A Winnipeg carrier hauling into Saskatchewan, Ontario and North Dakota does not have a local search problem, and the load board sitting above it in the results is not a neighbour. Both facts are visible in ordinary Search Console and rank-tracking data — if you read the splits instead of the totals.
Most owner-run firms here already have the data. Search Console was verified when somebody set up the site, and the numbers arrive whether anyone opens them or not. What is missing is a way of turning four columns into a decision that survives being questioned by the person paying for it.
The gap is not a shortage of reporting. It is averages built from mixed traffic, and the assumption that the results page looks the way your own report implies. Both are expensive for a company selling one service across three provinces and two countries from a single prairie address.
Two feeds, and the questions only one of them can answer
The Semalt panel carries Search Console analytics across eight views and Google SERP rank tracking across six. They are often treated as duplicates. They are not, and the difference matters before any number is read.
Search Console reports your own history. Every impression it counts happened: someone searched, and a result belonging to your domain appeared. That makes it evidence rather than estimate. It also means the report is blind to everything that is not you. A term you have never ranked for produces no row.
Rank tracking works the other way round. It asks what the result page contains for a term, on a schedule, whether or not you appear on it. The SERP views carry a ranking score across sites, average position, keyword count, top keywords by search volume and position, best pages, and a competitors view listing the domains in your keyword space with their Domain Authority and shared keyword counts.
| Question | Search Console | SERP tracking |
|---|---|---|
| Did anyone click us for reefer freight to Fargo? | Yes, with counts | No |
| Who else holds the first page for that phrase? | No | Yes, with authority scores |
| Which of our pages did Google pick? | Yes, per URL | Partly, via best pages |
| Are we invisible for a term we never targeted? | No — silence looks like absence | Yes |
| Did impressions in Saskatchewan grow? | Yes, by country and region | No |
| Is a directory outranking us for our own service? | No | Yes, directly |
Background workers sync both feeds, so neither dashboard needs a manual refresh. Keep both open, because a plan built on one alone has a hole in the same place every time: everything happening outside your own listings.
Impressions, clicks, CTR and average position, each wrong on its own
These four columns appear on almost every screen. Read singly, each produces a confident and incorrect conclusion, and the errors are consistent enough to name.
Impressions count appearances, not interest. A page surfacing at position 31 collects impressions from people who never scroll that far. A 3PL that publishes a glossary entry for cross-docking can watch impressions double while nothing reaches the enquiry inbox. Impressions rise with reach and with irrelevance equally.
Clicks are the honest number, and the smallest. For a specialised carrier a strong month might be forty clicks on the commercial pages. At that scale a week-to-week comparison is noise: one rival's holiday, one storm on Highway 75, one customer checking your address twice.
Click-through rate is a ratio, so it improves for bad reasons. Lose two hundred junk impressions and CTR climbs while clicks stay flat. It also moves for reasons outside your control: an answer box, a map pack or sponsored listings above you change the rate without anything on the site changing. Treat CTR as a prompt, never as an outcome.
Average position is the most quoted and the least useful. It averages every impression across every device, country and query variant that touched a URL. For a supplier ranking second inside Manitoba and nineteenth in Ontario, the blended figure lands near nine, describing a situation that exists nowhere.
Read the four in order: impressions for reach, clicks for demand captured, CTR for how the listing reads, position last and always split. Reversing that order is how a review ends with everyone agreeing things look roughly fine.
Queries and pages are different reports, not the same one twice
The keyword view carries position history, click trends and device and country splits. The pages view attributes the same traffic to URLs. Teams live in the first and skip the second, where a common failure hides.
The query view shows what people asked for: that "flatbed carrier Manitoba" earns steady impressions while "freight broker Winnipeg" earns almost none, or that a food processor's traffic runs on ingredient terms rather than the co-packing service that pays.
The pages view shows which URL Google chose to answer with, and that choice is often not yours. A printing company with a services overview, a wide-format page and three posts about substrates often finds a blog post taking the commercial impressions while the service page sits unused.
Demand and phrasing
Which words customers actually use, and which service terms bring nothing. Positions here split by device and country.
- Find unserved phrasings
- Separate brand from service terms
Which URL Google picked
Where the traffic landed. Two pages splitting one topic between them is visible only here.
- Spot cannibalised topics
- Check the page you meant to rank
Query plus page together
Filter a query, then look at which URLs served it. Mismatches between intent and landing page surface immediately.
Ranking the wrong thing
A warehousing firm ranking with its careers page for a storage query has not won a ranking. It has lost a conversion.
The cross-read belongs in the routine. Take the five commercial phrases that matter — not the fifty that produce impressions — and check which page each lands on. If one arrives at a blog post or a careers page, you have found a change worth making, with evidence rather than opinion.
Your market is three provinces wide, and the country split proves it
This is where standard local-search advice fails a Winnipeg business outright. City keywords, a map listing and a page per neighbourhood suit a business whose customers are within driving distance. A carrier, a broker or a 3PL is not that business. Its customers are shippers in Regina, Thunder Bay, Brandon, Minneapolis and Grand Forks, and none of them search for a city.
The countries view is the fastest evidence available. A firm that believes it has a Winnipeg market often finds a quarter or more of its impressions come from the United States, with a further slice from Saskatchewan and northwestern Ontario. That is not noise. It is the shape of the business, in the only report that never flatters anyone.
The matching trap is the location page. "Trucking company in Portage la Prairie" and thirty more like it are cheap to generate, near-identical, and name addresses you do not have. Service and lane language performs instead: commodity, equipment type, border crossing, corridor.
| Split | What it usually reveals | Decision it changes |
|---|---|---|
| Canada vs United States | A real share of southbound demand nobody planned for | Whether US lane pages are worth writing |
| Manitoba vs other provinces | Out-of-province impressions at far weaker positions | Where the next service page points |
| Mobile vs desktop | Mobile CTR at half the desktop rate on the same query | Template work before content work |
| Brand vs non-brand queries | Most clicks arriving from people who already knew you | Whether growth is actually happening |
| Service pages vs blog | Editorial pages absorbing commercial impressions | Which page gets rewritten first |
The device split deserves one paragraph. Freight and manufacturing buying happens on a desktop in business hours; drivers, dispatch and plant staff use phones. A high mobile share with mobile CTR at half the desktop figure is usually a template problem, not a ranking problem — and template work is cheaper than content work.
When the result you want is held by a platform
Ask an owner who they compete with and you get two or three regional firms. Look at the first page for their main service terms and you find something else: a load board, a national directory, a marketplace listing. These are not commercial competitors. They compete for the click, often with a page carrying your own company name on it.
The competitors view makes this explicit. It lists the domains occupying your keyword space, their Domain Authority and how many keywords you share, which is the clearest picture of the field either feed produces. The authority gap is the informative part. A regional rival at similar authority is a target. A national platform several bands above, ranking a page assembled from thousands of listings, will not be displaced by a rewritten service page.
- Separate the two kinds of rival. Firms with comparable authority are worth competing against directly. Platforms only where the query carries clear commercial intent.
- Read the shared-keyword count. A platform sharing four hundred keywords owns the category. One sharing eight is glancing off it, and those eight are winnable.
- Concede the aggregator queries. Terms meaning "show me a list of carriers" will be answered by a list. Fighting for them costs more than the traffic returns.
- Take the specific ones. Equipment, commodity, lane, certification and compliance phrasing is too narrow for a directory template to answer well.
- Check the listing above you. If it carries your own details, it is part of your search presence whether you like it or not. Keep it accurate.
The AI Analytics views add a second reading of the same field, but they belong at a different stage of the argument.
Competitiveness score
Arranges rivals as top-tier, mid-tier and niche. Useful for seeing which band you are actually fighting in.
Generated, not measured
Competitor strengths, content gaps and expansion suggestions are model output. They direct where to look; they do not prove what happened.
Date ranges, the two-day lag, and a sales cycle measured in months
Date ranges come with presets at 28 and 90 days, and the Search Console analytics section already accounts for the two-day delay, so yesterday is never half-empty and misread as a collapse.
The comparison needs discipline. This week against last week, on a site earning forty clicks a month, produces a story every time and the story is usually wrong. Ninety days against the previous ninety is real, at the cost of noticing slowly. Grain, construction materials and food processing are seasonal here, so compare with the same window a year earlier too, or freeze-up reads as a ranking loss.
Keyword dynamics earns its place in a monthly routine. It tracks which keywords enter or leave the top three, top ten and top thirty. A term slipping from four to eleven has left the first page, which is material; a term moving from 44 to 38 has changed nothing. The keyword dynamics report marks the first and ignores the second — the opposite of what a position column encourages.
Exports run to 10,000 rows in CSV or JSON and 250 in a server-rendered PDF, with a configurable report builder and branding. For an owner who wants one page a month the PDF limit never binds. For a keyword audit before a rebuild, take the CSV.
What the campaign tiers do with the same data
Reading data is separate from acting on it, and in an owner-run company the acting is what stalls. No marketing staff, a site last touched properly in 2019, and any advice assuming weekly publishing is dead on arrival. Both campaign tiers are built around that.
AutoSEO
For a company with no marketing staff that needs analysis to become action without a meeting.
- Keywords found and prioritised automatically. The pool draws on Search Console, live SERP data and your own seed terms, so lane and commodity phrasing you had not considered appears as a candidate.
- Each candidate approved, rejected or deferred. Nothing enters a campaign silently, which matters when the owner signs off.
- Backlink building runs automatically. Placements draw on a partner network of more than 230,000 websites.
- Full analytics and on-site AI suggestions. The same eight Search Console and six SERP views, plus live AI chat bound to project data.
FullSEO
For a firm that wants a person deciding which terms and which donors, not an algorithm.
- Manual keyword selection with automatic fallback. Useful when the commercially valuable terms are narrow and an automated pool would spread too wide.
- Manual placement with a Domain Authority target. You set the bar donors have to clear.
- Human review before on-site changes go live. Nothing reaches the site until someone has read it — the usual condition on an owner-run domain.
- A team behind the automation. SEO specialists, developers and writers, which closes the gap when there is nobody internal to hand work to.
Add-ons are priced per slot: Wikipedia placements at 10 USD in blocks of 0, 1, 5 or 10, and PBN placements at 1 USD in blocks of 0, 20, 100 or 500. Prices are in US dollars as listed; at recent rates roughly 200 and 680 Canadian dollars, approximate. Stream, the assistant in the My SEO area, takes keyword and URL lists in bulk and keeps answers, reports, placements and to-dos in one chronological feed, which suits an owner who checks in fortnightly. The unified Semalt panel holds it behind one login.
Common questions
Our average position got worse but clicks went up. Which one is right?
Both, and the combination is usually good news. It normally means you started appearing for a wider set of terms at lower positions, which drags the average down while the terms you already held keep converting. Split by page and country before calling it a decline.
Should American impressions count towards our targets?
If you run the lanes, yes — tracked as their own line rather than folded into a Canadian total. For a carrier running into Minnesota and North Dakota that is real demand. For a firm serving only Manitoba, it is a reason to check whether the pages attract the wrong audience.
A directory outranks us for our own company name. What can be done?
For the exact company name, usually a lot: the homepage should win a branded query, and when it does not the cause is normally a thin homepage, missing structured information, or a title that does not lead with the name. For generic service terms held by platforms, the honest answer is often that the position is not recoverable and the effort belongs on narrower phrasing.
How often should an owner-run company look at this?
Monthly, with a 90-day window and the same four checks: country split, keyword dynamics, the five commercial phrases, and which pages served them. Weekly checking on a small site produces reactions to noise.
Why do Search Console and rank tracking show different positions?
They measure differently. Search Console averages real impressions, personalised and location-dependent results included; rank tracking runs a standardised query from a fixed setup. Both can be correct at once. Use the first for what happened and the second for what the page looks like.
Where the numbers stop and judgement starts
Data settles a narrow set of questions well: what people searched, which page answered, where they were, what device they held, which terms crossed a threshold. It settles nothing about whether a lane is worth pursuing, whether dispatch can take that shipper, or whether a first-page position on a term with thirty searches a month is worth a month of work.
For a Winnipeg company two structural readings do most of the work. The first is geographic: your market is wider than your city, and the country and region splits say how much wider. The second is competitive: some results belong to platforms, and knowing which ones saves budget for the terms still open. Write-ups sit on the blog, the implementation side under our services.
To see your own domain arranged this way, connect it in the Semalt dashboard and let a full quarter accumulate before drawing conclusions. First measurable movement typically takes four to eight weeks. The first thing worth reading is never the average position. It is the list of countries your impressions came from, and how little of it was Winnipeg.