Forecasting digital outcomes for hospitality

In a big data panel at the HSMAI Digital Marketing Strategy conference held recently in New York, Peter Kim from start-up MightyHive commented that all of your competitors have access to the same third party data you do, and that your own data is much more valuable. Peter’s comment really struck me as I considered digital behavior data for hotels. Many hotel companies don’t even own their own digital behavior data stream, having outsourced it to web analytics providers. What would it take to justify bringing that data back in house? What new analytic capabilities could emerge from access to this detailed data? For my answers I spoke to Suneel Grover, senior solutions architect for digital intelligence at SAS about how to harness digital behavior data and what kinds of new analytic capabilities are available once that data is captured.

Why aren’t more hospitality organizations using predictive analytics with digital behavior data?

Suneel explained that typical web and digital analytics tools primarily aggregate and report on historical information and do not enable predictive analysis. “While data-driven marketers and analysts have used powerful advanced analytics for many years to perform sophisticated analyses – such as regression, decision trees, or clustering – they have been limited to using offline data, primarily due to restrictions on access rights to online data from third-party technology vendors,” he said. The challenge facing the marketing industry today is progressing beyond the multi-channel analytic limitations of aggregated data collection methods used by traditional web analytics, he explained.

The opportunity is to have a digital data collection framework that enables both business intelligence and predictive analytics, Suneel said. “This methodology requires organizations to collect and OWN the data to allow the analysis of the “who,” “what,” “where,” “when,” and most importantly, the “why” of digital experiences.” This data, if collected and prepared appropriately (e.g. considerations for data integration, quality, and governance), can be merged with your company’s first-party (or company-owned) customer data, and then streamed into your analytics, visualization, and marketing automation systems.

Once you have this data collected, what analytic capabilities can you use?

Suneel told me about a digital analytic capability that focuses on search marketing that leverages onsite behavioral data. He explained that one of the common questions debated in digital marketing is how increasing the paid search ad budget will impact website traffic and hopefully conversions. “When you reflect on this question, it touches on the ability to accurately predict not only what website visitation will look like in the future two week time period, but also have the ability to ask “what if I do this”, or “what if I do that,” Suneel said. “In business and visual analytics, this is known as forecasting, and scenario (or what-if) analysis,” he explained. The unique twist comes when you apply this set of analytic techniques to a digital marketing challenge.

“If your organization owns the digital behavior data stream that captures your website’s traffic behavior, this data can be fed into a forecasting model to first predict what will happen in the next two weeks,” Suneel said. Then, marketers can see how overall site traffic might increase or decrease at varying velocities – by week, day, or even by hour – all wrapped up in a 95% upper and lower confidence interval to highlight the most likely, best case, and worst case scenarios. “Given this information, marketers can determine the impact on predicted traffic patterns of allocating more ad dollars to one channel (such as paid search), versus another (email),” Suneel explained. He told me that by using scenario analysis in conjunction with forecasting, marketing analysts can easily inflate the potential impact of a 10%, 20% or 30% increase in paid search traffic, and then view how this will modify the forecasted prediction of overall site visitation.

“You can end up with takeaways that look like this,” Suneel said:

- A 10% increase in Paid Search visitors would provide a 13% increase in overall site traffic in the forecasted time period (Incremental 3% lift)

- A 20% increase in Paid Search visitors would provide a 28% increase in overall site traffic in the forecasted time period (Incremental 8% lift)

- A 30% increase in Paid Search visitors would provide a 56% increase in overall site traffic in the forecasted time period (Incremental 26% lift)

Access to this data, and the use of predictive analytics can help you understand how adding just a little bit more into one digital channel can have varying, and sometimes, dramatic changes in overall digital traffic.

Every hospitality company could use the results of this type of analysis, but just how difficult is it to perform?

Suneel explained that the emergence of visual analytics in the data visualization technology space makes leveraging these approaches much easier to achieve. “We are living in a time period where analytics cannot be meant for only a select few data scientists,” he said. “The opportunities are too many, and the democratization and accessibility of analytics has begun.”

tags: Big Data, Customer Intelligence, Hospitality Analytics, Hospitality Marketing, Suneel Grover

One Trackback

  1. [...] In a big data panel at the HSMAI Digital Marketing Strategy conference held recently in New York, Peter Kim from start-up MightyHive commented that all of your competitors have access to the same third party data you do, and that your own data is much more valuable.  [...]

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <p> <pre lang="" line="" escaped=""> <q cite=""> <strike> <strong>