DMC software: what to actually check before you buy (a buyer’s guide for 2026)

If you run a DMC, tour operator or specialist travel team, you’ve probably evaluated operating software at some point. Maybe you’ve trialled Safari Portal. Maybe you’ve sketched something out in Airtable. Maybe you’ve sat through three demos in a week and come out less sure than when you started.

Here’s the awkward truth: most demos are designed to make the software look great in 25 minutes. They’re not designed to surface the things that matter at month three. This guide is a checklist of what to check, what to ignore, and what to ask in the quiet bit after the demo, when the real answer comes out.

We’re building FieldDesk for exactly this audience, so we’re biased. We’ve also evaluated everything in the market and rebuilt our own ops three times. This guide is what we wish someone had handed us in 2023.

1. Is it actually built for tour operators, or repurposed CRM?

The single biggest tell is the vocabulary. Open the software and look at the navigation. Does it say Pipeline · Trips · Suppliers · Agreements · Quotes · Agents? Or does it say Leads · Deals · Companies · Contacts · Tasks?

If the vocabulary is generic CRM, the workflows underneath will also be generic. You’ll spend your first six months bending fields to look like a trip. Travellers will become “contacts”. Suppliers will become “accounts”. Quotes will become “deals”. Every report you write will reach for fields that don’t quite mean what they should.

Travel-native systems give you travellers, trips, stays, transfers, flights, suppliers, agreements, agents and guides as first-class records — not stretched-CRM fields. That difference compounds.

2. How does the enquiry-to-trip-to-quote flow actually feel?

The thing tour operators do every day is move an enquiry through to a quote, then to a booked trip. Watch this carefully in the demo:

• How many clicks to convert an enquiry into a trip?
• Does the quote follow the enquiry-to-trip conversion automatically?
• Can you have multiple versions of a quote and keep them all attached to the trip?
• When the quote is sent, does the system lock it so a junior team member can’t silently edit a number after it’s gone out?
• Does the trip remember the original enquiry source (agent, market, designer)?

A lot of systems do the first step (capture an enquiry) and the last step (run an itinerary) competently, but the middle — the bit where commercial decisions get made — is where they break. Make the demo do that bit at full speed.

3. How does the supplier library actually model the supplier?

A supplier library that just stores name + email is a contacts list, not a supplier library. What you actually need:

• Rooms / room types with descriptions, max occupancy, child policies
• Rates with seasons — high, shoulder, low, and your own custom periods
• Levies and extras (park fees, conservancy levies, single supplements)
• Contacts — reservations, ops, accounts, and the WhatsApp number that actually gets answered
• Transfers and schedules — what flights they connect to, how long the transfer takes, what days it runs
• Linked agreements — STO terms, commission, cancellation policy
• Linked trips and history — “which of my trips use this supplier?”

If the supplier record can’t model those, you’ll re-enter the same supplier data into every quote you write for the next decade.

4. Where does AI fit — and what does it actually do today?

Every vendor’s deck has an AI page. Most of the time the AI is doing one of three things: generating itinerary prose, drafting client emails, or doing a one-shot “summary”. Helpful in moments — but rarely the difference between weeks of admin and days.

The questions that separate marketing from product:

• Is the AI feature live or a roadmap promise? Ask to see it work, on their data, in the demo.
• Does the AI ever auto-send, auto-save commercial data, or publish anything without a human reviewing it? If yes, run.
• Is the AI scoped to one job each, or is it one giant chatbot trying to do everything? Narrow tools beat general chatbots on quality.
• Is commercial data (margin, supplier costs) sent to the model? It probably shouldn’t be.
• If the AI breaks tomorrow, can the team still do their job? The honest answer should be yes.

For comparison: the four AI actions live in FieldDesk today — missing-info checker, operations brief, follow-up drafts, traveller profile summary — are each scoped to one job, land as a draft for review, and never auto-send.

5. What do agents and partners actually see?

Most tour operators have agents. Some have lots of them. The agent portal is the part of the software your most important relationships interact with — and it’s the part demos gloss over.

Questions worth asking:

• How does an agent log in? Magic link is good; password resets every visit is bad.
• Can an agent see only the trips you’ve shared with them, not your whole pipeline?
• Can an agent approve or request changes to a proposal inline, without email tennis?
• Do they see your internal notes? Your margins? Your supplier costs? (Should be no.)
• What happens if the same agent works with two of your workspaces?

6. How does branding work on client-facing outputs?

The internal system is the operator’s tool. The proposal, itinerary and trip brief that the client sees should be in your brand, not the vendor’s. Most vendors stamp their logo somewhere on the output. The good ones let you fully white-label proposals with your trading name, accent colour, logo and signatory.

7. Multi-workspace, roles, and the actual permission model

If you run more than one brand, or you’re thinking about it, ask:

• Can a user belong to more than one workspace?
• What roles exist out of the box? (You’d expect at least Owner, Admin, Operations, Travel Designer, Finance, Read-Only.)
• Can a Read-Only member be safely shown a trip without seeing margins?
• How is data isolated between workspaces — at the database level, or just the UI?

8. Mobile and on-the-road

You don’t need a full app. You need to be able to open the dashboard on a phone in an airport and answer a guide’s question without three minutes of zooming. Test this in the demo. Open the demo URL on your phone. If it feels like 2014, that’s a real signal about how much the team uses their own product.

9. The migration story

This is the question most demos rush past. Ask directly:

• Do you help map our current data into your model?
• What does a first-90-days look like?
• Will we be supported by a person, or a help centre?
• What does churn look like — how many customers leave in year one, and why?

10. The awkward final questions

Sit in the silence and ask them anyway:

• What can your product not do yet?
• What feature on your roadmap would you not bet money on shipping this year?
• Who else have you lost a deal to recently, and why?
• Can I talk to a customer who tried you and chose someone else?

Vendors who answer those questions honestly are the ones you can trust to ship the next feature you actually need. Vendors who deflect — usually can’t.

If you’re early in the search

Make a one-page list with three columns: must have today, nice to have soon, can wait. Don’t let any demo move things between columns. Most regret in this category comes from buying for the demo, not the daily work.

And if FieldDesk sounds like a fit, we’d love to show it to you. Book a demo and we’ll do the awkward questions live.