Regular columnist Tom Bacon shares marketing lessons from Choice Hotels, and taps his own personal experience to find that marketing spend is more complicated than you think

How do you assess the value of marketing spend? The ‘multi-touch attribution model’, a sophisticated attribution technique, was presented by Olga Nielsen, director of marketing and distribution analytics for Choice Hotels, at the EyeforTravel Smart Analytics conference in Atlanta earlier this year. 

Rather than ‘last click attribution’ this model can assess both internet-based and more traditional marketing approaches and statistically separate the incremental revenue value of each. Every touchpoint can be incorporated in the analysis, which is increasingly necessary as customer purchase decisions flow across multiple sites and touchpoints (see Accurate attribution remains travel’s biggest headache, EyeforTravel June 5). The model can identify the value of the initial web ad versus the special offer versus the frequent flyer link.

Potentially, marketing initiatives have different impacts at different airlines. Southwest Airlines is known to use fare promotions more strategically and jetBlue relies more heavily on emails. Some airlines, on the other hand, are said to spend more on pay-per-click advertising. This raises important questions:

  • How sophisticated is each of these carriers, however, in assessing the value of their different approaches? 

  • Does each fully exploit analytical techniques like the multi-touch attribution model (MTM) to optimise the level and allocation of their spend - as they seem to reach such different conclusions?

Of course, marketing initiatives themselves are often not the biggest driver of bookings. Factors that explain most of any growth or change in bookings include elements like: i) economic cyclicality or seasonality, ii) competitive factors like relative capacity growth and iii) market segmentation factors like growth in target demographic groups. 

As Nielsen noted, the MTM model could attribute as little as 5% to marketing initiatives. She notes that most companies, without the benefit of a sophisticated attribution model, attribute much more than they should to targeted promotions. In many cases, the other factors – macro-economic, competitive environment, long-term brand awareness – represent the largest factors in booking volume – all without any current spend!

View from an airline

At one airline I worked with, during a cash crisis brought on by bankruptcy, we effectively terminated most marketing spend, and relied almost completely on local customer loyalty and online travel agency distribution. It worked! Revenue stayed strong throughout the bankruptcy period despite few marketing initiatives.

In retrospect, much of our historic marketing spend had actually supported our brand – even spend that was designed for a short-term boost in bookings more fundamentally conveyed that we were a uniquely ‘fun’, low fare carrier in comparison with our major competitors.  Such spend often drives a consistent level of demand over a longer period of time; it contributes as much to a base level of demand as it does to short-term demand. 

Arguably, in fact, marketing spend was not very productive on a short-term basis – we could keep revenue strong simply relying on our long history of investment.

Despite the multitude of factors that impact bookings, the MTM model is nevertheless able to focus on the ‘controllable’ factors – the separate impact of each of the short-term marketing initiatives across the various touchpoints. In fact, the end result of the model is to recommend an ‘optimal’ mix – in other words, what allocation of spend across channels, across touchpoints, across segments is likely to produce the best overall booking result.

The model will recommend increased spend on certain initiatives and decreased spend on others to drive the highest forecast revenue, and associated return on investment. Frequently, however, it will identify significant reductions in spend for certain touchpoints, or even overall, given relatively little short-term incremental value attributed to those initiatives. 

  • Nielsen has some recommendation including:

  • Use the sophisticated attribution model to assess the relative incremental value of different initiatives

  • Don’t blindly accept the analytical results as the optimal long-term answer.

  • Recognise the way the model is prone to under-estimate the longer-term impact of reducing spend in any individual medium

  • Test recommendations against experience with customer behaviour and purchase processes. The ‘optimised’ mix derived by the model needs to be adjusted for short-term versus longer-term goals, potentially increasing spend for those initiatives that may have a longer term impact on demand, but which would be unlikely to be captured directly by the model.

Tom Bacon has been in the business for 25 years, as an airline veteran and industry consultant in revenue optimisation. He leads audit teams for airline commercial activities including revenue management, scheduling and fleet planning. Questions? Email Tom or visit his website

Related Reads

comments powered by Disqus