How many licensed operators does Ontario have?

Hard Rock Bet's launch on 22 July 2026 made it the province's 48th licensed iGaming operator, across 82 active gaming sites. When Fanatics secured its licence in August the site count had reached 83. The operator count has been broadly flat through 2026: the province stood at 48 operators and 82 sites in April and again in July, despite launches in between — new entrants have largely been replacing exits rather than expanding the field.

Does a new operator launching create new search demand?

Not for category terms. Branded search for the new name rises because the launch is covered in the press, but phrases like 'online casino canada' have a fixed, and fairly small, monthly search volume that does not expand when another operator joins. A new entrant is competing for a share of existing demand, not adding to it.

Can a new brand simply outbid incumbents on category keywords?

Our own testing says no. We raised bids on nine high-demand head terms by 140% for 24 hours and the change produced one additional impression. Discovery platforms gate delivery on relevance signals — catalogue, engagement history, semantic fit — before price enters the auction. An unknown name is filtered out before the bid is read.

What should a market-entry acquisition plan budget for instead?

Branded and modified-branded demand, which is the only inventory that reliably serves an unfamiliar name, plus the slower work of building the relevance signals that unlock category inventory later. In our data the one term that produced a conversion was a geo-modified phrase, not a category head term.

Why measure first deposits rather than registrations?

Because the gap between the two is where acquisition quality actually shows up. Across 41 months of our own install data the same funnel produced a 15.3% conversion rate on paid-search traffic and 1.2% on incentivized traffic — a 12x difference invisible at the registration step.

Ontario just added its 48th operator. Who actually searches for it?

Ontario licensed its 48th iGaming operator in July 2026 and its 83rd site in August. Getting a licence is now the predictable part of market entry. The unpredictable part is demand: our own three-month paid-search test across three properties found that exact-match category keywords served 41 impressions and produced no clicks at all, and that raising bids 140% bought exactly one more impression.

How crowded is Ontario’s operator market now?

Hard Rock Bet launched in Ontario on 22 July 2026, becoming the province’s 48th licensed iGaming operator and its 82nd active site, with a catalogue of more than 3,000 titles. Three weeks later Fanatics secured its own Ontario licence, taking the site count to 83, with an operating agreement with iGaming Ontario still to be executed before it can accept a wager.

The market they are entering is large and still growing. Ontario’s private operators produced over $4.2 billion in gross gaming revenue in Year 4, and the province channelled 91.1% of surveyed online gambling activity onto regulated sites, up from 83.7% a year earlier. In Q1 2026 alone, players wagered $27.8 billion and operators recorded $1.13 billion in revenue, generating more than $226 million in provincial tax at the 20% rate.

What has not grown at the same pace is the number of operators. The province stood at 48 operators across 82 sites in April 2026 and at the same 48 across 82 in July, despite launches in the interval. New licences have largely been replacing departures rather than expanding the market, which means each entrant is competing for a share of a fixed pool of player attention rather than an expanding one.

Does search demand expand when a new brand launches?

Only for the brand’s own name. A launch generates press coverage, and press coverage generates branded queries for the new name. Category demand behaves differently: the number of people typing a generic category phrase in a given month is set by the size of the market, not by the number of competitors in it.

That distinction matters because category phrases are what most entry plans budget for. A vendor ASO dataset we reviewed scored the three highest-intent Canadian category phrases at 45, 42 and 27 on search popularity, against keyword-difficulty scores between 54 and 74. Those are small numbers on both sides of the trade: not much demand, and what exists is expensive and well defended.

Why can a new brand not simply outbid for category keywords?

This is the part our own data speaks to directly, because we ran the test on ourselves.

Over three months we ran paid search across three separate iOS properties in Canada, deliberately including exact-match commercial category terms alongside broader discovery inventory. The result was lopsided.

Match type	Impressions served	Clicks	Share of delivery
Exact-match category terms	41	0	~1%
Broad and automatic matching	~5,850	58	~99%

Roughly 99% of everything we were shown for came from loose, platform-chosen matching. The commercial terms we actually selected — the ones a market-entry plan would name in its first slide — served 41 impressions in three months and never produced a single click.

The obvious hypothesis is that we were underbidding. We tested it. We took nine high-demand head terms and raised their bids from $2.50 to $6.00, a 140% increase, holding the daily budget constant so the change reallocated spend rather than adding it. After 24 hours the nine terms had produced one additional impression between them. We reverted the change.

That is not an auction we are losing on price. It is a relevance gate: the platform decides whether a property is a plausible answer to a query before the bid is considered, using catalogue, engagement history and semantic fit. An unfamiliar name fails that test and is filtered out upstream of the auction entirely. Budget cannot reach it.

What actually served for an unknown property?

One pattern did work, and it was not a category term.

Query type	Impressions	Clicks	Conversions
Geo-modified brand-adjacent phrase	439	4	1
Provincial operator names	270	4	0
Generic category head terms	41	0	0

The single conversion we could trace to a specific keyword came from a geo-modified phrase — a province plus a category word — at a cost per install of $5.61. Specific, lower-competition, intent-bearing language served and converted. The broad category vocabulary did not serve at all.

This is the practical shape of a new-entrant demand curve: narrow, specific and modest, sitting well below the head terms that dominate planning documents.

Why measure deposits rather than registrations?

Because the difference between traffic sources only becomes visible after the registration step.

Across 41 months of our own first-party data — 168,916 app installs producing 2,033 first-time depositors — installs sourced from paid search converted to a first deposit at 15.3%, while incentivized installs converted at 1.2%. That is a 12-fold difference in quality between two channels that look identical on an install dashboard, and roughly identical again at registration.

Any market-entry model that optimises to installs, clicks or registrations will systematically over-value the cheapest and worst traffic. The metric has to sit at the deposit, because that is the first point in the funnel where the channels separate.

What does this mean for an entry plan?

Three things follow from the data above.

Budget for branded demand first. It is the only inventory that reliably serves an unfamiliar name, and after a launch it is temporarily inflated by press coverage. That window is short and worth capturing deliberately.

Treat category keywords as a later unlock, not an opening move. They are gated on relevance signals that accumulate over months — catalogue depth, engagement, semantic association. Spending against them before those signals exist converts budget into nothing at all, which is materially worse than converting it into expensive traffic.

Instrument to the deposit before spending. Given a 12x quality spread between channels that are indistinguishable earlier in the funnel, a measurement setup that stops at registration will misallocate the entire budget. This is unglamorous plumbing and it is the highest-leverage work in the first month.

The licensing timeline for an Ontario entrant is now measured in weeks and is largely procedural. Distribution is the constraint, and unlike a licence it cannot be bought on a fixed schedule.

A note on scope

The tests described here were run on media properties rather than operator brands, in Canadian markets, over a three-month window in mid-2026. The relevance-gate mechanic is a property of the discovery platform rather than of any one advertiser, so we expect it to generalise; the specific volumes will not. Anyone planning entry should re-run the bid-elasticity test on their own inventory before accepting our numbers, and should expect a licensed operator brand with a large catalogue to clear relevance thresholds faster than a media property does.