The Problem With Airline Tendering

Author: Tony O’Connor 

If you spend more than around $350,000 on air travel you can get a discount deal from the two major carriers covering domestic and international travel. However, at entry level, the terms are fairly rigid and formulaic. If your spend is a couple of million or more, the terms loosen up and can be improved through negotiation.

The airline offers consist mainly of an array of percentage discounts off public retail fare levels. The discounts vary per fare type. The cheaper restricted fare types carry low or no discount. Apart from extra goodies like status matching, accelerated loyalty points and discounted lounge fees, that’s basically it. What’s the main thing missing? The fares themselves.

A 20% discount is better than a 15% discount, but not if it applies to a $500 fare compared to a $440 fare. In this case, the first offer results in a fare of a $400 and the second offer produces a fare of $374. To even come close to comparing the two airline offers, you need to factor in the base fares to which the discounts apply.

Most effort in airline tendering goes towards improving the terms of the standard airline offer formats; maybe improving discounts here and there by a few percent and squeezing out a few improvements in the secondary benefits. This can be the case even when you get assistance from your TMC. In the meantime, there can be differences of hundreds of thousands or even millions of dollars in the true value of the airline offers that lie unrecognised. I suggest that the main game should be evaluating the actual total value of the airline offers in terms of the total annual airline spend they would result in. If you can estimate that your spend under airline A‘s offer would be around $4.6 million compared to $4.25 million under airline B, you should know that. Unless your spend is very large, the truth is that airlines are usually quite firm with their initial offers. Your efforts to improve the discounts might save you some dollars. But the amount is usually minor compared to the difference in the true total value of the two airline offers.

So, we need to input the retail fares. This is not complicated but there are a lot of numbers to deal with, and you need a nice spreadsheet. The model is equally applied to estimate total spend under each airline offer. It has to estimate total annual spend for each fare type for each route. So, say for Brisbane Perth, for fare type N, you input a retail fare of say $565, the airline’s offered discount of 12%, and your annual flown sectors of 82 (how often you flew this fare type on this route over the last year, as the best estimate of how often you will in the future). This results in an estimated annual spend of $40,770.

You simply do this for each route and each fare type, including those NOT mentioned in the airline’s offer, and add up all the estimated total spends. And, to make life bearable, you would include only your top 10 or 15 domestic routes that account for 80% or 90% of your domestic air spend. The task at hand is comparing the offers. You don’t need 100% coverage. It’s like safely calling an election result with 70% of the votes counted. For international, you would include all routes/classes with significant spend.

For accuracy, the model needs to manage a few important details, such as having a system to make the two airlines’ different arrays of domestic fare types equivalent, and a way to include Jetstar, Tiger and Rex.

For assessment on the international side, qualitative factors are very important. As well as total cost per route, we need to assess things like service frequency, total travel time, times of departure and arrival, stopovers and transits, seating specs and fare conditions.

Then, the model should monitor deal compatibility. Airline discounts come with conditions; some combination of minimum spend and market share. With domestic, they are usually overall requirements. For international, they can be overall, per route or per region. The best international outcome is usually some combination of cherry-picked offers per route from one of the two main airline deals, plus those of other airlines where you have negotiable per-route volume. But cherry picking can cause offers from other airlines on other routes to crash. International cherry picking of offers per route can also cause the entire domestic offer or even the whole combined offer of one of the two airline groups to crash. The deals have to be compatible. You have to be careful. In fact, the most effective target of your negotiations should be relaxation of the market share and spend conditions to accommodate more cherry-picked international route deals, rather than the discount levels.

And so, we have our model. But there is a third layer we need to add to discounts and fare levels. That is fare type availability. There is no value in contracting to a great discount on a fare type if the airline typically rarely makes that fare type available. This is important. And so, when you input your sector numbers per route per fare type, the number has to factor in likely availability. If it doesn’t, you can substantially over-state the true actual value of the airline’s offer.

Where to get the various inputs? The discounts are obviously in the airline’s offer. The many fare levels that you need per route per fare type can be obtained from the airline. They won’t volunteer this data, but if you give them the routes¸ if they want your business, they should give you the retail fare amounts for ALL the alpha fare types. Availability per fare type per route? This important input is trickier. The airlines can’t provide it because they don’t know it themselves. It’s a constantly changing thing driven by computerised airline yield management. But good approximate inputs here are much much better than none at all. The best source is therefore historic data. If your airline spend isn’t large, the sectors per airline per route per fare type will be too scratchy. Too many of the inputs will be zero or single figure numbers that are not meaningful or accurate. Such inputs will distort the outcome. You need availability figures from a large data set, as could be provided by a consultant or a TMC. A TMC might have issues in providing this level of detail on airlines from which it receives commissions. If you cannot get large-data availability inputs, it’s better to leave this layer out of the model. But that’d be a pity.

This all sounds like a fair bit of work. But once you have the model it’s a pretty straight forward exercise, and it yields a good procurement result. The pay-off can be substantial. And as a side benefit, once you have a good airline assessment model you can use it for other useful things. With a few clicks you can see exactly the value of an airline offer compared to the status quo, and the actual value of targetted improvements to an airline offer, such as a better discount on a particular fare type. And you can accurately estimate potential savings through better fare type usage by your travellers.

To get best airline deals without unreasonable effort, you need to have data and a model that is at least as good as the airlines’. Otherwise, you’re flying on one engine.

btTB-GBTA Annual Conference will be held on 1st & 2nd May 2019 in Sydney. Read more here. 

Tony O’Connor has run Butler Caroye since 1998, the independent travel management consultancy. He specialises in TMC, IT, card and airline advice and tendering, and offers the Zoolu airline assessment system. He is also the Director for Australia and New Zealand of the Global Business Travel Association.