Big Data Marketing's Next Frontier: Paid Search

frontiermarkets

It's tough to be a marketer in today's always-changing online environment. It seems that every time we master one new channel, a new and more "promising" channel emerges.

We witnessed a prime example of this with the May 2012 introduction of product listing ads (PLAs). In the face of quickly moving, high-cost changes, marketers were left facing the difficult calculus of reacting to the new user behavior, new advertising formats, new budgets and new marketing strategies – all while measuring return on ad spend.

Right before the 2012 holiday season, some retailers allocated up to 50 percent of their ad spend toward PLAs. The average allocation eventually peaked at 17 percent of overall paid search in the fourth quarter of 2012, leveling off at 15 percent by Q1 2013. While most in the digital commerce world are expecting PLAs to grow in usage compared to text ads, both are considerably important and will represent a large line item in marketing budgets this year.

But, what's behind the curtain, after a consumer clicks on an ad – whether text or visual? It's often pretty poor experiences that have no ability to learn from the consumer's behavior or intent.

It's not like the in-store experience where a salesperson can pick up on cues from shoppers by observing browsing habits or judging reactions to conversations. How could it be?

With PLAs, companies throw usually messy product feeds at Google, bid for traffic and let them do their magic.

In the world of online search, the top engines have done an incredible job with anticipating individual consumer's queries and predicting the most likely match that fits their intent.

Using big data in the truest sense of the phrase, machine learning and advanced algorithms, we almost feel that Google is our personal butler who's worked with us for years. They have essentially emulated a real-life experience. Shouldn't the companies on the other end of a paid click do the same?

A Big Data Answer to a Big Data Problem

Thus, we arrive at big data's next opportunity – recreate in-store experiences by applying a big-data-driven strategy to PLAs. By their visual nature, PLAs are already a more lifelike and relevant experience.

Consumers express what they want in countless ways, and it's virtually impossible to manually match the language to product feeds for the thousands of SKUs that many retailers have. The resulting language possibilities are mindboggling.

At the same time, consumers don't often have specific products in mind. If a company doesn't offer the best fit with an acceptable set of other options on a landing page, it will lose the consumer – forfeiting the revenue and ad dollars.

What's worse is that multiples studies have found that consumers think negatively of a brand that provides poor experiences and will actually pay more for good ones. With many paid search budgets approaching or exceeding millions of dollars, every loss is a bad loss.

Data Reveals Intent

Think about when a consumer goes to a dressing room. They try on a garment, but it doesn't quite work for any number of reasons. A salesperson can bring you something in another size, color, can suggest a sale item or similar item that others like the consumer have bought. They may even remember the last time you were in the store.

This type of intuitive thinking is based off of all of the cues – or data – that reveal intent. Together, you synthesize the best possible options to convert them to a sale.

In the digital world, you could have access to all of the same data points in some form, if interpreted correctly.

Consider a customer who moves away from certain racks, toward others, or lets you know "definitely not this one" as bounce rate or time-on-page. Consumers that quickly bounce off of a landing page or convert more on a particular suggested product page are providing small digital cues where an intuitive system can learn.

In addition, knowing a consumer's previous behavior on a site, such as a tendency to purchase sale items or certain sizes, is synonymous to the insight applied by a great in-store experience. The list can go on and on, but the underlying point is that ecommerce companies have all of the data needed for every consumer individually to present the most relevant content forward. The hard part is simply stitching it together; and at the scale of millions of potential customers, here in lies the big data challenge.

Deliver a Premium Onsite Experience

Brands that provide superlative experiences and "know" their customers to provide better content quickly will differentiate themselves from the pack.

Think about premium clothing retailers like Neiman Marcus versus mid-level department stores. You expect to have a premium experience. The irony is that online technology can be the great equalizer – no matter if a company is high-end or not. Any brand can provide that experience if it can stitch it all together, and you can bet that the world's largest e-commerce companies are investing millions to "get to know" their customers and give them what they want.

In the world of PLAs, brands have an excellent opportunity to capture more customers, but what are they doing after? Companies have focused gargantuan resources to widen the top of the marketing funnel – attracting as many as possible to a site, but have left the lower part of the funnel too narrow. In other words, they have dedicated money, research and data resources to predicting and optimizing for consumer searches, but have failed to use a data-driven approach to deliver onsite experiences for prospects that match their intent.

Search nearly any long-tail keyword, and you'll see a bad landing page experience that requires clicking multiple times to find what you are looking for, assuming you have the patience to keep looking. It's like paying for signage outside a store to market new blue jeans and then having the customer walk in to find t-shirts on the shelves. The customer then is expected to search the bins until they find the advertised jeans. A significant portion of the targeted audience will leave and won't show up again – and their impression of your brand will be shoddy.

How to Analyze Paid Search Website Experiences

Here are some tests and questions to consider when analyzing PLA website experiences – or any paid-search experience for that matter.

  • What will keep a customer on your page if they don't find what they want? Sometimes the consumer doesn't know exactly what they are looking for after they enter a query. For instance, a customer may search for a "backless black dress" and click on a PLA with a black dress. However, once on the page, the customer may realize that they wanted a cut-out dress, not a backless dress. The web page should contain other content or product suggestions that will help the customer find what they want easily based on what they say and do. Just like brick-on-mortar retailers provide related options next to certain products, web retailers should also think about offering personally relevant options to show customers they have what they want. If not, then they've just wasted valuable advertising dollars.
  • Does a site suggest products intelligently for the business? If it is offering other product suggestions on the page like many retailers do to enhance PLA product landing pages, they should be offered in a way that follows a business's goals. It's important to consider the ranking on the page, using site behavioral data and transaction history to determine if a certain product included is not performing well – then continuously refresh and optimize for this metric.
  • How does a site adapt as its customers change the way they act? Consumers do not act in a static manner, and online marketers should be ready to change what they offer in near real-time based on how they behave on a site. Continuously monitor and optimize for "happiness" metrics like time-on-site, bounce versus conversion rate, and search and navigation behavior.

Answering all of these questions for every possible experience across every paid click customer isn't easy. You would need to know about every product you have and match that to every other possible product on your site, in addition to every keyword and the actual queries that a consumer uses to get to that keyword. Again, it's a big data problem that teams of people couldn't handle.

Also, it's important to remember that you can learn a lot from text ads. They help marketers analyze the language of consumers, which can feed strategies for PLAs. In addition, with Google no longer providing keywords in organic search, text ads offer invaluable insight into organic search, which still represents a majority of the links clicked on by consumers.

Conclusion

This holiday season will likely bring another rise in cost-per-click prices. However, marketers shouldn't forget their post-click onsite experience, since they've invested hoards of time and money into their bidding strategy.

Make the money spent on PLAs and other paid-search experiences create a return by using data to create a relevant and positive experience for customers. Then it really will be a happy holiday season for everyone.