The Million-Dollar Digits

Which repeater phone number actually owns America's memory — the first jointly-scaled measurement of legal vanity numbers. Ten years of Google data, 2016–2026.

Every big TV firm in America has made the same bet: pay for a repeater number — 444-4444, 222-2222, 888-8888 — sing it a few hundred thousand times, and rent a permanent room in the public's memory. Nobody has ever measured whether it works. Or which number works best. We just did.

How to read these numbers

Every score in this report is a Share of Recall: the percentage of all searches for the phone numbers being compared that went to this one. If five numbers are on the board and one holds 33%, then when Americans typed any of these numbers into Google over the past decade, a third of the time they typed that one. It is a slice of a pie — not searches per month, and definitely not calls.

The number people dial is ten digits: an area code plus the repeater. William Mattar in Buffalo is 716-444-4444; in Rochester, 585-444-4444. Lerner & Rowe in Chicago is 708-222-2222. The area code changes market by market; the memorable seven-digit stem stays the same, and different firms license the same stem in different area codes across the country. We measured the bare stem — searches for "444 4444", "222 2222" — because that is how the jingle lives in people's heads. Nobody memorizes the area code.

That means a national Share of Recall is pooled: every firm singing that stem, in every market, plus cultural noise. State by state, the pool separates into owners — and that state cut is how everything below should be read.

The national board

Share of Recall, United States, January 2016 – June 2026, all five stems scaled against each other in a single request. Believed anchor advertisers: 222 — Lerner & Rowe (multi-state); 444 — William Mattar (New York); 777 — contested across industries; 888 — formerly Cellino & Barnes.

A third of the pie belongs to 222-2222. The most famous legal number in American history holds less than a tenth.

Before anyone frames this chart: the top of the board carries contamination, and we will name it rather than hide it. 222, 444 and 777 are also "angel numbers" — massive spirituality search topics with zero legal intent. The single highest month in the decade belongs to 222-2222, and it lands exactly where you would expect: February 2022, when the internet searched 2s for a day. Culture inflates the top of this chart. So how do you separate a jingle from a horoscope?

Geography is the lie detector

Angel-number searches are geographically flat — people search 444 for spiritual reasons at about the same rate in Boise and Brooklyn. Legal vanity searches are concentrated — they spike only where the number airs on television. That difference is a filter. Run the same jointly-scaled comparison state by state, and the advertising signal separates from the noise.

New York. 444-4444 takes the state decisively, capturing 63% of all repeater-number recall against 28% for 777 and 9% for 888-8888 — and it registered in 31 of 126 months versus 5 for 888. Nationally, 444 sits behind 222. In New York, it is king. That flip is William Mattar's thirty years of "Hurt in a car?" showing up in the data, cleanly separated from numerology.

Texas. 222-2222 wins with 49% of the state's repeater recall to 444's 37% — but the telling number is persistence: 222 registered in 75 of 126 months, 444 in 73. Two numbers under active, sustained spend in one state. This is not a Twosday artifact. Somebody is buying that recall in Texas every single month.

The finding with a price tag: recall is rented, not owned

The most valuable chart in this dataset belongs to the loser. 888-8888 was, for two decades, the most famous legal phone number in America. Cellino & Barnes built an empire on it. The firm dissolved in 2020, the television spend stopped — and in New York, the number now clears Google's reporting threshold in 5 months out of 126.

The most drilled digits in American legal advertising, and the recall evaporated within a few years of the checks stopping. The number is not an asset you own. It is a room you rent, and the rent is your media budget. Stop paying and the market forgets faster than you think.

The winner is still writing checks

Which makes the top of the board more impressive, not less. 222-2222's national lead is not a relic — the state data shows sustained, current, multi-market spend behind it, month after month. And the winning pattern is not one firm in one city: it is a multi-state footprint, the same seven digits with a different area code in every market, one jingle traveling the country and pooling into one national score. That is a fundamentally different strategy than owning one number in one town, and the board says it is working.

Why this never shows up in your search reports

One more finding from building this: Google Keyword Planner — the data source behind nearly every agency's search report — systematically buries phone-number queries. William Mattar's number shows almost nothing in Planner data even in Buffalo, his home market of three decades. Trends sees what Planner suppresses.

Which means every search-based brand measurement in legal marketing has been quietly understating TV-heavy, phone-first firms the entire time. Their demand does not route through a search box. It routes through a memorized number, straight to a dial pad. If your brand report says the TV guys are weak, check what your data source cannot see before you believe it.

What we are building next

This study measured five numbers in three geographies. The full map does not exist anywhere — not at the ad agencies, not at the number-licensing brokers: which firm holds which repeater stem in which area code, and what each holding is worth in accumulated recall. We are building it into the BrandTerritory index as the Vanity Number Registry.

Because here is what this data just proved: your jingle has a balance. Every year of TV spend deposits recall; every quiet year it drains — and 888-8888 shows how fast the account empties when the deposits stop. Firms have been paying into these numbers for decades without a statement. Now the statement exists.

If your firm holds a vanity number, three questions are now answerable with data: what share of your market's recall your number actually owns, whether it is growing or decaying, and how it stacks against every other number sung in your state. If you are considering licensing one: what the pattern you are eyeing is already worth before you pay for it.

Request your number's recall report

Methodology: all comparisons pulled via a single Google Trends request per geography so every number shares one 0–100 scale — separately-pulled Trends series are each normalized to their own peak and cannot be compared, which is why this had not been done before. Share of Recall is each stem's percentage of the combined search interest across the full 126-month window; proportions of relative interest, not search counts or call volumes. "Months observed" counts months a term cleared Google's reporting threshold. Figures as of August 2026. Phone-first advertisers generate demand search data cannot see, which understates every number on this board. Anchor-advertiser attributions reflect published firm contact numbers and are noted as believed where unconfirmed.