Five boroughs are five auctions, and your budget fits maybe two of them
One campaign aimed at New York spreads a small budget across an enormous and uneven market until it stops buying anything.
Manhattan does not price like the Bronx. Queens does not behave like Staten Island. Add Nassau, Westchester and the New Jersey side of the metro and you have a set of separate auctions with separate competitors sharing one line in your reporting.
Concentrate. Pick the two or three geographies your practice already converts in, and put the money there rather than spreading it evenly and finding out in ninety days that everything was underfunded at once.
Defend your own firm name first. It is the cheapest inventory you will ever buy, competitors and directories bid on it, and a caller who typed your name is not a prospect you should be losing at the last step.
Everything else gets added deliberately, one geography and one case type at a time, with a reason attached to each addition.
Target where your callers sleep, not where their phones sit at lunchtime in Midtown
Location settings in this city quietly buy you the wrong person, because half the metro spends the day in a borough it does not live in.
Set targeting to presence rather than presence-or-interest unless you have a specific reason. Otherwise you buy searches from people who are merely curious about New York, which is a very large group.
Commuting scrambles the picture further. Someone searching from a Midtown office may be doing it for a parent in the Bronx or a spouse in Astoria, and the ad that speaks to a family situation will do better there than the ad that speaks to a location.
Report by borough from day one, not as a later refinement. The cost per booked consultation from Brooklyn and from Westchester will diverge, and the account cannot be steered until you can see the split.
Watch the edges. Radius targeting drawn from a Chelsea office reaches into two other states long before it reaches the far end of Queens, and geography in this metro is not a clean circle.
In a city with real free legal help, the free searches are the biggest single leak
A large share of immigration search traffic wants a government form, a case status or a nonprofit clinic, and every one of those clicks bills at full price.
Build the negative keyword list before launch, not after the first invoice. Government and agency terms, form numbers on their own, case status checks, free and low cost and pro bono, clinic, volunteer, and the whole family of searches from people looking for work rather than counsel.
Employment terms are the classic trap. Keywords built around sponsorship attract job seekers in volume long before they attract anyone with hiring authority, and the two look identical in a click report.
Read the search terms report weekly for the first two months, then monthly. Broad match with automated bidding will find creative ways to spend money, and the only defense is somebody looking.
Exclusions are not a tidying task. In this auction they are the largest single lever on cost per file, usually larger than bid strategy or ad copy.
Run one campaign per case type you can staff this month and pause the others by Friday
Different case types are different products with different buyers, different urgency and different economics, and blending them produces an account that optimizes toward the cheapest and least valuable inquiry.
Split at the case type level, keep ad groups tightly themed, and send each one to a landing page about that specific situation rather than to a general practice page. Message match does more work here than it does in cheaper markets.
Do not advertise work you cannot staff. An account that generates inquiries in a category with no attorney hours behind it burns money twice, once on the click and once on the consultation you have to decline.
Say the practical things in the ad. Whether the consultation is paid, which languages are answered live, and whether the matter can be handled by video. Filtering in the ad is cheaper than filtering on the phone.
Revisit the split monthly against capacity. Categories that eat consultation hours and rarely convert to signed files should be paused without sentiment, and turned back on when the practice has room.
A daily budget in this auction is gone by mid-morning unless you defend the evening
Pacing in an expensive market is a structural decision, not a monthly reconciliation.
Left alone, a modest budget in New York exhausts itself during business hours and never sees the evening, which is when a working caller with a document in their hand finally sits down to search.
Cap budgets at the campaign level so a single high-cost case type cannot swallow the account, and check delivery patterns rather than trusting the monthly total to distribute itself sensibly.
Match spend to when the phone is answered. Buying clicks at ten at night when nobody picks up until nine the next morning is a choice, and it is a defensible one only if you have an answering service or a booking flow that genuinely works alone.
Hold bid strategies steady long enough to learn. Changing targets every week in a low-volume, high-cost account guarantees the system never gets past its learning phase and never gets past your patience either.
Cost per booked consultation is the honest number here, and it is going to look high
Cost per lead flatters everyone and tells you nothing, and in New York the honest number will look alarming next to a national benchmark it has no business being compared to.
Define the conversion as a consultation booked and attended, not a form submitted. Two firms with identical cost per lead can be running businesses that differ by a factor of several once no-shows and unqualified inquiries are removed.
Feed signed matters back into the platform as offline conversions where you can. Without that, the bidding learns to buy whatever produces form fills, which is rarely the caller who retains.
Expect lag. An immigration inquiry can take weeks to become a file, so a four week report on a New York account is a report about noise. Look at cohorts, with case type and borough attached.
When the number looks expensive, compare it to the value of a signed file rather than to a national average. Clicks here cost what they cost because the market is what it is, and discounting the argument does not change the auction.
Questions we actually get
- What is the minimum budget worth starting with in New York?
- There is no honest fixed number, and anyone quoting one has not looked at your case types. The practical test is whether the budget can sustain enough clicks in one or two case types, in two or three geographies, to produce a readable pattern over a full quarter. If it cannot, narrow the scope rather than accepting thinner coverage across everything.
- Why are our clicks so much more expensive than the benchmarks we read?
- Because the benchmarks average in markets with a fraction of the competition. New York carries national brands, aggregators and lead brokers in the same auction as local firms. The right comparison is the value of a signed file in your practice, not a national average cost per click.
- Should we run Meta or LinkedIn as well as search?
- They serve a different job. Search catches people who already know they need help. Social can reach employer-side decision makers and can retarget, but immigration-related advertising sits in restricted categories on some platforms, which limits targeting options and changes how you have to build audiences. Worth confirming the current policy on each platform before planning around it.
- Do we need a separate account for our second language?
- Separate campaigns at minimum, and often a separate budget. The auctions price differently, the search phrasing is not a translation of the English list, and the campaign should only run during hours when someone answers in that language. Mixing them into one campaign hides both.
- Should we bid on other firms' names?
- It is legal in most cases to bid on a competitor's name as a keyword while restrictions generally apply to using it in the ad text itself, and attorney advertising rules add another layer, so confirm the specifics with your own counsel. Practically, defend your own name first. Competitor campaigns tend to be expensive and low-converting compared with almost anything else you could fund.
