Almost no firm can state how many inquiries became attended consultations
Before anything is tested, the stage counts have to exist, and in most offices they do not.
The real funnel has more steps than a form fill and a signature. Inquiry received, contact made, consultation booked, consultation attended, fee quoted, matter signed, first payment collected.
Each step leaks at a different rate. A firm losing a third between booked and attended has a scheduling problem. A firm losing them between quoted and signed has a fee conversation problem. Fixing the wrong one is a wasted quarter.
Counting is a discipline problem rather than a software problem. A shared sheet filled in daily by whoever answers the phone beats a case management system nobody updates.
Run it for six weeks before changing anything. The leak is almost always somewhere the partners did not suspect, and six weeks of honest counting is cheaper than any redesign.
The first callback is a staffing decision, not a script problem
Speed to a live human is the highest-value variable in intake, and it is solved with a rota.
Inquiries arrive when people are free, which for working clients means early morning, lunch and after nine at night. Offices staffed nine to five then wonder why form fills go quiet.
Assign responsibility for the queue by hour and write it down. Cover lunch. Cover the hour after the office closes, which is frequently the heaviest hour for submissions.
When an hour cannot be covered, say so on the site and set an expectation you will meet. A stated two hour callback that happens outperforms an implied instant response that does not.
Voicemail is a leak and a five-option phone tree is a bigger one. Both are worth removing before anybody argues about a headline.
Give the consultation a fixed shape and send the scope home in writing
Attorneys who are excellent in the room are often inconsistent in how they close, and consistency is the cheap fix.
One consultation ends with a defined scope and a fee. The next ends with think it over and let us know. The difference shows up directly in the signing rate and has nothing to do with legal skill.
Use the same structure every time: what happened, what the options are, what you would do, what it costs, and the next step with a date attached.
Send a one page written scope the same day, in plain language, in the client's language where you can. People deciding with a spouse or a parent cannot forward a conversation, and the family member who was not in the room is often the one who approves the spend.
No lawyer should predict an outcome, and clients understand that. What they cannot work with is vagueness about price and process, which is a different problem and entirely within your control.
An ice week empties the calendar unless every booking has a video fallback
Winters here are real, and attendance is the stage most exposed to anything that interrupts a drive.
Set the video option at the time of booking rather than on the morning of the storm. Every confirmation should carry a link and a line saying the meeting can move online if travel is bad.
Weather is only the visible version of the problem. Shift work, childcare and a drive in from Murfreesboro or Hendersonville cancel far more consultations across a year than ice ever does.
Reschedule rather than cancel. When somebody says they cannot make it, the next sentence from your staff should offer two specific times, not an invitation to call back when convenient.
Track no-shows by day, hour and county. The pattern usually names its own fix, and it is frequently a time slot nobody thought to question.
The phone log is your largest sample, so run the experiments there
A single metro cannot power a page test for a small effect, but a busy phone produces readable evidence weekly.
A firm may take a few hundred inquiries in a quarter. That is far too few to detect anything but a very large difference on a landing page, and most tests get stopped early on noise that looks like a result.
Calls are different because you can listen. Twenty calls a week is a sample a human can read, and the changes it suggests are substantial: an opening question about what happened rather than how can we help, a question about the date on the notice, an offer of two specific times instead of an open invitation.
Change one thing in the call handling, hold it for a month, and compare consultations attended against the prior month. Crude, honest, and more informative than an underpowered split test.
Where you do test on the site, test structural things: fee published or not, long form or short, calendar booking or callback request. Small copy variations are not worth the wait at this volume.
Price the leak in signed matters before you spend a week on it
Rank the fixes by the work they would recover, which is arithmetic anyone can do on a napkin.
Take the stage counts, choose a leak, and estimate what closing half of it would produce in signed matters over a year. Multiply by your average matter value. Compare against the staff hours the fix costs.
Answering the phone in the hour after closing usually wins that comparison outright. Redesigning the home page usually does not, which is inconvenient because the home page is the thing everybody has an opinion about.
Revisit the numbers quarterly, because the biggest leak moves. Fixing callback speed routinely exposes a booking problem that was hidden underneath it the whole time.
Keep one figure visible to the entire office: consultations attended this week. Teams move what they can see, and it is the one number that sits between marketing spend and billable work.
Questions we actually get
- What should we fix first?
- Count for six weeks before deciding, because the answer differs by firm. In practice the most common winner is speed to first human contact and coverage of the hours when inquiries actually arrive, which costs staffing changes rather than money. Chasing a page redesign before the stage counts exist is how firms spend a quarter and learn nothing.
- Should our consultation be free or paid?
- Both work, for different practices. A fee filters and improves attendance, and everyone who arrives has already transacted with you once. Free fills the calendar and shifts filtering onto your staff, which is fine if you have the capacity and a disciplined way to end conversations that are not matters. What does not work is a free consultation the firm resents.
- Can we run A/B tests on our website?
- Not usefully for small effects at the volume a single metro produces. Tests get stopped early on noise and the firm acts on a result that was never there. Run structural changes one at a time, hold each for a month, and compare against the prior period. Save real split testing for large differences.
- What is a normal no-show rate for immigration consultations?
- There is no honest benchmark anyone can hand you, and published figures for this practice area are worth treating with suspicion. Measure your own baseline over a couple of months, then work on it with confirmations by text and email, a video fallback offered at booking, and specific reschedule times rather than open invitations.
- Do we need a CRM to do any of this?
- No. Start with a shared sheet updated daily by the person answering the phone. Every firm that buys software before establishing the counting habit ends up with an expensive system containing partial data. Once the habit exists, software makes it faster, and the requirements will be obvious by then.
