Quick Answer: For map pack visibility your Google Business Profile does the heavy lifting: Whitespark's 2026 study weights profile signals at 32% of local pack ranking. For AI answers and organic results, your website leads at 24% and 33%. They are not substitutes. Fix whichever surface your own buyers actually use first.
| Signal group | Map pack | Localized organic | AI visibility |
|---|---|---|---|
| Google Business Profile signals | 32% | 7% | 12% |
| On-page signals (your website) | 15% | 33% | 24% |
| Review signals | 20% | not top five | 16% |
| Link signals | 8% | 24% | 13% |
| Behavioural signals | 9% | 10% | not top five |
| Citation signals | not top five | not top five | 13% |
Read that table across the rows rather than down the columns and the argument settles itself. Neither asset ever falls to zero. Your profile carries a third of the map pack and still contributes 12% to whether an AI assistant surfaces you. Your website carries a third of localized organic, a quarter of AI visibility, and 15% of the map pack people assume belongs to the profile alone. These are not competing products. They are two halves of one machine-readable identity, and the useful question is narrower: with the money I have this quarter, which half do I repair first?
The question is wrong, but the budget behind it is real
Owners rarely ask this out of curiosity. They ask because a web designer quoted $4,000 for a rebuild the same week somebody on a forum told them the website is dead and Maps is all that matters. Both parties have an incentive. Neither answered the question the owner was actually holding: where does a finite amount of money produce the most enquiries this year?
Treat the versus framing as a budget question and it becomes answerable. The profile is free, fast, and capped. The website costs money, compounds slowly, and has no ceiling. Those three properties decide the order of operations for almost every business we look at, and they decide it differently depending on what you sell.
What the ranking factor data actually shows
Whitespark's 2026 Local Search Ranking Factors report, published November 2025, surveyed 47 practitioners across 187 factors and split the results by surface. That split is the useful part, because most arguments on this topic collapse three different result types into one word: "Google."
The inversion is the headline. Profile signals lead the map pack at 32% and collapse to 7% in localized organic. On-page signals do the reverse, climbing from 15% to 33%. For AI visibility the panel put on-page first at 24%, profile at 12%.
The finding people skip
On-page signals appear in the top three of all three surfaces. Profile signals appear in all three too, never below 12%. No surface exists where one asset does all the work and the other none. The weights shift; the dependency does not. Source: Whitespark, 2026 Local Search Ranking Factors, November 2025.
One caution: these are practitioner estimates, not numbers extracted from Google's ranking systems. They describe expert consensus about relative influence. Treat them as a well-informed weighting rather than a formula, and be sceptical of anyone quoting them to the decimal point.
Google's own documentation says the map pack depends on the web
Set the survey aside and go to the source. Google's guidance on local ranking names three factors: relevance, distance, and prominence. Relevance is how closely your profile matches the query. Distance is how far the searcher is from you. Prominence is the interesting one, and Google's wording is specific: it is "also based on info like how many websites link to your business and how many reviews you have."
That sentence undercuts the popular claim that Maps rankings are a closed system you win from inside the profile dashboard. Links from other websites are a web signal. They accrue to a domain, get earned through content and relationships, and cannot be toggled on in an edit screen. A business with no website has no obvious destination for anyone to link to, which quietly caps its prominence ceiling.
Why the seam matters more than either half
Search systems confirm a business exists by triangulating mentions of the same entity across independent sources: your profile, your site, directories, reviews, and third-party citations. When your profile lists a service your site never mentions, or your site lists an address your profile does not, you have not given the machine two supporting witnesses. You have given it a contradiction to resolve, and contradictions resolve downward. In the 12 Vectors framework this is Vector 2, Anchor, and it is the reason we treat the two assets as one job rather than two line items.
Which one do AI assistants actually read?
Your website, predominantly, and the mechanism is worth understanding rather than taking on faith. Assistants and AI Overviews answer by retrieving passages they can quote and attribute. A profile is a structured record: hours, category, address, star rating, a short description. It holds almost no quotable prose and lives inside a Google product rather than at a URL an assistant can cite back. Your service page is exactly the shape these systems retrieve from.
Google's developer documentation on AI features is blunt, and worth quoting because so much agency copy contradicts it: "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary," and "You don't need to create new machine readable files, AI text files, or markup to appear in these features." Anybody selling you an llms.txt file or an "AI rewrite" is selling something Google has said it does not use.
The academic work points the same way. Aggarwal and colleagues at Princeton, in the paper that named generative engine optimization, tested content changes across roughly 10,000 queries and found that adding quotations, statistics, and citations to authoritative sources lifted visibility in generative engine responses by up to 40%. Every one is a page-level edit. None exists inside a profile. We go deeper on the profile side in our research note on Google Business Profile and AI visibility.
That does not make the profile irrelevant here. It still carries 12% on the Whitespark panel, and the reason is consistency: it is one of the cleanest, most-mirrored records of your name, address, phone, and category on the web. It is how a model checks the business on your site is a real, locatable operation rather than a page.
What a customer does between finding you and calling you
Here is the part no ranking table can measure. Discovery and decision are separate events on separate surfaces.
Someone searching for an emergency plumber at 11pm discovers you in the map pack and calls from it. Both events collapse into one tap, and the website barely participates. Someone pricing a $22,000 basement underpinning discovers you the same way, then does something else entirely: opens three tabs, reads your process page, checks your insurance, looks for photographs of work resembling their own house, and decides which two of the three get a call.
That research interval is where a website earns its money, and it is invisible in rankings. You never see the visitor who read your thin four-page site, learned nothing that separated you from the other two contractors, and phoned one of them. The profile got you onto the shortlist. It has no mechanism for winning it. The longer and costlier the consideration, the more of the decision moves onto pages you control.
Side by side: what each asset can and cannot do
| Capability | Google Business Profile | Your website |
|---|---|---|
| Cash cost to set up | None, it is free | Typically four figures upward |
| Time to visible effect | Days to weeks | Months to quarters |
| Appears in the map pack | Yes, its home surface | Only as a supporting signal |
| Appears in organic results | Rarely | Yes, its home surface |
| Citable by AI assistants | Weakly, little to quote | Yes, passage-level retrieval |
| Works beyond Google | No, it is a Google property | Yes, platform-neutral |
| Room to explain a complex service | Very limited | Unlimited |
| Captures leads you own | Calls and messages only | Forms, bookings, email, tracking |
| You control it | No, Google can suspend it | Yes, an asset you hold |
| Ceiling on growth | Capped by proximity and category | Effectively uncapped |
The control row deserves more attention than it gets. Profiles get suspended, sometimes in error, and reinstatement is a queue, not a phone call. Any business whose lead flow sits entirely inside a Google product carries a single point of failure it did not choose. If your profile has gone quiet, our page on why a Google Business Profile stops showing covers the usual causes.
Fix this first: a sequencing answer by business type
Generic advice to "do both" is true and useless to somebody with $600. What follows is the order we would actually recommend, by what you sell, assuming you cannot fund both at once.
| Business type | Fix first | Why | When the other becomes urgent |
|---|---|---|---|
| Walk-in retail, restaurants, salons | Profile, decisively | Proximity-driven, low consideration. Buyers decide in the map pack and often never click | When you need ordering or bookings, or once review growth flattens |
| Emergency trades: plumbing, locksmith, towing | Profile, decisively | Urgent intent converts to a call in seconds. Nobody reads an About page during a flood | For scheduled higher-ticket work, and for AI answers to "who do I call for X" |
| Considered trades: roofing, foundation repair, HVAC, renovations | Profile first, website close behind | The profile earns the shortlist, then buyers research for days before choosing | Straight after the profile basics are done. Do not wait a year |
| Professional practices: legal, accounting, dental | Together, website leading | Credentials and specialization drive the decision, and none of that fits a listing | The profile still gates map pack entry, so do not skip it |
| B2B, manufacturers, niche suppliers | Website, clearly | Buyers rarely use the map pack. Research is long and spec-driven | The profile is a legitimacy check, not a channel. Claim it and move on |
| E-commerce or fully remote, no service area | Website only | You are likely ineligible for a profile under Google's guidelines | Not applicable |
| Brand-new business, any category | Profile, always | Free, the fastest legitimacy signal, and the only asset that can produce enquiries in week one | Once verified, start the site before month three |
Two patterns run through that table. Where consideration is short and proximity matters, the profile wins the first dollar. Where consideration is long or credential-based, the website carries more of the decision than any ranking table suggests, because it does persuasion rather than discovery. Service-area businesses spanning several municipalities have an extra wrinkle, unpacked in local SEO versus national SEO.
Not sure which half is actually broken?
We will check your profile and your site against the surfaces buyers actually use: the map pack, organic results, and what ChatGPT, Perplexity, Gemini, and AI Overviews say when someone asks for a business like yours. Reply within one business day. If the answer is that you should spend nothing right now, we will tell you that.
Run the arithmetic before you spend
The sequencing table gives you a default. Your own numbers can override it, and the calculation takes four minutes. You need two figures: gross profit on your average job, and the cash cost of each fix. The numbers below are illustrative only. Substitute your own.
Worked example (illustrative figures, not client data)
- Average job gross profit: $900
- Profile fix: $0 cash, about six hours of your time in month one, then one hour monthly
- Website build: $4,000 one time
- Break-even on the profile fix: $0 ÷ $900 = zero extra jobs. Any additional enquiry is pure return
- Break-even on the website build: $4,000 ÷ $900 = 4.5 extra jobs, once
- The honest comparison: the profile needs no incremental job to pay for itself; the site needs about five, spread over its lifetime
Now flip it. If your average job carries $12,000 in gross profit, as on a structural repair or commercial installation, the $4,000 site breaks even on one additional closed job. Payback is not a fixed property of websites. It is a function of ticket size, and high-ticket businesses reach it so fast the usual sequencing advice inverts.
The profile still comes first even for those businesses, and the reason is not return on investment. It is that the return is free, and spending money before spending free effort is poor capital allocation at any ticket size. For what search work costs in Ontario, see how much SEO should cost.
When the honest answer is to spend nothing with us
This page exists partly to talk some readers out of a purchase, so here are the cases where hiring anybody, us included, is the wrong next move.
Do these yourself, free, before you pay anyone
- Your profile is unclaimed or unverified. Claim and verify it today. It is a form and a verification step. No agency should charge for it, and several do.
- Your primary category is wrong or vague. The primary category is one of the highest-impact fields on the entire profile and changing it takes ninety seconds.
- You have fewer than about ten reviews. Review signals carry 20% of map pack weight and 16% of AI visibility weight on the Whitespark panel. Asking costs nothing and outperforms most paid work at that stage.
- Your hours are wrong or your phone goes unanswered. Visibility multiplied by a broken intake produces missed calls, not revenue. Fix the operation first.
- You are a busy walk-in shop with a full calendar. You may genuinely not need a bigger website. Capacity, not visibility, is your constraint.
We publish a Google Business Profile optimization checklist so owners can work through this without paying anyone, ourselves included.
A pattern worth naming: businesses arriving convinced they need a website rebuild frequently have an unresolved profile problem sitting one screen away. Rebuilding is the more expensive fix and the slower one, and quoting it before checking the free work is how a good deal of Ontario marketing money gets spent on the wrong half.
The compounding curve, and why month three misleads
Owners most often abandon website investment because they judge it on a profile timeline. Profile work produces feedback quickly: a corrected category or a run of fresh reviews can shift map pack position inside a fortnight. Site work behaves nothing like that, and expecting it to is what turns a sound investment into a cancelled contract at month four.
An anonymized Ontario shipping container dealer we work with shows the shape. Across a 16-month window the account recorded 4,810 clicks against 627,000 impressions. The distribution is the lesson: single-digit daily clicks for roughly the first year, one sample day showing 6 clicks against 1,682 impressions, then seventy-click days against 6,000 to 9,000 daily impressions by early summer 2026. Blended average position sat at 26.3, which reads badly until you notice it mixes freshly published pages entering at the bottom with commercial pages climbing. Judged at month three it looked like a failure. Judged at month sixteen, an asset. Attribution: Google Search Console, 16-month window, shared with permission. How quickly your own curve bends depends on how contested your category is and how much ground the site has to make up.
Matt Griffin, Formative Digital: "The pattern I see over and over in audits is an owner quoted for a full site rebuild when the real bottleneck was an unverified profile with four reviews and the wrong primary category. My rule is that we spend the free hours before we spend your money, and if I cannot find something free that moves your numbers before I quote you, I have not looked hard enough. That costs us revenue on some calls. It is still the right order to work in."
How Formative Digital sequences the two
Our order of operations is consistent, and stated plainly enough that you could run it without us. Audit both surfaces to establish what the map pack, organic results, and AI assistants currently say. Repair the profile completely, because it is free and fast. Reconcile the two records so categories, services, hours, and contact details agree exactly. Then build the site against the questions buyers ask before they call, and measure all three surfaces rather than one rank number.
That sequence is what our local SEO service executes, and the engagement record sits in our case studies. Our standing commitment applies: Results Guarantee: if your existing domain shows no measurable organic search results after 12 months of work with Formative Digital, we work for free until you see them.
Frequently Asked Questions
Can I rank on Google without a website?
Yes, in the map pack. A verified profile can rank in Maps with no website attached, and for some walk-in businesses that alone keeps a calendar full. What you cannot do without a website is rank in ordinary organic results, appear in AI answers consistently, or control what a buyer reads before calling.
Does having a website improve my Google Maps ranking?
Indirectly, yes. Google's local ranking documentation says prominence is based partly on how many websites link to your business, which is a web signal rather than a profile one. Whitespark's 2026 survey puts on-page signals at 15% of local pack weight. Your site is not the main lever in Maps, but it is not neutral.
Which should I fix first if I only have a few hundred dollars?
The profile, in almost every case, because it costs nothing but time and it is where local buying decisions get made fastest. Claim it, verify it, set the correct primary category, load real photographs, and ask every satisfied customer for a review. Spend cash on the site once that free work is actually finished.
Do AI assistants like ChatGPT read my Google Business Profile?
Profile data does travel through aggregators and directories into the material these systems draw on. But assistants cite pages they can retrieve and quote, and a profile gives them very little to quote. Whitespark's 2026 panel scored on-page signals as the largest AI visibility factor at 24%, with profile signals at 12%.
Is a free Google Business Profile website enough?
That option no longer exists. Google switched off the simple websites it generated from Business Profiles on 1 March 2024, and the business.site addresses stopped resolving later that year. If you relied on one, you now have a profile with no destination behind it.
How long before a new website affects my rankings?
Plan in quarters, not weeks. A corrected primary category or a burst of fresh reviews can move map pack position within days. Site work compounds far more slowly: in one anonymized Ontario engagement the search console record showed single-digit daily clicks for roughly a year before the curve bent upward. How quickly your own curve bends depends on how contested your category is and how much ground the site has to make up.
Do I need a Google Business Profile if I am purely online?
Probably not, and you may not even qualify. Google's guidelines require a physical location customers can visit or a business that travels to them, so a purely online operation with no service area is usually ineligible. Then the website is not the first priority, it is the only one.
Sources
- Google Business Profile Help. Tips to improve your local ranking on Google. Link
- Shaw, D. (November 2025). 2026 Local Search Ranking Factors. Whitespark. Survey of 47 practitioners across 187 factors. Link
- Google Search Central. AI features and your website. Google Developers. Link
- Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2023). GEO: Generative Engine Optimization. arXiv:2311.09735, accepted to KDD 2024. Link
- Google Business Profile Help. Guidelines for representing your business on Google. Link
Get your free AI visibility audit
Formative Digital, Brantford, Ontario · (226) 450-2065
If you have read this far and still cannot tell which half is weak, that is the audit's job rather than yours. We check both against the map pack, organic results, and what the major AI assistants say, then hand you the order of operations we would follow. A recommendation to spend nothing is a legitimate outcome.