Shipping runs in English. Your German customers work in German. That is not a contradiction — it is a reporting problem, because the tool you are using assumes a country and a language are the same thing, and in this trade they are not.
An agent in Singapore, a forwarder in Rotterdam, a carrier's representative office in Piraeus: all three sell into the German market and all three write their websites in English, because English is the working language of the industry. Their Search Console reports look weak for Germany. The conclusion drawn is usually that German demand is thin. That conclusion is wrong roughly half the time.
What the numbers actually show is a market where the operational language and the procurement language diverge, and where the reporting has no column for that distinction unless somebody builds one.
The trade language and the buying language
A Hamburg-based operations manager reads your English documentation without difficulty. The purchasing department at the same company issues its tender in German. The quality function checks certifications in German. So the same customer generates queries in two languages depending on who inside it is looking — and only one of those is served by an English-only site.
The practical consequence is that a headline figure for Germany means very little. It aggregates a strong English-language position among operations staff with a weak or absent German-language position among the people who sign. Reported as one number, the two cancel out into something unremarkable, and the territory looks unpromising when in fact half of it has never been addressed.
Country crossed with language
The cross that most operators never make
Four combinations, four different people, and only one of them served by an English-only site.
- Germany, English query. Operations, chartering, documentation staff. They read your site as it stands and this cell usually looks healthy — which is exactly why it gets mistaken for the whole picture.
- Germany, German query. Purchasing, quality, customs and legal. They search in German, and an English page competing for a German query starts behind regardless of how good it is.
- Outside Germany, English query. Frequently the largest cell by volume and the least useful commercially, unless you actually serve those origins. Separate it before it flatters the total.
- Outside Germany, German query. Small, and mostly Austrian and Swiss. Relevant only where you can genuinely handle those flows, which is an operational question rather than a content one.
The second cell is where tenders originate and where an English-only site performs worst. A purchasing department qualifying a forwarder is not reading a service description; it is looking for named authorisations, a stated liability position and evidence that somebody at your end answers in German. None of that is in the English material either, so translating the English material produces nothing useful.
"LCL consolidation, Far East to Hamburg, weekly"
Three facts an operations person recognises instantly and a purchasing department can put into a comparison.
- Names a service and a lane
- States a frequency
"Tailored logistics solutions for your global supply chain"
True, unfalsifiable, and identical to what four hundred other agency sites say in the same words.
- Matches no specific query
- Distinguishes you from nobody
The gap between those two cards is where most of the available improvement sits, and closing it requires no budget approval from anybody. Every fact on the upper card is already known to the operations desk — it appears in quotations and in service schedules and simply never made it onto a public page. Publishing it is an afternoon's editing, and it improves performance in the English-language cell at the same time as it opens the German-language one.
- Lanes, named individually. Not "worldwide coverage" but the corridors where rates, partners and experience actually exist. Naming everything results in being retrieved for nothing.
- Services and exclusions. What is handled, and explicitly what is not. The exclusion carries more weight than the list, because it makes the list credible.
- Authorisations with scope. Designation, extent and issuing body in running text. As a downloadable file they are, for search purposes, absent.
- Response commitments with a unit. "Quotation within four working hours" can be checked and compared. "Prompt" can be neither.
Which view answers which question
Within the Semalt workspace the Google reporting arrives as eight separate screens rather than one report with filters. For a bilingual property that arrangement helps, because every interesting comparison here is a cross-section rather than a total.
| Who asks | The question | The screen |
|---|---|---|
| Principal or head office | Is Germany worth a dedicated effort? | Property overview with clicks, impressions and health flags |
| Commercial | Which lanes and services are people searching for? | Keywords with position history and click trend |
| Whoever maintains the pages | What is carrying the result? | The same figures broken down by landing page |
| Anyone debating translation budget | Does German-language demand exist here? | The country split, cut by the language of the query |
| Whoever signs the invoice | Is the position improving or slipping? | Arrivals and departures across the leading three, ten and thirty places |
Of the five, only the bottom row reports movement instead of a snapshot, which is why it alone warrants weekly attention — and a quarter of an hour covers it. Everything above it moves slowly enough for a monthly or quarterly look, and operators who try to review all eight every week stop reviewing anything by the third month.
The competitor with no sales team
Search Console stops at your own domain. Which other domains hold the same lane and service terms, how strong they are, and whether the field is tightening is entirely absent from it. Six further views on the position-tracking side cover that gap, and in this trade they surface the same finding with tedious regularity.
The forwarder you meet at conferences
A comparable house on the same lanes, known to your commercial team, discussed at every review.
- Shares surprisingly few search terms
- Competes at the tender, not at the shortlist
The freight exchange or trade portal
Holds the lane and procedure terms, often with fifteen years of domain history and no interest in your relationships.
- High count of shared terms
- Frequently a channel rather than a rival
The second card deserves a decision rather than a reflex. Competing head-on with an established portal for a generic lane term commits several years to an uncertain outcome. Being listed on it produces enquiries within weeks. For an operator with a small marketing function, the listing is the defensible route — and it rules nothing out: narrower terms where the authority gap is small still justify a page of their own.
One practical note on frequency for a bilingual property: check once a month that the two language trees still reference each other correctly. Language annotations break quietly during platform work, and the symptom — the German page disappearing from results while the English one stays — looks exactly like a content problem and is not one. It takes five minutes and prevents a quarter of misdirected effort. Index coverage sits beside positions in the workspace, which makes the disappearance visible rather than inferred.
Half your enquiries are invisible
In this trade a large share of first contact arrives by email to a named person, through an agent, or at a conference. The person read three pages first. Nothing in any analytics package records that, and nothing ever will.
The remedy is one field in whatever system records enquiries, filled in during the first exchange. It is tedious, easily skipped, and it is the only bridge between what the reporting shows and what the business decides. Without it, a company argues for a year about whether German-language pages are worth writing while the evidence sits unrecorded in a commercial inbox.
A review cadence that survives peak season
| Cadence | What gets reviewed | What it decides |
|---|---|---|
| Every week, a quarter of an hour | Which terms arrived in and dropped out of the leading positions | Repair technical losses; point internal links at the new arrivals |
| Every month | Landing pages, and the country figures split by query language | Whether demand in German is actually building |
| Every quarter | The domains sharing your terms and the consolidated score | Which German-language pages to commission next |
| Once a year | Enquiries with a recorded source, against the prior year | Whether the whole exercise covered its cost |
The weekly slot exists for two reasons and no others: catching a dropped position while the cause is still traceable, and noticing a term that has risen on its own — the point at which an internal link from a well-visited page yields the most and six weeks later yields little. Traffic counts, impression totals and the mean position are kept out of that fifteen minutes on purpose: at these volumes they shift for reasons that have nothing to do with anything anybody did.
Making the German number legible elsewhere
Reporting that leaves the country serves a different purpose from reporting used at the desk. It will be read by somebody who has never looked at a German results page and who is placing this territory against five others on the same slide.
Five items are usually enough: the enquiry count with recorded origin against the same period last year, whichever three pages carried that result, which terms entered and left the leading positions, how much of the demand was genuinely German-language, and a single explanatory sentence. Full term lists, competing domains and country breakdowns belong in the export for whoever does the work. Reports carry your own mark and colours, which decides more often than it should whether a document gets forwarded or filed.
Four things that move the figures and are not your doing
Each of them belongs in the report as a marker, or the comparison will be read as performance.
- Rate volatility. Sharp movements in spot rates generate a burst of searching that has nothing to do with your site and everything to do with buyers reconsidering their arrangements.
- Disruption events. A strike, a canal closure, a congestion episode: within hours, terms appear that did not exist before and vanish afterwards. The entries-and-exits view catches these; a monthly total does not.
- Contract season. Annual tender cycles concentrate qualifying activity into a few weeks. Comparing a window that contains one against a window that does not produces a percentage with no meaning.
- Regulatory changes. A change in a customs or reporting requirement generates search demand for the procedure, not for a provider — but a page that explains it well attracts exactly the audience that will later need a provider.
Marking these four on the chart takes a minute and changes how the report is read. Without them, a rate-driven spike gets attributed to marketing and a quiet quarter gets attributed to failure — and both readings survive into the following year's planning, where they do real damage.
The record that outlives the person
Marketing responsibility in this trade changes hands frequently, and each change erases whatever was not written down: which terms were chosen, why, and what the previous holder had concluded. A record that survives that change is worth more than most of the analysis it contains.
Each project inside My SEO Stream keeps a dated timeline: the assistant's answers, generated reports, new placements with the referring domain's rating and traffic, items open and closed, and campaign notices. A routing stage decides per question which blocks are relevant — Google reporting, position data, campaign state, or figures you supplied — and loads between none and three of them. Output appears as it is written, twenty prior messages stay in view, and the whole record can be searched as text. That last function is the one nobody values in month one and everybody uses in month eighteen.
Term selection, placement activity and the suggestions made against existing pages all live alongside the reporting inside My SEO, so nothing has to be exported from one system in order to be compared with another. Where the German-language terms come from in the first place is a separate question, and our approach for keyword research describes it; the editorial split between English and German material sits in content strategy, and the prerequisites under technical SEO.
Open the German figures in the dashboard
Questions from agents and forwarders
Our whole industry works in English. Do we need German pages?
The operational side of your customers works in English; the side that awards contracts frequently does not. Settle it with data rather than opinion: split the German figures by the language of the query and look at the German-language portion on its own. An empty cell has two possible explanations — nobody is asking, or nobody has anywhere to arrive. Publishing a single page distinguishes the two within a quarter, and in this trade the second explanation is the more common one.
Which German pages should come first?
Not the company presentation. The pages that get read by a German purchasing or quality function: named authorisations with their scope, the lanes actually served, liability arrangement, and the procedural explanations around customs and dangerous goods. All of that already exists internally in tender responses, which makes it an editing task rather than a writing one — and the review is quicker because the text has been approved once already.
Our figures swing wildly. Are we doing something wrong?
Probably not — you are reading small numbers at high frequency in a market driven by external events. At thirty clicks a month a single visitor moves the percentage substantially, and a rate movement or a port disruption moves it further. Lengthen the window: quarters rather than months, year-on-year rather than month-on-month, with weekly attention reserved for entries and exits alone.
A freight exchange holds every term we care about. What now?
Check the count of shared terms and the authority gap first. A wide gap means the term will not come within reach inside any timeframe you are planning against, and pretending otherwise wastes a year. Put the money where no incumbent sits. It is also worth asking whether the exchange competes with you at all: appearing on it tends to generate qualified enquiries within weeks, whereas a page built to displace it may never arrive.
What happens when our marketing person leaves?
Whatever was written down survives; the rest does not. Three things are worth maintaining deliberately: a record of which terms were chosen and on what grounds, an enquiry log with recorded origin, and access issued to individual addresses rather than a shared password. All three cost very little while somebody is present and cannot be reconstructed once they are not.