Local SEO vs SEO: what changed, and why rankings slip

Type “plumber vancouver” into Google and look carefully at what comes back. Near the top there is a map with three businesses pinned to it. Below that, a list of blue links. Most owners read that as one ranking. It is not. Those are two separate systems, ranked on different signals with different weights, and you can win one while losing the other completely.

That distinction explains a lot of confusing situations. It is why a business can say “we are number one on Google” and still have a quiet phone. It is why the fix that worked for your competitor did nothing for you. And it is why some of the largest content operations on the internet lost most of their traffic over the last two years while their rankings, on paper, looked fine for a while.

This is a long piece because the honest version of this story does not fit in a listicle. We have used primary sources throughout: Google’s own documentation, published research from Pew, large-scale click studies, the annual survey of local search practitioners, real company disclosures, and our own Vancouver ad data. Everything is linked at the bottom.

Part one: two rulebooks on one screen

Local SEO decides the map. That box of three businesses, called the Local Pack, is ranked mostly on your Google Business Profile rather than your website. Google names three factors publicly: relevance (how well your profile matches the search), distance (how far you are from the searcher), and prominence (how well known your business is).

Organic SEO decides the blue links. That list is ranked on your website: your content, the sites that reference you, whether Google can crawl and render the page at all.

Google’s three-word summary is true but thin, so practitioners measure the rest. Whitespark runs the long-standing survey on this, and the 2026 edition (published November 2025) asked 47 local search specialists to weight 187 individual factors. The interesting part is not any single number, it is the gap between the two columns:

• For the map box: Google Business Profile signals about 28%, reviews 18%, behavioural signals 16%, on-page 14%, links 12%, citations 8%.
• For local organic results: links about 24%, on-page 20%, Business Profile 18%, reviews 14%, behavioural 12%, citations 8%.

Read those two lines against each other and the practical lesson falls out. Links to your website are worth roughly double for the blue links compared to the map box. Your Business Profile is the single biggest input for the map and only the third biggest for organic. The top individual factor for the map box is your primary Business Profile category, followed immediately by proximity. The top factor for local organic is something completely different: having a dedicated page for each service you offer.

Two systems. Same screen. One rewards a well-built profile close to the searcher, the other rewards a well-built website that other sites reference.

So when an agency tells you “we will do your SEO,” the reasonable next question is which of those two they mean, because the work barely overlaps.

This matters more in Greater Vancouver than in most places, because the region is a chain of distinct municipalities rather than one blob. Someone searching from Surrey and someone searching from North Vancouver are shown substantially different map results for the same words. If you serve the whole Lower Mainland from one address, you are competing in each of those markets from a fixed point, and only relevance and prominence are yours to move.

Part two: what actually happened to search

Owners keep telling us the same sentence: “we did not change anything and our traffic went down.” Usually that is literally true. You did not change. The rules did, and the biggest content operations on the internet got hit first and hardest. Their stories are worth knowing because the same forces reach down to a five-person business in Burnaby.

HubSpot: the textbook case

HubSpot is a large, legitimate marketing software company with a blog that was, for years, held up as the model everyone should copy. Then it fell apart in public. Depending on whose measurement you use, the blog lost roughly 81% of its organic traffic between early 2024 and early 2025. Semrush measured the blog at 14.8 million monthly visits in January 2024 and 2.8 million a year later. Ahrefs measured a similar collapse on its own data.

The independent consultant Aleyda Solis dug into the details and found something more useful than the headline. HubSpot’s blog accounted for 77% of the company’s organic traffic in January 2024 and 42% by December. The decline was not one bad day: it started with the March 2024 core update and accelerated through the November and December 2024 updates.

The cause was not a technical fault or a penalty. HubSpot had been ranking for enormous, popular queries that had nothing to do with marketing software. Pages about famous quotes. How to type the shrug emoji. Resignation letter templates. Interview weaknesses. Google reassessed whether a CRM company was really the best answer for those questions and decided it was not, handing the rankings to sites with a more obvious claim to the topic.

Chegg: when the answer replaces the click

Chegg is an education company whose business depended on people finding its answers through search. In February 2025 it sued Google, alleging that AI Overviews were taking the traffic its business ran on. The numbers in its own disclosures are stark: fourth-quarter 2024 revenue of about US$143 million, down 24% year over year, and non-subscriber traffic down 49% by January 2025. The company retained advisors to explore a sale.

You do not have to take a side in that lawsuit to take the point. Chegg was not penalized. It did not stop ranking. The results page simply started answering the question before anyone needed to click.

Forbes Advisor and friends: the rules moved under them

A third group lost rankings for a reason that did not exist a few years ago. Google introduced a “site reputation abuse” policy aimed at big domains renting out their authority to third-party content, then began issuing manual penalties. Forbes Advisor had entire sections deindexed. CNN Underscored, WSJ Buyside, USA Today and LA Times were among the publishers hit. These were not obscure spam sites. They were household names, doing something that had been normal and profitable, until it was against the rules.

And someone won: Reddit

None of this traffic evaporated. It moved. Reddit’s search visibility rose roughly 191% during 2024 (SISTRIX data reported by Search Engine Journal), following a content partnership with Google announced in February 2024. If you have noticed Reddit threads sitting above real businesses for questions in your industry, that is not your imagination.

There is a broader pattern in that redistribution worth naming: the very largest sites held roughly flat, and the losses landed hardest on the middle. That is cold comfort for a small business, but it does mean the story is not simply “everyone lost.”

Part three: the click math, and why local is different

The change that touches every business is the AI summary at the top of the results. This is measured from several independent directions, and the measurements agree on direction while differing on size.

• Pew Research Center tracked the actual browsing of 900 US adults in March 2025. When an AI summary was present, people clicked a traditional result in 8% of visits, against 15% without one. Links inside the summary were clicked in 1% of visits. People also ended their session entirely more often: 26% versus 16%.
• Ahrefs compared 300,000 keywords, half with AI Overviews and half without, using aggregated Search Console data from December 2023 against December 2025. It found the presence of an AI Overview correlated with a 58% lower clickthrough rate for the top-ranking page. Their earlier study, eight months prior, had measured 34.5%. The effect is getting worse, not settling.

Sit with what that means. Your position can hold perfectly still while the clicks that position used to deliver fall by half or more. Every rank-tracking report will look fine. This is the single most common thing we get called about now, and it is usually misdiagnosed as a ranking problem, which leads to spending money on the wrong fix.

The part that is genuinely good news for local business

Here is where most coverage of this topic misleads local businesses, because it generalizes from publisher data. Whitespark ran a study in May 2025 across 540 queries in three cities and six industries (plumbers, lawyers, dentists, optometrists, clinics, real estate agents), sorting queries by intent. The split is dramatic:

• Pure local intent (“plumber near me”): AI Overviews appeared 15% of the time. The Local Pack appeared 93% of the time.
• Informational queries (“how long does an eye exam take”): AI Overviews 92%, Local Pack 6%.
• Hybrid queries: AI Overviews 97%, Local Pack 17%.

So the map box is not being eaten by AI. When someone in Vancouver is ready to hire and searches like it, they still get a map with three businesses on it, and that placement is still the most valuable real estate in local search.

What you are losing is the top of the funnel. Every “how much does X cost,” “how long does X take,” “do I need X” query that used to bring a stranger onto your website now gets answered on the results page. That is where the informational articles most local businesses were told to write have quietly stopped paying. The correct response is not to write more of them. It is to make sure you are the business the summary cites, and to be unmissable at the moment intent turns local.

Part four: why rankings actually drop

Now the diagnosis. Google publishes its own list of causes for a search traffic drop, and after a few hundred audits we can confirm it is genuinely the right list. Six causes:

• Algorithmic updates. Google reassessed your content against everyone else’s. Nothing is broken. This is what happened to HubSpot.
• Technical issues. Google cannot crawl, index or serve your pages: server outages, a robots.txt blocking the wrong thing, pages returning errors. Boring, common, and usually the fastest thing to fix.
• Security issues. Malware or a hacked site, so Google warns people before they reach you.
• Spam policy violations. Something breaks Google’s rules, so pages rank lower or vanish. This is what hit Forbes Advisor.
• Seasonality and changing interest. Demand for the search itself moved. Roofing enquiries in Vancouver do not behave in February the way they do in September.
• Site moves and migrations. URLs changed and Google is still recrawling. This turns a redesign into a six-month problem when redirects are handled carelessly.

The genuinely useful trick, which Google also documents, is that the shape of the drop tells you which cause you have. A sudden cliff points to an algorithmic update or a site-wide security or spam problem. A slow steady slide points to technical problems or fading interest in the query itself. A dip that recurs at the same time every year is seasonal, and the correct response to it is nothing at all.

Add one modern cause Google’s list does not cover, because it is not a ranking drop at all: impressions holding while clicks fall. That is the AI summary signature. If your Search Console shows steady impressions and falling clicks, no amount of on-page work will fix it, because your ranking never moved.

Half the “Google penalties” we get called about are a redesign that dropped its redirects. A good number of the rest are February.

The local-only ways to lose

• Keyword stuffing in your business name. This one deserves honesty rather than a slogan. In the 2026 practitioner survey, “keywords in the business title” ranks as the third most influential factor for the map box, so specialists really do observe that it works. Google’s guidelines are equally clear that your profile name must be your real-world name, and that padding it “could result in the suspension of your Business Profile.” So the accurate statement is: it is effective, against the rules, and one competitor report away from taking your entire map presence offline. We do not do it for clients, and you should know the trade-off you are being offered when someone suggests it.
• Review starvation. BrightLocal’s 2026 survey of 1,002 US consumers found 74% only consider reviews from the last three months relevant, 32% want the last two weeks, and 18% the last week. A profile with forty five-star reviews whose newest is from 2023 now reads as a business that might not be open. In the same survey 97% read reviews and 68% will not use a business under four stars.
• Inconsistent contact details. Name, address and phone need to match everywhere. “Suite 200” on your site and “Ste 200” on a directory is the kind of small mismatch that erodes confidence in your data.
• A profile that stopped moving. Behavioural signals are about 16% of the map-box weighting in the 2026 survey. Hours never updated, no recent photos, customer questions unanswered for a year. An inactive profile looks like an inactive business.

Part five: what this is worth in Vancouver

Abstract advice is easy, so here is the local reality in numbers. These are Google Keyword Planner search volumes and top-of-page bid estimates with City of Vancouver targeting, pulled from our own Google Ads account in July 2026. Bid estimates are what advertisers pay for a top ad slot, which is the clearest available proxy for how contested a search is here.

• plumber vancouver: 1,900 searches a month, CA$12.87 to CA$67.17 a click.
• family lawyer vancouver: 880 a month, CA$5.78 to CA$17.57.
• electrician vancouver: 720 a month, CA$7.01 to CA$40.40.
• heat pump installation vancouver: only 50 searches a month, but CA$29.54 to CA$103.34 a click.
• restaurants vancouver: 14,800 a month, and CA$0.43 to CA$1.81 a click.

Two things fall out. First, volume and value are unrelated. “Restaurants vancouver” is searched roughly eight times more often than “plumber vancouver” and is worth a fraction as much per click. Second, and more useful: those bid ranges are the price of skipping the work. If you are not in the map box for heat pump installation, the alternative route to that customer can cost north of a hundred dollars a click in this city. Our free lead calculator runs that math by industry and budget.

Part six: the order to work through this

If your rankings or your calls have fallen, resist the urge to start rewriting pages. Diagnose first, cheapest and most diagnostic first.

• Look at the shape, not the number. Pull sixteen months in Search Console. Cliff, slide, or annual dip. That single observation rules out most of the six causes.
• Compare impressions to clicks. If impressions held and clicks fell, you have an AI summary problem, not a ranking problem, and the fix is different.
• Separate map from links. Did you lose the Local Pack, the organic listings, or both? Different systems, different fixes, and treating one as the other wastes months.
• Check the boring technical things. Is the site reachable, are pages indexed, did a redesign change URLs without redirects.
• Audit your topical honesty. What share of your traffic comes from content unrelated to what you sell? That was HubSpot’s whole problem, and it is worth knowing your own number.
• Then the profile. Review recency, name compliance, categories, hours, photos, unanswered questions.

If you would rather someone did that properly, our free 7-day audit covers exactly this ground and you keep the document whether or not you hire us. If the map box is your problem, local search and AI visibility is the service page. If the blue links and paid channels are where you are losing, digital growth is the one. And if you want to know what the AI engines currently say about your business, the AI Overview checker is free and takes a minute.

The honest bottom line

Local SEO and organic SEO are not two names for one job. One is mostly your Business Profile and your proximity to the searcher. The other is mostly your website and who links to it. The practitioner data shows the same signals weighted very differently across the two, which is why work that lifts one can do nothing for the other.

What changed since 2023 is not that ranking became impossible. It is that search stopped forgiving three things: content outside your genuine expertise, profiles left to go stale, and the assumption that a ranking equals traffic. HubSpot, Chegg and Forbes Advisor learned those lessons at a scale that made the news. The same forces are working on a much smaller scale in Vancouver right now.

The good news, and it is real: for the searches where someone is actually ready to hire you, the map box still shows up 93% of the time, and it is still won with unglamorous work that most of your competitors are not doing.

Sources

Google documentation

• Google: how to improve your local ranking (relevance, distance, prominence, and no paid placement)
• Google: guidelines for representing your business, including the business-name rule and suspension warning
• Google Search Central: debugging drops in Search traffic (the six causes and the chart shapes)
• Google Search Central: updating the site reputation abuse policy (November 2024)
• Google Search Status Dashboard: March 2026 core update (March 27 to April 8)
• Google Search Status Dashboard: May 2026 core update (May 21 to June 2)

Independent research and studies

• Pew Research Center (July 2025): Google users are less likely to click links when an AI summary appears (900 tracked US adults)
• Ahrefs (2026 update): AI Overviews reduce clicks by 58%, across 300,000 keywords
• Ahrefs (April 2025): the earlier 34.5% measurement, for comparison
• Whitespark (May 2025): prevalence of AI Overviews in local search, 540 queries across three cities and six industries
• Whitespark 2026 Local Search Ranking Factors: 47 practitioners weighting 187 factors, published November 2025
• BrightLocal Local Consumer Review Survey 2026 (1,002 US adults, published February 2026)

The company cases

• Search Engine Land: HubSpot’s SEO collapse, what went wrong and why (note: Search Engine Land is owned by Semrush, whose data it cites)
• Aleyda Solis: independent analysis of the HubSpot blog ranking drop, with the traffic-share breakdown
• Search Engine Land: Chegg sues Google over AI Overviews (February 2025)
• EdTech Innovation Hub: Chegg’s 24% revenue decline and the traffic figures from its own earnings disclosure
• Search Engine Land: enforcement of the site reputation abuse policy, with publishers delisted
• Search Engine Journal: SISTRIX data on Reddit leading search visibility growth

Our own data

• Google Ads Keyword Planner: search volumes and top-of-page bid estimates, City of Vancouver targeting, July 2025 to June 2026, pulled July 2026 from our own account. The full table is published on our lead calculator.