Travel 15th October 2020 - 5 min read

Avoid the data dump

By Jason Kramer

I’m Jason, your Data Insider – a former travel management consultant turned data geek. In my new gig at leading data intelligence firm PredictX, I have seen what an effective data strategy can do for all-round travel programme improvement. I want to share this knowledge to those wanting to start or improve their current data strategy. Each month I will be sharing a different tip. This month’s tip is: Avoid the Data Dump

It’s the first week of the month. My TMC travel data reports and dashboards are coming and my anxiety is rising. Is it accurate? What’s important? What’s the data telling me I should do? What do I communicate to my stakeholders? Do I have to ask my consultant for insights?

This is a common refrain from Travel Managers for a phenomenon I like to call the “data dump” – a plethora of data arriving with minimal verification process and little-to-no context for what is most important. And that’s just TMC data. Also coming is reports on credit card and expense data. This data is delivered weeks after the trips have taken place leaving you little opportunity to change traveller behaviour.

Accurate and timely data has been proven to empower a travel programme with better strategies, improved performance-to-goals, increased leverage with suppliers, and more meaningful and effective engagement with both travellers and stakeholders. But what kind of data makes the difference?

The traditional transfer of reports and dashboards from TMC’s to clients reflects an “analyse and fix” model to travel management, whereas a more impactful approach to travel management is to “predict and prevent.” This involves using accurate data communicated succinctly together with advanced technology like machine learning that allows you to see, not only what is happening in the past, but what is happening in the present and what will happen in the future.

How do we get our programme data to no longer be a “data dump” but to “predict and prevent”?

First, we need to start with the basics: getting past data right. Sounds easy enough but most travel data is complex and can easily lead to incomplete analyses with poor data quality and no context to make it meaningful. Follow the steps below to build a solid foundation before we can move on to more advanced predictive analyses.

Step 1: Use more than just one source

Leakage is not a bad word. Rather, unmanaged leakage is. Studies show roughly 40 to 45% of programme spend falls outside of TMC channels. In today’s data-led world, we simply need to use more than one data source to capture these bookings.

What other sources should we include? I always say start with the big three: TMC, Credit Card and Expense data. These three sources will provide a complete picture of your spend. The next logical data source is the HR feed to deliver insight into individual department and traveller spend.

The combination of these data sources will offer a complete, all-in view of your travel programme and lay a strong foundation for “predict and prevent” programme management.

Step 2: Introduce an automatic data verification process

What good is data if you can’t trust it? If it’s wrong once, it raises the concern for past or future data issues. Everyone can agree data verification can be a nightmare of a process, but it is essential. A comprehensive data verification process can be simplified when you catch it early and from the source.

A couple of examples: hotel property names can be coded in different ways and in different formats. Even dates can have different formats.

Another example for potential data errors is when we combine data from two data sets. Two amounts in two different currency can be merged into one report, yet both may only be calculated in one currency with no conversions made.

The answer to normalize these differences? Advanced technology.

Computer and data science techniques along with machine learning models can be used to automatically reconcile your source data against a master database. The more data the system processes, the more it learns and is able to catch issues early.

A trusted and strong data verification process from initial ingestion and to final output is another strong building block to “predict and prevent.”

Step 3: Use your data to tell a story

Now that we have incorporated multiple data sources and applied a data verification process, it’s time for your data to come to life. Data needs to tell a story, it needs to drive actionable insights, and it needs to be presented in a compelling way.

Charts, graphs and other visualisations are a must in capturing your audience’s attention. When presented with data and analysis, you and your audience should always be able to easily see meaning behind it. Does this analysis answer the questions you and your stakeholders care about? It’s great to report travel policy compliance across the company. It can reflect that the policy is effective and employees are following it. However, in an environment with data from multiple sources you can further isolate compliance by department and by traveller. With these tools we can understand compliance on a deeper level and isolate departments and/or travellers to correct behaviour sooner rather than later.

Data is all about context and, in travel, it should be no different. Let your data tell a story at a high and granular level.

Data should never be a dump where you and your stakeholders have to dig for the nuggets of insight you care about. Rather, use these steps so your data is accurate and capable of answering questions your stakeholders want to know. In an industry that has more internal visibility than ever before, there is no better time than now to implement a data strategy that adds value to the business.

Now that you know the foundations of ideal data visibility, our next blog will focus on moving away from “analyse and fix” to “predict and prevent” and using predictive analytics and forecasting.

PredictX combines, cleanses and verifies data from multiple sources to provide engaging analyses and keeps stakeholders engaged through The Story – an NLP generated automated reporting feature. If you want to see how we can help you avoid the data dump, book a consultation with us.

Jason Kramer
By Jason Kramer
5 min read

Leave a Comment

Your comment will need to be approved.

PredictX uses cookies to improve your experience on our website. Cookies enable you to enjoy certain features and they also help us understand how our site is being used. Find out more here. By continuing to use our site you consent to using cookies. More information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close