Ruby vs. Dialkeep for a DMV solo law practice: human-first vs. always-on

The best argument for Ruby isn't its price. It's the voice on the other end.

A caller who's never used a virtual receptionist before, calling a law firm for the first time in her life, gets a real person who says "let me make sure I get this right" and means it. That's not a marketing line — that's what a live answering service is built to do, and for a lot of solo practices, it's exactly the right call. This post isn't a takedown. It's a walk through where that matters most, where a flat AI line starts to make more sense, and how to tell which side of that line your practice sits on — because the honest answer depends on your call pattern, not on which company is writing the post.

What Ruby actually is

Ruby sells live, human answering coverage, and as publicly listed in July 2026 on ruby.com's pricing page, it's priced in four tiers by monthly minutes — Ruby doesn't name them, it prices them: $250/mo for 50 minutes, $395/mo for 100 minutes, $720/mo for 200 minutes, and $1,725/mo for 500 minutes. Every tier includes what Ruby calls "24/7 live answering (including after-hours, on weekends, and holidays)" — this isn't a premium-only feature; it's on the entry plan too, and as Ruby puts it themselves, "the only difference between our plans? The number of receptionist minutes."

That's a real strength, not a footnote. A live receptionist can pick up on tone in a way a script can't fully match yet, reassure a nervous caller without being asked to, and never once face the question "wait, am I talking to a robot?" If your practice runs a steady, moderate call volume and the thing you value most is a human voice on every single call, Ruby's model does exactly what it says.

Where the minutes matter

Here's the part worth sitting with before you sign up for a tier: Ruby bills by the minute, and the minutes aren't evenly distributed across the calls a solo family-law practice actually gets. A caller asking to reschedule a document-signing is 90 seconds. A caller describing a custody emergency at 11pm — shaking, talking in circles, needing to be walked back to the actual facts twice — is not. Crisis calls run long because crisis callers need more time to get to the point, and family-law intake is disproportionately built from exactly that kind of call.

The /law page puts it plainly: after-hours is where cases are won, because a crisis caller hires whoever answers first. That's also, structurally, when calls tend to run longest — nobody in a custody emergency is watching the clock. Under a minutes-metered plan, the calls that matter most to your business are also the calls most likely to push you toward your tier's ceiling or into whatever comes next. Ruby's own pricing page doesn't say what happens past your plan's minutes — but their Help Center does: overage runs $5.40/minute on the 50-minute tier down to $4.00/minute on the 500-minute tier, billed in 60-second increments with no rollover of unused minutes (Ruby Help Center).

Dialkeep's answer to that same math is structural, not promotional: $199 a month, flat, unlimited calls, regardless of how long any individual call runs or how many come in during a busy week. A 15-minute custody-crisis call costs the same as a 90-second reschedule. That's the whole mechanism — nothing to track, nothing that changes the bill based on how bad someone's night is going.

The honest version of what Dialkeep is today

I'm building Dialkeep, so here's the disclosed pitch, and the limits with it. Today it answers the call, tells a caller plainly it's an AI if they ask (it doesn't volunteer that, but it won't dodge either), and captures the facts an intake file actually needs — case type, when it happened, how urgent, how to reach them back. It's scripted never to give legal advice, never to promise an outcome, and never to imply an attorney-client relationship formed on the call. It does not perform a conflicts check, does not book anything on a calendar, does not text a recap, and nothing the caller tells it reaches your firm by any route today — there's no bridge from the call to your inbox at all. And it hasn't taken a single live family-law call yet: zero customers, a demo line that isn't public yet, nothing to point you toward calling and judging for yourself tonight. That's the state of it, plainly, not dressed up.

Where a firm decides to touch anything that isn't purely fact-capture — anything about what a firm can or can't say to a prospective caller before representation exists — that's genuinely a question for your own bar's published rules, not something to take from a blog post. DC, Maryland, and Virginia each set their own guardrails here, and a five-minute check with your bar beats guessing.

So which one fits your practice

If your call volume is steady and moderate, and the thing you're actually buying is the reassurance of a real human voice on every call, Ruby's model is doing what it's designed to do, and the metering is a fair trade for that. If your crisis calls cluster nights and weekends the way DMV after-hours coverage tends to, and a bad month means both more calls and longer ones, a flat unmetered line removes the one variable that would otherwise scale against you on your worst week.

Neither answer is the "right" one in the abstract — it's your own call log that decides it, the same way screening the call itself is what actually decides whether a consult slot was worth booking. Pull your last few months of after-hours calls, if you have them, and look at how long the real ones ran. That number, not this post, is the one to trust.