Run the ad schedule on your phone coverage, not on the valley's clock
In a metro with a large shift-working population, the ads should be live when a human can actually pick up.
Pull your answer rate by hour before you set a schedule. Most firms discover a block of unanswered calls at hours nobody planned for, and buying more clicks into that block is buying voicemail at auction prices.
You have three honest options. Widen staffed hours, hire an answering service whose script takes the facts and stops short of anything resembling advice, or restrict the schedule to the hours you cover. All three are defensible. Pretending the phone is answered is not.
If you use a service, script it for this trade: case type, the date printed on any document they are holding, whether anyone is detained, best hours to call back, and a promised callback window you will actually meet.
Sunday morning and early weekday mornings are real hours in this valley. Test them deliberately with a small budget before you decide they do not work.
A radius drawn in a visitor city buys clicks from people flying home Sunday
A destination metro fills a location-targeted campaign with travelers, convention attendees and out-of-state searchers who will never be your clients.
Set targeting to people who are present in your locations rather than people merely interested in them. It is a single setting and it removes a meaningful share of the waste in this particular city.
Then decide per campaign how wide you actually go. Some immigration work is served remotely without difficulty. Some depends on the client sitting in your office. Build the campaigns separately so the geography can differ, rather than compromising with one radius that serves neither.
Report by submarket, not by metro. Henderson, North Las Vegas, Summerlin, Spring Valley and Enterprise behave differently in both cost and quality, and a blended metro average hides which one is carrying the account.
Small exclusions pay for themselves. If you never take a case out of Boulder City or beyond, say so in the settings instead of paying to learn it monthly.
Tourist visa questions are the tax a destination city puts on your keyword list
The negative list here has to handle the travel questions a visitor economy generates, on top of the usual searches for the government.
Start with the searches for a government office, a form download, an appointment portal, a case status check, or free help. Those are the classic budget drains in this vertical and they arrive relentlessly.
Then add the layer this city generates: travel authorization questions, tourist visa extensions, and general entry questions from people planning a trip. If you do not sell that work, none of it belongs in your account.
Job seeker searches are the other quiet leak. Queries about employers who sponsor bring people looking for work, not employers looking for counsel. Keep the employer side in its own campaign with its own vocabulary.
Read the search terms report weekly for the first two months and monthly after that. The list is never finished, and an account nobody reads drifts toward the cheapest, least useful traffic available.
Broad match writes its own keyword list, and it is not the one you approved
Automated match types and blended campaign formats will find every immigration-adjacent question in the Mojave and bill you for the education.
Begin with exact and phrase match in tight ad groups, one per case type. It is slower and it is the only way to learn what this market actually costs you.
Automated formats can work later, once the account has a real signal to optimize against. Feeding them form fills teaches them to find people who fill in forms, which is not the same population as people who sign retainers.
Watch what automated assets write on your behalf. Headlines generated from your site can drift into claims about results or comparisons. Lawyer advertising rules generally govern what may be said, and what is permitted is worth confirming with the State Bar of Nevada or your own counsel before it runs.
Exclude your own brand from the general campaigns and run brand as its own line item so you can see what you are paying to defend.
You are bidding against firms with no office anywhere in the valley
Part of the price on the obvious keywords is set by national practices and lead sellers who will never appear locally.
That changes where you choose to fight. On the broad metro term you are competing with budgets built on volume economics. On a case type plus a real local detail, you are competing with firms that have to be true about the same things you are.
Local proof is the asset. A staffed phone at unusual hours, an office people can drive to from Henderson or Centennial Hills, staff who speak the language the caller starts in, and a named attorney who will be in the room.
Watch your own name in the auction. Other firms and lead resellers bid on it, and a brand campaign is usually cheap insurance for traffic you already earned.
If the head term costs more than your economics support, say so out loud and spend the money on three case types instead. There is no prize for appearing on the most expensive query in the vertical.
A small budget buys three case types well or twelve badly
Concentration is the only real advantage a modest immigration account has, and the number to defend is cost per signed file.
Pick the case types with the best combination of fee, staffing capacity and close rate. Fund those properly. Everything else can wait for a quarter where the numbers justify it.
Pace against consultation capacity rather than the calendar month. An account that books more consultations than your attorneys can take produces cancellations, and cancellations produce reviews you did not want.
Push the outcome back into the platform. Import the consultations that were attended and the matters that were signed, not the form fills. If offline import is more than you want to build, a dated spreadsheet keyed to call date will still beat optimizing toward leads.
Expect lag. People click, read, talk to family and retain weeks later, so any weekly judgment on this account is mostly noise. Set a review cadence long enough to contain the actual sales cycle.
Questions we actually get
- How much should we budget to start?
- Work it backward instead of picking a number. Decide the three case types you want, estimate what a signed file in each is worth to the practice, and fund enough clicks to learn a cost per consultation before you judge anything. Then scale only the case types where the math holds. Any agency quoting a required monthly figure before that conversation is guessing.
- Should we run Local Services Ads alongside search?
- They can work, but they compete for the same searcher and they carry their own verification and eligibility requirements that are worth confirming directly with the platform and with your own counsel. If you run both, tag them separately so you can see which one actually produced the retainer rather than assuming the last click did.
- Can we advertise immigration services on Meta?
- Platform policies around sensitive categories and targeting change often, and immigration advertising has drawn restrictions in the past. Confirm the current rules with the platform before building anything that depends on narrow targeting, and assume broad targeting with strong creative rather than tight audience segments.
- Do we need a separate campaign for Henderson?
- Not usually a separate campaign, but definitely separate reporting. Split the account by case type and let geography be a reporting dimension and a bid adjustment. If one submarket turns out to close at a very different rate, that is when it earns its own campaign and its own landing page.
- Our cost per lead is falling but signed files are flat. What is happening?
- Almost always the account is optimizing toward the easiest form fills it can find, which in this vertical means status checks, free help searches and people who are not eligible yet. Feed attended consultations or signed matters back as the conversion, tighten the negative list, and accept a higher cost per lead in exchange for a lower cost per file.
