A higher conversion rate can leave the practice worse off
Rate treats every signed file as identical, and in immigration work they are nothing alike.
Conversion rate rises fastest when a firm stops declining. Say yes more often and the number on the dashboard improves while the paralegal hours, the deadline exposure and the collection problems all get worse.
Track signed files by case type, by fee collected, and by a rough estimate of staff hours. Once those three sit side by side, most firms find one or two case types that quietly subsidize everything else and one that consumes a year of somebody's attention.
Set the goal as mix, not volume: more of the work you do well, fewer of the files that arrive underfunded and stay open. Sometimes the improvement is declining faster and more politely.
Declining well is its own conversion skill. A same day referral out, with a name attached, produces goodwill and returning referrals in a way a slow no never does.
Let people book at midnight without talking to anyone
Self serve scheduling with two screening questions in front of it turns the quiet hours into consultations.
A large share of immigration research happens after work and after the children are asleep. Most firms answer that traffic with a form and a promise, then call back the next afternoon, by which time the person has spoken to two other offices.
Put real scheduling on the site, gated by a short screen: case type, where the person is, language, and whether they want video or the office. Only offer slots you can actually staff, including interpreter coverage.
If you charge for consultations, take payment or a deposit at booking. If you do not, confirm twice and send what to bring, because a free slot that nobody paid for is a slot somebody skips.
Keep the human path visible for people who cannot or will not use a booking tool. The goal is to add a door, not to close the one that already works.
Your answering service is deciding which cases you take
The vendor picking up after five is doing intake work that nobody in the firm has ever audited.
Pull a month of recordings and listen to twenty of them. Check the basics: did they capture a callback number, spell the name correctly, offer the Spanish path, and get the message to the right person before morning.
Then check the expensive thing. Count how many callers were turned away or told to call back who were describing a matter you would have taken. Every one of those is a signed file you paid to generate and gave away.
Give the service a one page decision list rather than a script that drifts toward advice. Non-lawyers must not be answering legal questions, and where the line sits is worth confirming with your own counsel before you hand anything over.
Measure the handoff itself: of the after hours messages taken last month, how many reached a calendar slot. Firms are routinely shocked by that fraction the first time they compute it.
Interpreter hours, not demand, are what cap the calendar
In a multilingual practice, bookable capacity is set by who can be in the room, not by how many people call.
Publish which languages are covered and when. A caller who learns on the phone that nobody is available this week has already been converted and then lost.
Block the calendar around real coverage and steer interpreter-dependent consultations into the windows where it exists. Video interpreting widens those windows without adding headcount, and it travels better across a metro this spread out.
When a language has no route at all, say so on the first call and refer out immediately. Booking a consultation you will have to cancel costs more than the honest no.
Which languages deserve a route should come from your own intake log, not from a demographic table. The log is the only source that knows who is actually calling your office.
The follow-up sequence is the only surface with enough volume to test
One metro will never power a split test on a headline, but every single inquiry passes through your follow up.
A page that receives a few hundred visits a month cannot resolve the small differences a headline test is looking for. Waiting for statistical significance in a single market is how a year disappears.
Every inquiry, by contrast, receives the same sequence of messages. Change one element per month: the timing of the second contact, the channel it uses, whether it names the fee, whether it includes the document checklist.
Compare cohorts month over month and write down what changed and when. It is not a clean experiment and you should not pretend otherwise, but a written log of changes and outcomes beats an untested sequence that has been running since 2019.
Do the same with the phone. Twenty recorded intake calls scored against a simple rubric will show you more than a quarter of page experiments.
The week a case is approved is the only time anyone will write a review
Reputation is a conversion input for every future inquiry, and it has exactly one natural moment to collect.
Build the ask into the closing checklist so it happens on approval week rather than whenever somebody remembers. Waiting three months turns a grateful client into an unanswered email.
Expect many clients to decline being named, and treat that as normal rather than as a failure. A short review with no details still helps, and nothing about the matter should ever appear in it. What you may publish, and how you may ask, sits inside the advertising and confidentiality rules that apply to you and is worth confirming with your own counsel.
Referrals travel the same way and faster. Families in Westminster, Arvada, Golden and Broomfield talk to each other, and a client who leaves with a card and a clear description of what you handle will send the next one.
Track referred inquiries as their own source. They convert differently from paid clicks, they need a different follow up, and averaging them together hides both.
Questions we actually get
- What should we measure first?
- Three things, in order: the minutes between an inquiry arriving and a human speaking to that person, the share of booked consultations that are actually attended, and signed matters by case type. Most firms cannot produce any of the three on request, and each one is usually worth more than a redesign.
- Is a paid consultation better for conversion?
- It converts fewer inquiries and generally produces better ones, since paying filters out the browsing. Whether it suits you depends on your case mix and how full the calendar is. If you do charge, the booking and payment flow has to be effortless, because friction at the payment step will cost you people who were ready to hire you.
- How do we reduce no-shows?
- Take a deposit or a fee at booking, confirm twice including the day before, send what to bring and where to park, and give every appointment a video option that the client can switch to without calling. A meaningful share of no-shows are people whose day fell apart, not people who changed their mind.
- Should we run A/B tests on our website?
- Generally not on page elements, because a single metro rarely supplies the traffic to resolve the differences you would be testing. Put that effort into the follow up sequence and the intake call, where every inquiry gives you a data point. Change one thing at a time and keep a written record of when it changed.
- What is a reasonable conversion rate for immigration inquiries?
- Any benchmark quoted at you is close to meaningless, because it depends entirely on case mix, whether consultations are paid, and how the inquiries were generated. Compare yourself against your own numbers from last quarter with case type attached. That comparison is the only one that tells you whether the work is paying.
