A Copenhagen company's Danish pages and its English pages usually sit inside one Search Console property, and the figures arrive blended into a single line that describes neither of them. Separating the two before you read anything is the whole job.
Most advice about search analytics assumes one site, one market, one set of rivals. That assumption breaks the moment a Danish business publishes an English section. The two halves are not two views of the same audience: they face different searchers, different competitors and different economics, and the only thing they genuinely share is a hostname.
The buyer who never sees your Danish pages
Picture a furniture studio with a workshop in Valby. The Danish side of its site is read by dealers in Aarhus, by journalists putting together an autumn feature, by private customers who want to know whether the oak is oiled or lacquered. Everyone on that side already knows the brand exists.
The English side is read by somebody else entirely. An architect in Munich drawing up a specification needs a lounge chair with a documented fire rating. A contract-furnishing dealer in Rotterdam wants a Nordic supplier who can deliver forty units. A hotel group's procurement office in Edinburgh is comparing three studios it has never heard of. None of them will ever load a Danish page. They search in English, and if your English pages are not in that result set you were never in the running.
Denmark is a small place to sell into, so growth means Germany, the United Kingdom, the Netherlands and the rest of the Nordics. The English pages are the only shop window a foreign buyer, distributor, specifier or recruiter ever walks past.
One average, two businesses
Here is what goes wrong when you do not separate them. Danish-language queries are contested by a small field. The searchable phrases are narrow, the competitors are a handful of firms you could name from memory, and a decent page often lands in the top few results without much of a fight. English queries in the same subject area are contested by everyone on the continent, and a decent page lands somewhere on the second page.
Put both into one property and the average position lands in a comfortable-looking middle. That middle is not a description of anything. No searcher experiences it. It is the arithmetic midpoint of a strong domestic performance and a weak export performance, and it moves whenever the mix of traffic shifts, even when neither underlying market has changed at all.
Click-through rate distorts the same way. A Danish query with few searches and a first-place result produces a high rate; an English query with far more searches and a tenth-place result produces a very low one. Blended, they yield an unremarkable-looking number that hides one half of the site winning and the other half being invisible.
| What you are reading | Danish pages | English pages | Why the blend misleads |
|---|---|---|---|
| Competitive field | A nameable handful of Danish firms | Effectively everyone selling into the same category | The same rank means two different achievements |
| Typical search volume | Thin, long-tail, often single figures per month | Larger, but shared with many more sellers | Impression totals are dominated by the English side |
| Click-through rate | High when you hold the top of a narrow result page | Low while you sit below the fold | The mean sits where no query sits |
| Buyer intent | Mixed: retail, press, recruitment, existing customers | Concentrated: specification, sourcing, distribution | Equal weighting overvalues casual traffic |
| What a decline means | A domestic competitor improved a page | A market you do not monitor moved | The same downward line needs opposite responses |
Strong at home, invisible abroad
The domestic pages hold the top of a narrow result page while the English ones sit below the fold, and the blend reports a respectable middle.
- Growth targets depend entirely on the weaker half
- Every improvement to the strong half makes the average look better and changes nothing
No domestic buyer to speak of
A supplier whose customers are all outside the country still keeps Danish pages for recruitment and credibility, and they quietly dominate the impression count.
- The commercially important segment is the smaller one
- An unfiltered report inverts the priorities of the business
Cutting the property in two before you read anything
The split is not difficult, it is a habit you have to build. In the unified Semalt panel, Search Console data arrives across eight connected views: a dashboard over every verified site, then site overview, keywords, pages, devices, countries, traffic and keyword dynamics. Each can be narrowed, and the narrowing is where the value is.
- Filter by page path first. The pages view lets you isolate everything under the English directory. That single filter turns a blended property into two readable ones, and it is more reliable than any language guess.
- Read countries second. Germany, the Netherlands, the United Kingdom and Sweden appearing under your English pages is the export signal you are looking for. Denmark appearing there is a different story worth chasing separately.
- Use site tags as a standing filter. If you run several domains, tags act as a global filter across the whole account, so a portfolio of brands can be examined one segment at a time.
- Share the segment, not the login. Individual sites can be released to other email addresses, which means a distributor's marketing contact or an external consultant can see one property without touching the rest.
- Set the window deliberately. Twenty-eight days for operational checks, ninety for anything you intend to call a trend. Search Console publishes with a two-day delay, and the presets already account for it.
When forty impressions matter more than four thousand
Danish is a small language area, so raw volume is a poor guide even domestically. On the export side the arithmetic inverts. A phrase describing a niche technical requirement may be searched a few dozen times a month across Europe, and each of those searches is a person with a budget and a specification open in another tab.
That is why intent and specificity outrank volume here. The generative market-research views help: they run query research and classify search intent, so candidate phrases sort by what the searcher is trying to do rather than by how many of them there are. A competitiveness score with a market circle separates top-tier, mid-tier and niche competitors, which maps an export category better than a volume column ever does.
Design and furniture studios
The purchase decision is made by an architect or an interior specifier who needs documentation before a name.
- Queries centre on materials, certification and lead times
- Product pages must survive being read by someone comparing six suppliers
Architecture and engineering practices
Foreign clients and competition juries research a practice in English before any conversation happens.
- Project references carry more search weight than service descriptions
- Named typologies and standards are the phrases that get searched
Hospitality and food businesses
Bookings, press coverage and distribution enquiries arrive from people who have heard a name and want detail.
- Brand-adjacent queries dominate, so position one is the only useful position
- Seasonal swings distort any window shorter than ninety days
Biotech, cleantech and shipping suppliers
The entire customer base sits outside the country, which makes the Danish pages close to decorative.
- Every meaningful keyword lives on the English side
- Procurement vocabulary differs sharply from marketing vocabulary
One wrinkle belongs to this region specifically. The Øresund crossing means Swedish results mix into what a searcher on either side of the strait sees, so a Malmö-based rival can surface in a result set you assumed was domestic. Expect neighbours in the competitor list.
Keyword dynamics: what entered, what fell out
On thin export volumes, a trend line is close to useless. Twelve impressions becoming eight is noise; eight becoming twelve is also noise. What is not noise is a change of band. Keyword dynamics tracks precisely that: which phrases entered or left the top three, the top ten and the top thirty during the period you selected.
Band movement is legible when totals are not. A phrase crossing into the top ten has become visible to a buyer who scrolls; a phrase falling out of the top thirty has ceased to exist for that buyer. Both deserve a response even when the click count barely moved, and both vanish inside an averaged chart.
Eight views over one verified property
For anyone who needs the export half of a bilingual site read separately from the domestic half.
- Portfolio dashboard. Clicks, impressions, click-through rate, trend lines and health flags across every verified site at once.
- Keyword view with splits. Position history and click trends, broken down by device and by country, which is where an export pattern first becomes visible.
- Pages, devices, countries, traffic. Four separate cuts, each of which can be pointed at the English directory alone.
- Background synchronisation. Workers keep the data current, so a dashboard is up to date without a manual refresh before every meeting.
The Danish list and the everyone list
Rank tracking sits in a separate group of six views and answers a question Search Console cannot: not how you performed, but who you performed against. It holds a ranking score across sites, an average position, a keyword count, top keywords by search volume and current placement, the best-performing pages, and a global ranking across a domain portfolio with a twenty-eight-day trend curve.
The view that matters most for a bilingual site is the competitor one. It shows the domains contesting your keyword space, with domain authority and the count of keywords you share. Run it against the Danish side and you get a short list of familiar names. Run the competitor breakdown in the SERP module against the English side and the list changes character entirely: larger domains, older domains, and companies whose marketing budget is not comparable to yours.
| Scoreboard | Who is on it | What a shared keyword count means | Realistic ambition |
|---|---|---|---|
| Danish pages | Domestic firms, a few Nordic neighbours via Øresund | Direct overlap with a named rival you can study page by page | Top three on most commercial phrases |
| English pages | Continental and global sellers, trade publishers, marketplaces | Partial overlap with domains that outrank you on authority | Top ten on specific phrases, top thirty on broad ones |
| Domain authority gap | Small and closable | A page-quality problem | Fixable with content and internal links |
| Authority gap, English | Often large | A link and reputation problem | Fixable slowly, with placements and references |
Both scoreboards are legitimate. The mistake is judging one by the other's standard: celebrating a domestic top-three while an export phrase sits at position twenty-four, or despairing at twenty-four without noticing that the four domains above you are pan-European trade platforms.
Turning a figure into a decision
Analytics only earns its place when a number changes what somebody does on Monday. The rules below are deliberately blunt, and they are applied to the English segment and the Danish segment separately, never to the blend.
- Impressions rising, clicks flat. You are being shown and not chosen. The title and description are doing the losing, not the ranking. Rewrite them for the buyer described in the section above, not for the domestic reader.
- Position between eleven and twenty. Close enough to be worth investment. Strengthen the page itself and point internal links at it before spending anything external.
- Position four to ten with a healthy rate. Demand is proven and the page converts attention. This is where external placements pay for themselves fastest.
- Impressions at zero on a published English page. This is not a ranking problem at all. The page is not in the index, and no amount of optimisation will change a number that has no denominator.
Once the reading is settled, acting on it can be automated. Campaign automation in the panel comes at two levels, and the choice between them is mostly a question of who selects the keywords.
AutoSEO, for a two-language site with one operator
Suited to a company where nobody's job title contains the word search, and the English pages still have to compete.
- Candidates come from three sources. The keyword pool draws on Search Console, live result pages and your own seed phrases, and every candidate is approved, rejected or deferred one at a time.
- Placement runs in the background. Backlink building draws on a partner network of more than 230,000 websites, with Wikipedia slots at 10 USD each and network placements at 1 USD each as optional additions.
- On-site suggestions arrive as proposals. The step up to FullSEO at 500 USD per month adds manual keyword selection with an automatic fallback, placement against a target domain rating, and a human review mode before changes are applied.
Whatever you decide, the decision has to leave the screen. Exports run to CSV or JSON at up to 10,000 rows and to PDF at up to 250 rows, rendered server-side, with a configurable report builder carrying your own logo and colours. Country and device heatmaps carry an export story best, because they show a board or a distributor where demand physically is. Semalt's reporting and export tools also cover time series, metric cards and sortable tables.
Common questions
Should the English section live on a subdirectory or a separate domain?
A subdirectory keeps both languages under one property, so the analytics split is a filter rather than a reconciliation exercise, and the English pages inherit whatever authority the domain has earned. The cost is the problem described above: figures arrive blended, and you separate them yourself every time.
Our English pages get almost no impressions. Is the translation at fault?
Check indexing before language quality. A page that is not in the index generates no impressions regardless of how it is written, and that failure looks identical to a ranking failure in every chart. Only once the page is demonstrably indexed does the wording become the variable worth testing.
How long before a change in the English section shows up in the figures?
First measurable movement typically appears after four to eight weeks, and on thin export volumes you should expect the slower end of that. Watch band changes in keyword dynamics during the wait rather than click totals, because totals on a few dozen monthly searches will not produce a readable line in that time.
Can a distributor abroad see our figures without an account of their own?
Individual sites can be shared to other email addresses, so an external partner sees one property and nothing else. For an audience that only needs the outcome, a branded PDF at up to 250 rows is usually the better format, since it travels through procurement departments without requiring anyone to log in.
Is it worth tracking Danish keywords at all for an export-led business?
Yes, but for different reasons. Domestic search supports recruitment, press coverage, supplier relationships and the credibility a foreign buyer checks after finding you. Judge it on those terms, with its own targets, not inside an average dominated by the export side.
What to do first
Start by proving to yourself that the two halves behave differently. Filter the pages view to the English directory, note the average position and click-through rate, then invert the filter and note them again. If the two sets of figures are far apart, every blended report you have read until now has been describing a business that does not exist.
From there the sequence is straightforward: confirm the English pages are indexed, read band movement rather than totals, pull the competitor list for the English segment and look honestly at the authority gap, then decide where money goes. The first step belongs to technical SEO; the rest lives in the panel.
If you want to see the split on your own property rather than in the abstract, connect the site and run the two filters side by side: open the dashboard and connect your verified property, then compare the English directory against everything else across a ninety-day window. Whatever else the exercise produces, it will end the habit of quoting one average for two businesses.
