Being results orientated is something many airline pricing professionals strive for, and this is positive, but it’s not always the ideal approach. Tom Bacon has some warnings

What’s the bottom line? Did we correctly anticipate the no-show factor over spring break? Did we save enough seats for high fare demand? Did RASK (revenue per available seat kilometer) beat the forecast?

Sound familiar? What these questions demand is results and it’s true that many of us are results oriented. Indeed we all seek teammates who are results oriented. Not surprising really, give that results are often intuitively appealing. Rather than having to rely on theories or ‘intent’ or probabilistic models, ‘results’ feels like it is a highly objective measure. Unambiguous. Definitive. Unarguable. And, by definition, results are ‘good’: measured by higher revenue, greater profitability, improved customer satisfaction, increased accuracy.

Unfortunately, however, ‘results-oriented’ is not always the ideal.  In fact, ‘results oriented’ can be misused or misleading. ‘Results-oriented’ can steer you in the exactly wrong direction. There are two main problems with ‘results’.

1.  Results can be subjective. We may review and measure ‘results’ in ways that lead us away from the optimum. 

One common tendency is for us to look for results that support our hypothesis (business psychologists call this ‘affirmation bias’).  There’s an abundance of data and many different ways to look at or interpret it. But, with this natural bias, we focus on the data that seems to support our original view and dismiss or ignore data that doesn’t.

Similarly, we pay too much attention to anecdotal evidence. We conclude the data itself is inconclusive or messy and gravitate to specific instances – which may not have any statistical validity at all.

Even if we regularly apply statistical tests, many businesses look only at one-tailed tests rather than two-tailed tests. We prove to ourselves that our hypothesis can’t be dismissed – but not that its converse cannot also be dismissed. We don’t mind this ambiguous result if it is consistent with our affirmation bias above.

Here is an example:

At most airlines I’ve worked for, we have prided our insistence on ‘results’. And, in a pricing experiment at one airline, the pricing manager concluded from the data that, in fact, his hypothesis was correct. It had been his idea originally so, perhaps, he had ‘affirmation bias’. Actually, looking at the same data, I reached the opposite conclusion. But I too may have experienced ‘affirmation bias’ since the idea was risky so I was sceptical of his hypothesis from the beginning. We were both ‘results-oriented’! 

In fact, we should have agreed on the criteria ahead of time. The metrics need to be clearly stated before the experiment is conducted – and developed in a way that recognises the danger of affirmation bias.

2.  We correlate our actions with the results when economic relationships are generally much more complicated.

We all know how complicated pricing can be – an initiative can fail due to a poor demand forecast or misjudgment on elasticity - or because of an ineffective marketing effort or due to sales and distribution issues. Actual sales is subject to so many factors besides pricing. Too often, we correlate the ‘results’ with the change we made most recently rather than explore all of the possible explanations.  Economists like to identify relationships that hold ceteris paribus, or if everything else is held constant.  In the real world, nothing is ever truly held constant.

Here is another example: 

Currently, low fuel prices are making all airlines suddenly look ‘smart’. Thus, Alaska is doing fine despite new competition from Delta; Frontier is successfully pursuing its ultra low-cost carrier strategy; the American Airlines/US merger is brilliant. Each of these airlines could be truly pursuing winning strategies – strategies that would yield positive results even with higher fuel prices. But the tailwinds from lower fuel prices is certainly the biggest driver to current U.S. industry financial success.

We need to review results in ways that acknowledge the complexity of factors behind them. Whether this means multi-factor statistical analyses or more highly controlled tests or simply, greater scepticism regarding results, we need to avoid jumping to conclusions too quickly.

Certainly, being results-oriented is a positive trait, in RM and in other business situations. But we need to approach these results with some humility. We need to agree in advance on the metrics used to validate hypotheses and we need to become our own devil’s advocate as we continually test – and learn – by taking a rigorous approach.

Tom Bacon, 25-year airline veteran and industry consultant in revenue optimisation pens a bi-monthly column for EyeforTravel.   Questions?  Email Tom at or visit his website