Commentary

How Search Works With Social CRM

Search technology: Some companies will license it, while others build it from scratch. It depends on the egos of executives working at the company. Real-time search and social media have pushed technology to the forefront. Companies need sophisticated algorithms that can sort and index structured and unstructured data.

A recent Accenture report titled "Social CRM: The New Frontier of Marketing, Sales and Service" ties it all together. Joe Hughes, senior executive from Accenture's customer service and support business, confirms that enterprise companies have begun to build search engine technology that will integrate into software applications and consumer hardware to help marketers, advertisers, agencies and others sort through the mounds of data created by social media.

Hughes defines social CRM as the conversation data from social media networks. And as marketers continue to try and make sense of the mounds of data flooding in from real-time search, Twitter streams, Facebook status updates, and behavioral targeting tags, they will need a faster method to sort, index and access data. Wow, are you overwhelmed yet?

Marketers need technology that can move feedback from customers and call center agents between channels with as much automation as possible. That will become the only way to analyze the data. Natural language query processing will also become a focus, to search through documents of unstructured and structured data as the mounds of social media data continues to mount.

Last year, tools measuring buzz metrics in social networks emerged. This year, the focus turns toward integrating the social data into traditional CRM platforms from companies like software-as-a-service (SaaS) provider Salesforce, which late last year integrated the feature, allowing people to search on that data in real time.

Until now, CRM packages did not allow marketers to view data collected on Twitter alongside traditional queries. But the real-time search movement has sent companies looking to improve search results back to the drawing board to build engines that can process structured and unstructured data, as well as sentiment analysis, taxonomy, classification and entity extractions, according to Hughes. "The strategy of combining structured and unstructured data will become more important," he tells me.

Attivio, a company specializing in enterprise search, has developed an engine that can simultaneously analyze and index structured and unstructured data. In real time, the search engine identifies buzz in Twitter and connects with a CRM package to identify important trends. The technology comes from the same folks who founded Fast Search & Transfer, which Microsoft acquired in 2008 for about $1.2 billion.

Hughes tells me Attivio's technology reduces the time needed for identifying trends through Twitter or blog data from weeks and months to hours and minutes.

But Attivio isn't the only company working through the problem of sifting through mound of social data. The issue perhaps prompted companies like Hewlett-Packard and SAP to build search tools into software and hardware application. Some companies have begun to build site search features into databases that can more quickly index data.

Hughes speculates the HP patent on intent-based search engine technology, which GoRumors describes in a recent post, could integrate into consumer electronic devices like laptops or mobile phones. Now I'm thinking perhaps Apple could use this type of search engine for the iPad tablet launched Wednesday. "There are all kinds of reasons for HP to have a patent on search," he says. "This is the year to bring everything together."

Engineers at enterprise applications provider SAP have also been spotted in search engine forums asking about search engine technology. It's not likely the company will launch an engine to compete with Google, but it might build the technology into products and databases that will allow people to more quickly sort through the piles of data. Hughes agrees that this year marketing needs to connect with product development and focus on building a bridge from social networks to CRM tools that manage buzz and products.

Interesting insight: The Accenture report suggests the evolution of social networks and online communities has led to a collapse of the marketing funnel. Hughes writes that traditional mainstream marketing forces a message through the marketing funnel, moving through stages from building awareness to earning loyalty. It only allows dialogue and relationship building as the prospect or customer progresses through the funnel. By contrast, the digital revolution, and particularly social media, make it possible to engage in a dialogue with prospects or customers much earlier and at many more touch points. This creates a lot more data that marketers will need to search through.

Old and New World

4 comments about "How Search Works With Social CRM".
Check to receive email when comments are posted.
  1. Marc Engelsman from Digital Brand Expressions, January 29, 2010 at 10:21 a.m.

    Think the graphics say it all. And this will be helpful in reinforcing with clients why they need to be integrating search optimization, paid search and social media marketing under a "findability" umbrella and then integrating their findability efforts with their other online/offline marketing tactics.

  2. Avi Rappoport from Search Tools Consulting, February 2, 2010 at 2:34 p.m.

    I agree with the point of the article, and hope more strategic managers will learn from it.

    One issue is that various CRM packages have come with terrible implementations of text search, often licensed from Verity or Excalibur or Autonomy. I know some people gave up on search because of that.

    But I hope no one is contemplating writing search from scratch these days. It might be easy to create an inverted index of words and search it. But beyond that there are hundreds tiny technology and policy issues, from what to do with umlauts to autocomplete to spelling suggestions to search log analysis tools.

    There are plenty of effective and fast search engines with good APIs out there, both commercial and open source. They are fine tools and far better than any old OEM package.

    Avi

  3. Jeremy Bergstein from TheScienceProject, April 18, 2010 at 6:35 p.m.

    Fairly strong top to bottom assessment of the platform. The big question will be whether we can find strong enough CIOs, training professionals and visionaries to be able to roll out and implement revolutionary platforms such as Attivio within a complex brand structure and business organization.

    As with every piece of disruptive structural software, getting the organization to adopt the platform and the users to change their behavior is 95% of the battle.

  4. Gregory Yankelovich from Amplified Analytics Inc, April 19, 2010 at 11:30 p.m.

    There are too many terms are thrown around to discuss a relatively simple concept of translating unstructured data streams of human communications into structured information digestible by enterprise systems and processes.

    The context in which this task is undertaken is often disregarded in favor of serendipitous methodology of "I know what it means when I see it". It is like going on search of a needle in a hay stag hoping to find a farmers daughter - good luck, but probability of success is not very high.

    We decided to focus instead on what Customers say about the products or services they purchased and measure the delta between between their experiences and their expectations. These findings are very specific, accurate, consistent and actionable as they can be used to fix problems and improve customer experiences systematically as oppose to fishing for illusive influencers and trying to placate them in hopes of their positive WoM. I think this approach can produce real life ROI.

Next story loading loading..