By now, most everyone has heard the term “big data.” It captures the notion that today we have the ability to collect and analyze vast amounts of information – information from customer transactions, social media sites, public records, digital systems and sensors, and much more.
Here at Bing Ads we see big data up close every day and one of our core priorities is to enable you, our advertisers, to do more with this data. We have 158 million unique searchers, each month we serve 5.3 billion search queries and post tens of billions of ads each year.
All this data allows us to analyze consumer behavior in search of opportunities for advertisers. We’re often asked by advertisers to provide insights that can improve their campaigns. The most common areas that advertisers want to know more about are: prediction, attribution and optimization. More specifically:
Prediction: How do I better predict what my customers will do or want to do?
Attribution: How do I better quantify the contribution of each of my efforts?
Optimization: How do I continue to test and optimize my campaigns quickly and efficiently?
Here at Bing Ads, our team is using the data within our platform to help answer some of these questions advertisers are constantly asking for. Our team’s using Big Data to create new insights and opportunities for advertisers. For example...
Being able to accurately forecast clicks, search volume and spend can mean better bid and budget management and more sales for your business, but forecasting is difficult. Most often, advertisers only use their own data to create forecasts. This has many drawbacks, including not being able to control for user mix, ad position etc. At Bing Ads, we have the unique capability of taking our own marketplace data, looking at seasonal and historical trends, and combining it with an advertiser’s data, to better predict search volume, clicks and spend. This advanced forecasting gives advertisers a more accurate look ahead, so they can better anticipate what types of campaigns and strategies will be most effective.
Secondly, we want to help advertisers better quantify the contribution of each of their campaign efforts. Non-converting keywords are often paused or ignored, but what if they have some intrinsic value? To find out, we’re piloting with select advertisers something called Keyword Assist, which tracks search queries that may have influenced the converted search term. Understanding what users are searching for during their entire decision journey can help advertisers better quantify the contribution of each of their keywords.
Keyword Assist data is quite common, but at Bing Ads we’ve taken a more holistic approach. Not only do we provide data on keywords advertisers are bidding on, but we also provide data on keywords that an advertiser may not have in their campaign. This approach helps advertisers better attribute performance across their campaign while also identifying potentially new keyword opportunities.
Lastly, we seek to give our advertisers the ability to optimize their campaigns quickly and efficiently. For example, advertisers have a number of tests they can perform with their ad copy. But how can they quickly and efficiently take the first step forward? At Bing Ads, we’ve looked at our marketplace to determine which combinations of ad copy elements and ad formats are most effective. We’ve analyzed millions of ads within specific verticals and have identified the attributes of the ads that drive the most clicks. This helps advertisers save time, so they can begin testing these top ad combinations to see if they are effective for them.
These are just some of the sample insights that we’re bringing to advertisers. With so much data available to us we know it’s important to use it to benefit our customers. It’s our goal at Bing Ads to provide more and better insights to advertisers; make more data available to you so you can make better decisions; and use our data to make improvements in our marketplace.
For more information, be sure to check out our presentation, Unlocking Advertiser Insights with Big Data on Slideshare.
Questions? Comments? Feel free to leave them below.