Here at Koddi, we spend a lot of our time looking at performance metrics like impressions, clicks, bookings, revenue, conversion rates…. the list is endless.  We focus on these metrics because understanding performance allows us to identify new opportunities and optimize our campaigns. A laser like focus on key performance indicators often makes sense and can lead to big performance lifts. However, focusing only on KPIs can sometimes create a blind spot in the advertiser’s strategy. Accurate core data is a requirement to participate in the metasearch space. Without it you are likely losing impression share or even worse providing a bad customer experience. Advertisers who focus on improving their core data stand to see significant improvement in their metasearch performance.

What is Core Data?

We define core data as basic hotel information which includes hotel name, address, phone number, amenities, latitude and longitude etc. This information is usually not at the top of mind for marketing executives. Most assume that their data is good, or at least good enough. This mindset can be detrimental and can often lead to money left on the table. When we onboard a client, it is not uncommon for us to find data issues with 10-20% of a chain’s properties. That means that 1 in 5 of the properties currently being advertised are using bad data.

Best-in-class advertisers constantly maintain, optimize, and redistribute their core hotel data. A common misconception among advertisers is that core hotel data is a static data set. This often couldn’t be further from the truth! Over the course of a year, even an advertiser with a portfolio of only a handful of hotels will have openings, closings, remodels, expansions, changes in amenities, phone number changes, and address changes, just to name a few.

Why should you care about your Core Data?

When changes in the core data are not reflected accurately, the result is an outdated data set which can lead to significant negative consequences. Consider that metasearch engines can only display good ads when they can match your property to their own data. This pairing is done through a clustering of data points. When data is bad or missing, it can cause your ads to not be seen, or even worse, to be seen for the wrong property.  Using the picture below, let’s consider what happens when a poor match is made due to inaccurate core data.

Even though the data does not match exactly, the metasearch engine determines there is enough in common to list an advertiser’s property. The property in question recently had its hotel name changed, but that information has not been updated in the core data. The customer decides to book the room and takes note of the hotel name and the dates they booked. When it comes time to travel, they attempt to find the hotel by searching its name and city. As far as many search engines are concerned, that hotel no longer exists and the customer becomes frustrated as they attempt to locate a hotel that has a completely different name. Imagine if the phone number had also changed and the customer could not get in contact with the hotel.

What kind of customer experience would this misinformation contribute to? Core Data can often times be the first impression a customer gets of your property. It not only provides valuable information, but also sets the consumers expectation of what their travel experience will be like. Failing to meet a customer’s expectations guarantees the perception of a bad experience. Bad data could mean a bad first impression. Hotels are judged by their ability to create an enjoyable experience and it can be an uphill battle trying to win the approval of a customer who starts their trip out frustrated. Inaccurate data can be seen as a sign of incompetence, or even worse interpreted as an attempt at deception.  A customer expecting to have access to a pool, only to find out it is being renovated is likely to feel that they have fallen prey to a bait and switch.

How can we quantify bad data’s impact?

The full extent of the impact of inaccurate Core Data on a campaign can be difficult to quantify. More times than not, inaccurate data leads to silent failures that add up over time to become significant shortfalls. Perhaps the fact that these failures can be silent is the reason many advertisers overlook the importance of spending time with their data. Without hard data how can we begin prioritize maintaining our Core Data?

Well fortunately (or unfortunately depending on your viewpoint) the rise of social media and review sites has allowed customers to openly share their opinions and experiences. Hotels are now more than ever beholden to their customers and the reviews they leave behind. A good review will lead to more bookings, while one scathing review can be detrimental to an online campaign. We know that bad data can lead to bad reviews, but how do bad reviews affect a metasearch campaign?

We recently ran a test in order to determine the best way to set a bid when a property has not had a significant amount of traffic. Our theory was that we could use Google User reviews to accurately predict which properties would perform better and thus were worthy of a higher bid. As we conducted our experiment, we noticed that the average RPC (revenue per click) steadily increased as the hotel rating improved. In fact, we found that one click for a 4.5 star property was 5x as valuable as a click for a 1 star property. When we used this data to determine our bids, we were able to improve the conversion rate of the campaign by 13%. This proved the importance and reliability of review data as a predictor of hotel performance.

Now, we would not imply that perfect Core Data will lead to your hotels rising from 2 star properties to 4 star properties. However, it is not a stretch to suggest that the removal of one or two bad reviews could improve a property’s aggregate review score by .5 to 1 stars. Let’s imagine we have a large supplier that is currently advertising 5,000 properties. They pay a fair amount of attention to their data and their listings are accurate 80% of the time. So 1,000 of their properties currently suffer from inaccurate data. We will say that over the course of a year, this data has led to half of these properties receiving a bad review due to the bad information.  This gives us 500 properties that have received 1 to 2 negative reviews due to bad data.

So let’s say on average that the reviews have caused the hotels to receive a rating of 3.0 stars and without those bad reviews the average rank would be 3.5 stars. A property’s average RPC grows by 35% when a review rises from 3.0 stars to 3.5 stars. So let’s assume we have an RPC of $10 for a 3.0 star property. We would then expect an RPC of $13.5 for our properties that our 3.5 stars. So if the 500 properties generated a 50 clicks a month we would estimate those properties would generate 25k clicks and $250k in revenue….Not bad. However, if the properties had not received the bad reviews due to bad data, we would expect revenue of $337.5k. That is an incremental revenue gain of $87.5k per month. Extrapolate over the course of a year and we are talking a $1.05MM gain. Who couldn’t use an extra $1 Million?

How do we focus Core Data?

A strong Core Data program not only has mechanisms to track the quality of any outgoing Core Data, it also tracks how data shows up at its destination. Many publishers either merge data from several different sources automatically or have processes with less quality assurance than you might expect. Both of these scenarios can contribute to unexpected results.

The team responsible for this data should have a way to measure accuracy across partners over time and should seek out feedback explaining non-matches. The team should make updated, fresh, optimized data available to publishers and marketing partners, and ensure that those partners use it.

In addition to accuracy, the team should also strive for constant improvement of the quality of the data through an optimization plan. Location data (for better or worse) tends to move through organizations pretty slowly. It’s realistic to budget a few minutes a week to positively affect the quality of Core Hotel Data by expanding / validating amenities, optimizing content, and confirming key NAP (name, address, and phone number) information.

Best-in-class Core Data, by definition, is accurate, rich, and easily accessible. The more this is the case across all channels, the more effectively advertisers are able to respond to demand. If it isn’t realistic to manage this internally, these activities can be outsourced to any number of companies for relatively little investment.

On the surface Core Data may seem to be not very actionable or interesting. This misconception keeps advertisers form taking steps to make meaningful optimizations to their data. That’s why we believe a disciplined focus on Data excellence represents a significant growth opportunity for many advertisers.

If you are interested in learning more about core data  check out our eBook here.