Your answering service is making case selection decisions and nobody trained it
Whoever picks up when the office does not is deciding, in real time, which matters reach a lawyer.
Listen to a week of after hours calls. You will hear callers asked questions your staff would never ask, told things about your practice that are not accurate, and occasionally told you cannot help with something you handle daily.
Give the service a one page brief: the case types you take, the case types you do not, the states you are admitted in, how to describe the consultation, and the exact wording of what happens next. Include the boundary too, since a non lawyer taking messages must not stray into anything resembling advice.
The Illinois question belongs in that brief. Callers from Belleville and Edwardsville will ask whether you can help someone in another state, and the answer, hedged properly, is not obvious to a scheduler reading from a generic script.
Then check monthly. Scripts drift, staff turn over, and the recording is the only place you will see it happen.
Put a different number on the Illinois page or you will never see the leak
You cannot fix a funnel you cannot see, and phone calls are where most immigration inquiries live.
Assign tracked numbers by source and by page. One for organic, one for paid, one for the Metro East page, one for the Business Profile listing. Record the calls with the notice your counsel approves, and keep them long enough to review a full month at once.
Add a small set of outcomes that intake marks on every inquiry. Reached or not reached, in scope or out of scope, booked, attended, retained. Six words on a form, filled in consistently, will show you more than any heat map on your homepage.
Split the numbers by state. Missouri and Illinois inquiries behave differently, and averaging them produces a figure that describes neither.
Expect the first month of data to be embarrassing. The gap between what a firm believes about its intake and what the log shows is usually the largest single opportunity in the account.
Send the document checklist before the appointment, not after
A consultation where the client brought the paperwork is a different meeting than one where they did not.
Send a short list at booking, in the language the inquiry came in, of exactly what to bring or upload. Notices received, passports, prior filings, dates. Nothing exhaustive, just the items that let the lawyer say something useful in the first fifteen minutes.
Two things happen. Show rates improve because a person who spent twenty minutes gathering documents is more committed to the appointment. And the meeting converts better, because the attorney can be specific instead of hedging everything until documents arrive.
Attach the practical details in the same message: address, floor, parking, the video link as an equal option, and one tap to reschedule. For someone driving in from St. Charles County or across the county line, the logistics are half the decision to show up.
Send a reminder the day before with the same checklist repeated. Redundancy costs you nothing and recovers appointments.
Signature and first payment should not require a drive across the county line
Every extra trip between the consultation and the signed fee agreement is a chance to lose the matter.
Electronic signature, card and bank payment, and a plan option if you offer one, all available the same day as the consultation. The moment of highest intent is while the person is still in the meeting or an hour after it, and a promise to mail paperwork spends that moment.
Say the fee out loud in the consultation and put it in writing the same day. Immigration clients frequently need to discuss the money with family before committing, and they cannot do that with a number they half remember from a conversation.
Make the written scope readable by the person who was not in the room. Plain language, the work included, what is not included, what happens next and when.
If a deposit holds the file or the slot, say so plainly. Ambiguity about what money does is a common reason a warm consultation goes quiet for three weeks.
Referrals and reviews are made in the first month of the file, not at the end
The strongest source of new immigration matters is a client in the middle of a good experience, and most firms wait until the case closes to ask.
Immigration matters can run a long time, and by the end the client remembers the wait more vividly than the service. The first month is when they feel relief and momentum, which is the moment to ask for a review and to mention that you welcome referrals.
Build a routine into the file: a status message at a fixed interval even when nothing has changed. Silence reads as neglect, and neglect is where refund conversations and bad reviews originate.
Handle review requests carefully in this practice. Many clients will not want to be identified, so the ask has to be genuinely optional and the alternative has to be easy. A client willing to write two lines anonymously about the process is still worth more than a client who was never asked.
Track referrals by source the same way you track ads. A referral pipeline nobody measures is a referral pipeline nobody maintains.
Intake improves when one named person owns the number, not the agency
A metro this size will not supply the volume for meaningful split tests, so the discipline has to come from ownership rather than statistics.
An immigration practice does not generate enough inquiries per month to detect small differences between two versions of a page. Anyone claiming a statistically sound test on that volume is describing noise. Accept it and change the approach.
Fix the things that are obviously broken without testing them. Unanswered calls, missing callbacks, a form that goes to an unmonitored address, an out of scope consultation on the calendar. None of those needs an experiment. They need somebody to notice.
For the rest, change one thing, write down the date, and read the numbers over quarters rather than weeks. Immigration matters open on a long delay, so a four week read on a change is guesswork wearing a chart.
Give one person in the firm the weekly ten minutes to look at the log and say what changed. An agency can build the tracking and interpret it, but the practice that improves is the one where somebody internal actually owns the number.
Questions we actually get
- Should we use an answering service or hire someone?
- Either works if it is briefed and monitored. A service is cheaper and turns over more, so it needs a written brief and a monthly listen. An in house person costs more and understands the practice, which shows up in fewer wrongly booked consultations. What does not work is an unbriefed service nobody has ever listened to.
- What should intake ask before booking a consultation?
- Enough to know whether the matter is one you take: case type in the caller's own words, which state they live in, any deadline or notice date they are holding, preferred language, and how urgent it is. Screening questions must stay factual, since a non lawyer cannot assess a case, only route it.
- Do consultation fees help or hurt conversion?
- They do both, and the balance depends on your capacity. A fee reduces volume and raises the seriousness of who shows up, which helps when the attorney's calendar is the constraint. Free consultations fill the calendar faster and include more matters you will decline. Pick one deliberately and hold it long enough to read the results.
- What tracking do we need before starting conversion work?
- Call tracking numbers by source and page, call recording with the notice your counsel approves, a short list of intake outcome codes filled in consistently, and a link between signed matters and where the inquiry came from. Without those, any recommendation about your funnel is an opinion.
- How will we know a change actually worked?
- Honestly, not with certainty at this volume. Track a single change from a written start date, read it over a quarter alongside inquiry counts and case mix, and be willing to say the result is unclear. Beware anyone reporting a precise percentage lift from a few weeks of data in one metro.
