What an NEMT operator actually needs from a CRM
Most NEMT operators do not have a software problem. They have a memory problem that only shows up once they get busy.
You do not need enterprise anything. You need to stop losing track of people.
The failure mode, specifically
This is not abstract. It happened to us at The Doctor Ride when we got serious about marketing and the volume picked up.
We started running ads, doing walk-ins, putting the word out. It worked. The calls picked up and the lead forms picked up. And it got messy fast in four particular ways.
Two people touching the same customer. Somebody calls, one of us takes it and quotes them. They call back Thursday, the other one picks up, and now that person is being asked the same questions they already answered. Or worse, quoted a different number. Every time that happens you look like a company that does not have its act together, which is exactly the thing a family checking on their mother is watching for.
Quotes nobody wrote down. You give a price on the phone from the driver seat. Three days later they call to book. What did you tell them? You are guessing, and either you underquote and eat it or you overquote and lose the ride.
A ride confirmed on a text thread only one person can see. This is the one that actually costs you. The customer texted your cell. You said yes. It is on your phone. Nobody else knows that trip exists. You get sick, you get busy, you forget, and a person in a wheelchair is sitting by a door at 9am waiting for a van that is not coming.
Leads that just go quiet. Somebody asked for a quote three weeks ago and you never called back, because there was no list telling you to. That lead was paid for, it was interested, and it evaporated because nobody had a system for the second touch.
None of that is a character flaw. It is what happens when the system of record is a human being who is also driving.
What a CRM actually is, in plain terms
Strip the word down. A CRM is four things.
- One record per person. Every customer and every lead exists in one place, with a name, phone, mobility type, and where they go.
- One thread you can all see. Calls, texts, and emails on that person land in the same timeline with timestamps, and more than one person can open it. This is the fix for the text thread stuck on one phone.
- A stage on every lead. New, quoted, booked, lost. So you can look at a list and know who is waiting on you right now.
- Follow-up that runs without you remembering. A quote that goes quiet gets a nudge in two days because the system does it, not because you thought of it.
That is it. If a tool does those four things and you actually use it, you have solved the problem this page is about.
A CRM is not dispatch software
Be clear about this, because it is where the money gets wasted.
Dispatch software runs the day of the ride. Assigning trips to drivers and vehicles, routing, driver app, on-vehicle status, signatures, billing the trip. That is a different job, it is genuinely specialized, and if you run any volume you need it.
A CRM runs everything before the ride exists. Somebody heard about you, reached out, needed a price, went quiet, came back, and booked. All of that happens before dispatch has anything to schedule.
They are not competitors and neither replaces the other. Trying to run your sales pipeline out of a dispatch system is miserable, because dispatch systems assume the trip already exists. Trying to dispatch out of a CRM is worse.
What you want is a handoff. The lead lives in the CRM through the messy part, and when the ride is confirmed, it goes into dispatch to actually get executed. One direction. The CRM is not trying to be the system of record for the trip.
Why the private pay boundary is deliberate
One NEMT covers private pay customers only. Broker trips, Medicaid trips, and individual patient records referred to you by a facility are out of scope on purpose.
That is not a missing feature. It is a decision, and it is worth understanding why.
The moment a system starts holding patient records that came to you through a facility or a payer relationship, you and the vendor are both in a much heavier category of data agreement, with obligations that come with it. That changes what the software costs, what contracts have to be signed, and how long it takes to turn anything on.
Keeping the boundary at private pay keeps this an ordinary software subscription. You sign up, it works, nobody is negotiating a data agreement for six weeks.
To be direct about what NGM does and does not say here: we make no representations about your regulatory obligations. Those are yours, they depend on how you operate, and you should talk to your own counsel about them. What we control is the scope of what goes in the system, and we drew that line on purpose.
Practically, that means: private pay leads and customers, yes. Facility business contacts, the office manager at the dialysis center, yes. Broker and Medicaid trip data, no. Patient records handed to you by a facility, no.
Most private pay operators find that boundary is exactly where their sales pipeline lives anyway.
What to actually look for
If you are shopping, here is the short list that matters.
- Shared visibility. Can a second person open a customer and see everything that has ever been said to them? If not, skip it. This is the whole point.
- Timestamps on everything. You need to be able to answer “when did we tell her that.”
- Texting from a business number. Not your cell. In this business most customer conversation is text, and if it is on your personal phone it is invisible to your company and it leaves with you.
- A stage field you will actually maintain. Four stages you use beats fourteen you ignore.
- Forms that feed it directly. Your website request form should create the record. If somebody has to retype a lead into the CRM, that will stop happening within a month.
- Missed call handling. A rung out call should at minimum trigger a text back and create a record. See the page on missed calls for why this is the highest leverage thing on the list.
- It talks to your dispatch software. Even one direction. Retyping confirmed rides is how rides get dropped.
When you do not need this yet
Honestly, if you are one person doing a few trips a week and you have never once lost track of a customer, a notebook is fine and you should spend the money on something else.
The moment to move is when any of these is true: a second person is answering customers, you have caught yourself quoting the same person twice, or a lead went cold and you only noticed weeks later.
That is not a growth milestone. It is a warning light. Once you are busy enough for it to matter, you are also busy enough that fixing it in the middle of the mess costs a lot more than it would have last month.