A tourism metro fills your keyword list with people who only want to visit
Orlando produces enormous volume around the word visa from travelers, and none of it is your client.
Broad and phrase match will happily buy you visitor visa questions, travel authorization queries, passport renewal, and parents-visiting-for-the-holidays searches. In most metros that is a rounding error. Here it is a structural drain, because the whole world is looking up how to get into this city.
Build the negative list before the keyword list. Tourism terms, application status lookups, do-it-yourself filing, lottery entries, test preparation, government fee schedules and job-seeking queries all go in on day one.
Then keep cutting. Pull the search terms report weekly for the first stretch, not monthly, because the wasted spend concentrates early and compounds.
Exact match on your qualified terms plus a disciplined negative list will always look like a smaller account than the one an agency shows you in a pitch. Smaller is the point.
Language targeting is not immigration targeting across Orange and Osceola
A large share of Spanish speakers in this metro are US citizens, so buying the language is not the same as buying the need.
Central Florida's Hispanic population is large and growing and includes a substantial Puerto Rican community. Anyone in that community was born a citizen. Targeting Spanish language interest on paid social here reaches a very different composition of people than the same setting reaches in Hialeah.
Which does not mean skip Spanish. It means expect a lower proportion of the audience to have a benefit question of their own, and write creative that speaks to the petitioner as often as to the beneficiary. Citizens file for spouses, parents and siblings constantly.
Run the Spanish and English buys as separate campaigns with separate budgets and separate landing pages, so the numbers stay legible. Mixed campaigns hide which half is working.
Immigration ads generally fall into a restricted category on some social platforms, with narrower targeting available. Build for broad audiences and let the creative do the qualifying, and confirm the current policy position before you plan around any specific option.
Property managers around ChampionsGate and Celebration never type your keywords
The best employer accounts in this metro are reached by profile and by name, not by intercepting a search.
Osceola County's short term rental corridor, running through Kissimmee, Celebration, Davenport and ChampionsGate, sits on top of a management industry with cleaning, maintenance and landscaping vendors underneath it. Owners and operators there deal with workforce questions constantly and search for a lawyer almost never.
Professional-network targeting reaches them by job title and company size. So does a plain list-based approach on social, built from names you already know. The creative is a document or a briefing, not an offer of a free consultation.
Hotels, attractions and food service operators along the I-Drive corridor near the Orange County Convention Center behave the same way. The buyer is a general manager or an HR lead with a compliance worry, and the ask that works is a fifteen minute call about their situation.
Give this its own budget line and its own patience. Employer-side buys produce fewer, slower, larger matters, and judging them against a consumer campaign's weekly lead count will get them canceled before they mature.
Retargeting follows people onto shared phones, so write the ad for the room
Someone researching a deportation case may not want your firm's name appearing on a device the household shares.
Immigration clients frequently research on a family phone, a work device, or a relative's tablet. An aggressive retargeting sequence naming removal defense can surface at the wrong moment and cost you the file entirely.
Keep retargeting creative neutral. The firm name, a general practice description and a booking prompt are enough. Save the case-specific language for the landing page, where the person arrived deliberately.
Cap the frequency lower than you would for a consumer service. There is no version of this where being seen twelve times a week helps.
Exclude converters properly, including people who called rather than submitted a form. Call conversions are a large share of this category, and an account that only knows about form fills keeps paying to chase clients you already signed.
Hold a reserve back, because immigration demand moves with the news
Policy announcements move query volume in this category faster than any seasonal pattern you can plan around.
Pace to a weekly budget rather than a monthly one, and keep a reserve you can release inside a day. When a change lands, the searches that follow are the highest-intent traffic you will see all quarter, and a spent account cannot buy any of it.
Pace against consultation capacity too. If the calendar cannot absorb more this week, spending harder just buys unreturned calls and a worse reputation.
Dayparting is worth setting deliberately rather than by default. Buying clicks at hours when nobody can answer in the caller's language turns a paid consultation into a voicemail.
Reserve does not mean underspending all month and dumping it on the last Friday. Set the floor, set the ceiling, and hold roughly a fifth back for the weeks that actually matter.
The number that decides the account arrives weeks late, so build for the wait
Cost per lead is the metric that flatters every agency and informs nobody; cost per signed matter is the one that arrives too slowly to be convenient.
Import offline conversions. Consultation attended, matter opened and fee collected all need to travel back into the ad platform with the click identifier attached, or the bidding is optimizing toward whichever form is easiest to fill in.
Report by case category, not in aggregate. An employer-side matter and a family petition have different economics, and a blended cost per acquisition conceals a campaign that is quietly funding the other one.
Expect an attribution lag measured in weeks. Judge campaigns on rolling windows and resist restructuring an account on three weeks of data, which is the most common way a working account gets broken.
When a category genuinely does not pay, cut it and say so. Some of the cheapest clicks in immigration search are cheap for a reason.
Questions we actually get
- What should we budget to start?
- Enough to gather a decision in a reasonable window on one language and one case category, rather than a thin spread across five. We would rather run a single well-negated campaign properly than four that each produce too little data to act on. The right figure depends on your consultation capacity, which we would look at before quoting anything.
- Is Google or paid social better for immigration work?
- They do different jobs. Search catches people already in trouble and already looking. Paid social and professional networks reach people defined by who they are and where they work, which is how employer-side matters and community-based demand actually surface. A firm running only search is leaving the employer side entirely uncontested.
- Why do you refuse to report cost per lead?
- Because it rewards the wrong clicks. A campaign buying lottery and status-check searches will show a wonderful cost per lead and produce no fees. We report cost per attended consultation and cost per signed matter by category, which arrives later and tells the truth.
- Can we run ads targeting recently arrived communities directly?
- Targeting options in this category are generally restricted on major social platforms, and the available settings change. We build for broad audiences with creative that qualifies, rather than depending on a targeting option that may not exist next quarter. Current policy is worth confirming at the time of the build.
- How fast can you pause spend if our calendar fills up?
- Same day. Pacing against capacity is part of the setup rather than an emergency measure, and weekly budgets with a held reserve make it straightforward to throttle down and back up without the account relearning from scratch.
