock — is unique content worth keeping, but it belongs under central maintenance rather than with whoever created it.
Must every property sit on the same level?
Not at all, and it is usually wrong to do so. Since the charge attaches to each domain, combinations are free. Put the upper level where a real person takes part in choosing which queries to chase — for a company selling abroad, that is the English presence — and leave everything else on the lower one. Across four properties the arrangement described above works out at 997 dollars a month.
Can regional offices see each other's figures?
Only if you configure it that way. Markers behave as filters in every view, so each office can receive an output containing its market alone while head office retains the cross-market comparison. Both come from the same underlying data, which is what makes the two views consistent with each other.
What happens if a product line is divested?
Its marker allows the full history for that line's pages to be exported as a data file of up to ten thousand rows. That is exactly the documentation requested during diligence, and having it ready shortens the exercise from weeks to days rather than reconstructing it under a deadline.