When content became king: Marketers know what you want before you know

Marketing today bears no resemblance to the Mad Men era. Powerful analytic tools are parsing companies' web sites and content. Machine learning tools are studying buyers' behaviors to generate recommendations for even better web designs.
Written by Brian Sommer, Contributor

When did machine learning and marketing get together?

A couple of years ago, there was a pronounced shift in the way businesses wanted to buy software, goods and services. Buyers borrowed from the activities they did at home or on mobile devices when shopping for consumer goods. They applied these research tactics to the purchase of business or industrial products. What has happened is that business buyers use the Internet as the place to "research" future purchases and suppliers.

The big implication of this is that smart marketers must place significant amounts of "content" on the Internet today. Content (whether it be product slicks, short role-based videos or educational/thought leadership pieces) is what buyers need to decide which providers belong on the long list, which of these potential suppliers will make it to the shortlist, and, how will the buyer decide which solution best solves their business problem.

Sales professionals and marketers who believe that businesses still want to be bombarded with large volumes of unfocused sales collateral or demonstrations just don't get it. Likewise, when it comes to the sale of the ERP software, no one has the time or patience to sit through a 3-5 day function/feature marathon anymore. If people are willing to do any kind of binge watching today, it's for their favorite cable TV shows not software demonstrations. Smart marketers put their most important content online for prospective buyers to find, read and evaluate at their leisure.

But this is a relatively new shift. Marketers have been scratching their head guessing as to what content prospective buyers might need, want or value. The how do they decide?

A little over a year ago, I highlighted some of these marketing challenges in a ZDNet article, and how one company, BrightEdge was tackling the problem.

"Jim's firm looks at a number of data sources to figure out the questions, issues, concerns, etc. that people have in the market. BrightEdge looks at what sort of person is scoping out your website and your competitors' sites, too. They might also look at specific discussion groups or social networks where large numbers of your buyers hang out online. They look at the kinds of content these folks are searching for. They use social, mobile and search data to supplement their analysis. In the end, they can suggest an editorial calendar of sorts for the Marketing organization.

What's brilliant about this is that this is the material that buyers are looking for - no SEO needed here as the high number of hits on this material alone will automatically move this information to the top pages of search engines. Buyers will find your firm as they will be finding your assets more and more frequently.

Today's buyer wants to do business with a firm that understands them. If you're not communicating with them, they think you don't understand them. Modern Marketing is about creating very efficient and effective communication channels with prospective buyers. Guesswork has no place in a ruthless, competitive market.

There's more to it than this though. BrightEdge also possesses tools to track the topics requested, how the information is being distributed, which items are most desired/valuable/translate into more sales, etc."

BrightEdge has continued to innovate and has made some recent announcements of merit. But I'm getting ahead of myself. Let me talk first about another company: HighSpot.

HighSpot looks at the content problem slightly differently. In a conversation with their CEO, Robert Wahbe, he explained how businesses actually possess large amounts of content, but that marketing often throws it over the wall to sales without proper context. Sales representatives don't necessarily know which content that prospective buyers are receiving or appreciating let alone whether these prospective buyers are finding it relevant to their business.

Image Courtesy of Highspot

What HighSpot is doing is to understand what content prospective buyers are reading in great detail. For example, if content is placed on a dedicated webpage, then HighSpot can identify which portions of content the prospective buyer is lingering on. This serves as a signal as to the importance of this particular information versus other content within the same online brochure, slick, newsletter, etc. Understanding what portions of the content that readers are actively engaging in helps a salesperson understand the critical business problems a perspective buyer is trying to solve. As a result, sales representatives can craft a more relevant follow-up materials and phone calls with these prospects. The software, according to Robert, "Helps, not bugs, people".

HighSpot has only been on the market for 10 months yet already has 100 customers. The technology is already integrated with Salesforce. It uses a mix of data science and machine learning technology to improve sales effectiveness, as well as to deliver the right content at the right time. They also have a mechanism that helps marketing organize related and focused content into various pitch books. These pitch books can be placed on dedicated landing pages and are what sales professionals use to communicate with prospects.

Image Courtesy of Highspot

There are interesting parallels between this type of marketing technology and that used in the relationship/dating websites. Those "compatibility" algorithms on dating websites often pay more attention to how long a person lingers on another individual's picture versus what people overtly state as the ideal characteristics of their desired ideal mate. So what HighSpot is doing is adding another dimension or insight to the salesperson's toolkit. Even though a prospect may have said that they were interested in X, Y and Z functionality, the fact that they spend a lot of time reading about a different collection of functions and features suggests there may be some other sales factors to consider.

Let's return to BrightEdge.

When I recently spoke with Jim Yu, CEO of BrightEdge, we talked about how the content phenomenon has been exploding in the marketing world. He told me that as much as 70% of the content placed on the Internet is not used or seen in any appreciable way. He added that many sales marketing people have an unintentional bias towards their own content, data and sites and rarely know what content from competitors is working or working better than their own.

What he was referring to is that prospective buyers use search strings on Google, for example, and do their own homework. As a result, buyers rarely start their searches by putting in the name of your company. They're looking for answers to specific issues or thematic matters. Vendor names frequently are not part of that search. As result, content that is unfocused or not engaging rarely pops up in the first page or so of search results. The answer to this is not a search engine optimization (SEO) game, but rather in understanding exactly what it is that buyers are searching for and how well your competition is already providing those answers.

BrightEdge has built a data cube that understands the content that your firm and your competitors have placed on the Internet. This data cube is the mother lode for analysis by its two main products: Content Optimizer and Page Optimizer.

BrightEdge helps Marketers understand what elements of websites in their vertical or other segmentation slice are working with specific prospects and which are not. Rather than being limited to simply testing the performance of two of their own sites (an A/B test), Marketers can see what is/isn't working across all of their competitor's sites as well. This knowledge helps Marketers create/optimize web sites that will drive more sales.

Image Courtesy of BrightEdge

What should go on these optimized sites? The DataCube and Content Optimizer identify search terms, social signals, rich media and other items that are being used to find competitors' websites/content, engage with competitors and drive commerce. The Landing Page Optimizer helps Marketers create winning content landing pages. Machine Learning is also used. According to BrightEdge:

"Landing Page Optimizer leverages sophisticated natural language processing to identify all the headlines and CTAs across the web. Additionally, search volume trending capabilities spot trends to see what works. BrightEdge's predictive analytics offer data-driven insights and recommendations to take a company's landing pages to the next level."

Optimizing these landing pages can be done across multiple cuts: local vs. national markets, desktop vs. mobile devices, etc.

Image Courtesy of BrightEdge


Marketing today bears no resemblance to that of the television show Mad Men. Today, powerful analytic tools are parsing apart companies' web sites and content. Machine learning tools are studying buyers' behaviors to generate recommendations for even better web designs.

Think about it, when you and others enter a simple query into a search engine, a flurry of machine powered introspection and analysis is underway. Your query can trigger the creation of new web content or cause firms to re-title, update or enhance their web marketing content. And, should you click on this content, some clever bot is studying how and what you read. These inputs are helping marketers and sales professionals understand what your firm really wants even if you don't know or didn't want to tell them. Progress?

Editorial standards