Nobody types their status into a form on the first visit
The standard intake form for this trade asks for exactly the information a first-time visitor is least willing to give a stranger.
A dropdown asking for current immigration status, on a site somebody found ten seconds ago, is where a large share of inquiries quietly end. The hesitation is reasonable, and no amount of reassurance copy removes it entirely.
Ask less. Name, best number, whether a voicemail is safe, and one line about what happened. Everything else belongs in a conversation with a person who can explain why they are asking.
Offer a phone-only path with equal visual weight. Some people will never type anything, and some will call from a number you cannot call back, which is its own reason to answer live.
Say plainly who reads the form and what happens to it. One sentence beside the button, in the same size type as the rest of the page.
The person paying is often in Slidell and the person filing is in Kenner
Immigration fees are frequently paid by a relative, and most intake systems are built as though only one person exists.
A sibling on the Northshore paying for a case in Jefferson Parish is completely ordinary here. If the engagement letter, the payment link and the receipt all assume one name, the payment stalls while somebody works out the exception.
Build the second person in on purpose. A field for who is paying, a payment link that can be sent to someone other than the client, and a written note about who the client is and who receives updates. Confidentiality rules constrain the second part, so set that policy with your own counsel and put it in writing at signing.
Accept more than one payment method. Card, bank transfer and in-person options all appear in this work, and the method you do not take is sometimes the entire reason a file never opens.
Take the first payment during the conversation where the decision gets made. Momentum decays fast in a household deciding together.
One name owns each inquiry, or Thursday swallows it
Shared inboxes and shared phone lines are where inquiries go to be handled by nobody in particular.
Assign every inquiry to a person by name within minutes, automatically if your system can do it. Ownership is the cheapest change most firms can make and the one most often skipped.
Give the owner a visible list showing the age of each item. An inquiry with no contact attempt after a day should look wrong on a screen somebody actually looks at every morning.
Write down what counts as an attempt. A voicemail plus a text is an attempt. One unanswered ring is not, and counting it as one is how a firm convinces itself it followed up.
Set a hard rule for what happens when an owner is out. Immigration inquiries do not wait politely for a vacation to end, and the next firm on the list answers.
Your follow up cadence is regulated, so approve it once and reuse it
Follow-up messages, review requests and reactivation campaigns are lawyer communication, and Louisiana has its own rules about that.
Design the sequence once. Wording, timing, channels, and how someone stops it. Run it past your own counsel, then leave it alone until something actually changes.
There is a benefit beyond compliance. A fixed sequence can be measured. Improvised follow-up cannot be evaluated at all, because it is never the same twice and nobody can say what was tried.
Vary the cadence by case type rather than by mood. A matter with a deadline needs contact within hours. A family petition somebody has been thinking about for a year needs a lighter touch spread over months.
Stop when asked, log the stop, and make sure every person on the team can see it before they dial.
At this volume you are removing breaks, not proving lifts
Statistical significance is not available to a single-metro practice, so the standard has to be mechanism plus an honest record of what changed.
A firm in one metro will not gather enough consultations in a month to detect a small change in a button. Any tool reporting that it has is measuring noise and charging you for the confidence.
So change the things that need no proof. An unanswered phone line. A form that fails on a phone. A booking page with no evening slots. A confirmation message nobody sends. Each one is a break, not a hypothesis.
For anything genuinely uncertain, run it as a period rather than a split. Hold it long enough to cover a normal month of variation, note the start date, and count the same way before and after.
Keep a change log with dates on it. Without one, every result gets credited to whatever anybody happens to remember doing that month.
Silence during a pending case costs you the referrals that come after it
The waiting period is the largest part of a client's experience and usually the least managed part of the practice.
Clients who hear nothing for months assume nothing is happening. Some call every week, which spends staff hours. Some tell their family the firm went quiet, which costs the referral you never knew you had.
Set a fixed rhythm of updates, including the ones that say nothing has changed. Short, written, in the language the client uses with you.
Ask for a review at the point where the client is happiest, which is frequently not the end of the matter. Keep the request inside your advertising obligations and confirm the wording with your own counsel.
Referrals are the cheapest demand a practice in this market has, and they are produced by communication rather than by marketing spend.
Questions we actually get
- We get plenty of inquiries but few retainers. Where do we start?
- Count first. For one month, record every inquiry, who owned it, how long until a human made contact, whether a consultation was booked, whether it was attended, and whether a fee was paid. Most firms discover the largest gap sits between booked and attended, or between the first attempt and a second one that never happened. Fix whichever is biggest before touching the website.
- Should the consultation be free or paid?
- Both work, and they select for different clients. A fee filters out casual callers and protects attorney hours. Free fills the calendar faster and puts more of the sorting burden on your staff. The decision that matters is committing to one long enough to see what it does to attended consultations and signed files, and giving intake a script that matches.
- Do we need a CRM, or is a spreadsheet enough?
- A spreadsheet is enough to find the leak. It is not enough to hold the fix, because it relies on somebody updating it. Once you know which stage is losing files, move to a system that assigns an owner, timestamps the first contact and shows aging inquiries without anyone building a report.
- How do we measure any of this when cases take months to close?
- Use staged counting. Inquiry, contacted, consultation booked, consultation attended, fee paid, matter opened. The early stages give you a fast signal, the later ones give you the truth, and the gap between the two is where you find out which marketing channel is sending people who never sign.
- Can you run A/B tests on our site?
- Rarely with honest results at this volume. One metro will not supply enough consultations in a reasonable period for a small difference to be detectable, and tools that claim otherwise are reporting noise. Better use of the same effort is instrumenting the funnel, repairing the visible breaks, and running one deliberate change at a time with the dates written down.
