Travel management RFP: travel management company RFP template, TMC RFP questions and corporate travel RFP scoring

How many travel management companies should you invite to an RFP?

Five or fewer. GBTA advises defining your top five to ten programmatic goals, holding conversations with TMCs of various sizes, and then narrowing candidates to five or fewer before the formal process begins. FCM makes the practical version of the same point from the receiving end: one expert cautions against soliciting too many responses and suggests reviewing six or seven proposals at most, because past that the exercise stops being an evaluation.

The arithmetic is unforgiving. Each serious response is fifty to a hundred pages plus a pricing workbook. Five responses is a week of somebody's life read properly. Nine responses is not thoroughness, it is a guarantee that you will skim, and skimming favors the proposal with the best design rather than the best terms.

Getting to five means doing the disqualifying work before you write the document. Rule out anyone who cannot cover your geography, anyone whose smallest client is larger than your whole program, and anyone whose technology stack cannot connect to the expense or HR systems you are keeping. A thirty-minute call each does that, and it is far cheaper than reading a proposal you were never going to accept.

There is one composition trap worth checking on the way. Corporate ownership in this category is more concentrated than the brand names suggest: Amex GBT owns both Egencia and CWT, so a shortlist naming all three is one parent company presenting three times. If independence of bid matters to your procurement team, verify ownership before you finalize the list. Our Navan vs Amex GBT comparison covers that structure, and BCD Travel pricing covers the largest genuinely independent alternative.

What questions should a TMC RFP ask?

The useful questions are the ones where two good TMCs will give you materially different answers. Anything all five bidders will answer identically is filler. These are grouped by evaluation category so they map straight onto the scorecard, and they are phrased so a vague answer is visibly a vague answer.

• Pricing: complete our pricing table in the units specified, and also express your entire proposal as a blended cost per transaction against the booking data we supplied.
• Pricing: is the air transaction fee bundled or unbundled, and does a hotel booked alongside a flight generate a second fee?
• Pricing: what is charged separately from your fee, including booking tool licensing, and who holds that contract?
• Pricing: what are the minimums, volume commitments or shortfall charges, and what happens to our fee if travel volume falls thirty percent?
• Service: where are the agents who would answer our calls physically located, in what hours, and what is the escalation path outside those hours?
• Service: what is your average speed to answer and your abandonment rate for accounts of our size, and what SLA penalty attaches if you miss them?
• Service: name the account manager who would be assigned to us, and how many other accounts they carry.
• Technology: which booking tool would we use, do you own it or license it, and what does a change to our policy configuration cost and take?
• Technology: list the integrations you have live today with the expense, HR and single sign-on systems we named, and give a reference for each.
• Data: who owns our booking data, in what format can we extract it, and what happens to it if we terminate?
• Duty of care: how do you locate a traveler in the first fifteen minutes of an incident, including one who booked outside the tool?
• Unused tickets: how are unused ticket credits tracked and applied, and what happens to open credits if we move to another TMC?
• Implementation: give us a week-by-week implementation plan with a named lead and the specific inputs you need from us in each week.
• References: three clients of our size in our industry, and one client who left you in the last two years.

The unused ticket question earns its place. Airline credits are issued against the agency that made the booking and do not automatically follow you to a new one, so on a program of any size an unanswered version of that question is a five-figure surprise years later. Ask it in the RFP, when you have leverage, rather than at exit, when you do not.

The reference who left is the highest-yield question on the list and the one bidders most dislike. A TMC that can produce a departed client and talk honestly about why is showing you something real. One that cannot is telling you what its exits look like.

Do you need an RFP, an RFI or an RFQ?

They answer different questions and using the wrong one wastes a quarter. An RFI asks who is out there and what they can do, and it is the right tool when your shortlist does not exist yet or when the market has changed since you last looked. An RFQ asks for a price against a specification you have already fixed, which only works when you genuinely know what you want down to the line item. An RFP asks bidders to propose an approach and price it, which is the right tool for a service as configurable as travel management.

Most buyers need an informal version of the first and then a real version of the third. Thirty-minute discovery calls with seven or eight TMCs will tell you more than a formal RFI, and they let you disqualify on fit before anyone invests in a document. Then run one properly scoped RFP with the five who survived.

The case for skipping the RFP entirely is stronger than the category likes to admit. BCD lists running an RFP solely to get better pricing as a mistake and suggests an annual vendor review instead, and it is right: if your incumbent is performing and you only want a better number, a benchmarking exercise and a renegotiation costs a fraction of a competitive bid and does not put your program through a migration. Run an RFP when the relationship, the service model or the technology is wrong, not when only the price is.

What buyers get wrong in a travel management RFP

Five failures account for most disappointing outcomes, and four of them happen before the document is sent.

• Reusing last year's template unchanged. BCD lists using generic industry templates without customization as a top mistake, and FCM notes buyers still recycling questions built for a pre-2020 travel program.
• Running the RFP only to move the price. That is what an annual vendor review is for, and it does not cost you a migration.
• Giving bidders different information. If the incumbent has your data and the challengers have a summary, you are not comparing proposals, you are comparing access.
• Not knowing why travelers book outside the program. If leakage is unexplained, a new TMC inherits the cause and the same leakage reappears under a new logo.
• Deciding on the presentation. The best-presenting bidder is the one with the largest sales function, which tells you about their go-to-market and nothing about their after-hours desk.

There is a sixth that is less a mistake than a mismatch, and it is worth checking before any of the above. GBTA puts the threshold at which a TMC becomes beneficial at travel spend over roughly $250,000 annually, or a program large enough to negotiate directly with airlines and hotels. Below that, an RFP is a substantial project whose prize is a contract that cannot save enough to repay the staff time. The honest answer for a company at that size is usually a self-service platform, and we compare the trade-off in travel management company versus travel management software.

GBTA also frames the upside realistically: organizations can save anywhere from 5 to 50 percent of travel spend with a TMC, depending on the starting maturity of the program and the volume of travel. The width of that range is the point. A company with no policy, no preferred suppliers and no data is at the high end because almost anything helps. A company already running a mature program is at the low end, and for that company the RFP should be about service and technology rather than savings.

Where TripAgent.ai fits, and where it does not

TripAgent.ai does not bid on travel management RFPs and this page is not a pitch dressed as a guide. There is no agent desk here, no negotiated rate program, no sourcing consultancy, no meetings division and no global service footprint. If your RFP is asking for a partner who will answer a phone in Jakarta at three in the morning, that is a real requirement and it is not one TripAgent.ai meets.

What TripAgent.ai does is the part a TMC bills per transaction to handle: the trip itself. You give it the city, the dates, the budget and the purpose, and it produces a day-by-day itinerary, prices and sequences the days and hotels that make it work, and rebuilds any day on request instead of waiting for someone to reach a support queue. It is $19 a month, published, with no per-trip fee and no minimum commitment.

For a company weighing whether to run an RFP at all, that comparison is worth doing before the project starts rather than after. A transaction fee model at benchmark rates on 400 bookings a year runs into five figures and buys human service you may call on a handful of times. A subscription buys automation you use on every trip and no human at all. Neither is universally right, and the dividing line is usually whether your travel is complex enough that someone genuinely needs to be called.

Plenty of programs run both, and that is often the cheapest configuration nobody proposes. The TMC handles international and executive travel where service is the product, and a lighter tool handles routine domestic trips that never needed an agent. If that describes your travel, say so in the RFP and ask bidders to price the reduced scope, because a TMC quoting your full volume will price differently than one quoting the complex half of it.