Larry Cannell

A member of the Gartner Blog Network

Larry Cannell
Research Director
6 years at Gartner
29 years IT industry

Larry Cannell is a Research Director in the Gartner for Technical Professionals Collaboration and Content Strategies service. Mr. Cannell covers enterprise collaboration and social software, search, content management, and open-source collaboration and content solutions. Read Full Bio

Coverage Areas:

Search and the Enterprise Social Network

by Larry Cannell  |  April 9, 2012  |  3 Comments

Many people have pointed out the similarities between the messaging in email and enterprise social network sites. However, the synergy between search and enterprise social network sites is discussed far less. As enterprise social network sites integrate with more business applications, the similarities between enterprise social network sites and search will become evident from two perspectives. First, from the point of view of how workers will expect to interact with business application information via a social network site. Second, there are significant similarities between how business application information is aggregated into a social network site with how the same is done with search-based applications.

I’ve previously blogged about how enterprise frictionless sharing should become "the norm and what architects aspire to provide". In other words, information about what is happening in business applications should be surfaced within an enterprise social network site. For example, let’s assume that several business applications are integrated with an enterprise social network site. Entities within these applications (e.g., a customer in a CRM system, a project workspace, or a product in a release management system) are shadowed with non-person entities in the social network site. These entities should even have their own profile page, similar to a Facebook fan page except you would follow a customer’s CRM record instead of being a fan of Lady Gaga.

As changes occur to these business application entities (e.g., a customer makes an inquiry, a document is uploaded, or a product is released to pre-production) anyone following the shadowed social network entity are notified of these changes in their activity stream. For example, if I follow customer "ABC" within the social network site, then events involving the customer record "ABC" in the CRM system (e.g., a new opportunity for customer “ABC” is created) generate notifications within my activity stream.

Notifications can also spark conversations. For example, a notification from a CRM system could provide the opportunity for someone to offer additional information (e.g., "Be careful with this customer, I’ve had problems with them"). Or the conversation could be used to resolve an exception flagged by the business application (e.g., "I’m trying to determine why this release was kicked back to testing"). The notification and resulting interaction happening within the activity stream are compelling reasons for implementing these types of integrations.

In addition, these integrations offer longer-term opportunities to build the enterprise social network into a powerful knowledgebase that consolidates information across business applications; much like the role enterprise search can play today.

Think about how our previous example (several applications integrated with an enterprise social network site) will shape worker expectations for finding information. The social network site’s search function should naturally include results from these notifications. While workers may want to search for older events (before the business application was integrated with the social network site), it could be a matter of time before the social network site’s search results contain much of the workers’ needed information. The richness of the information contained within the notifications (e.g., changes to a customer record) and the descriptions reflected in the entity profiles (e.g., a customer’s demographic or contact information) will impact the effectiveness of these social network-based searches.

As multiple business applications become integrated with a social network site, a significant challenge will be the normalization of business entities across applications. For example, aligning a customer record in a CRM system with the same customer record in a warranty claims system. Without this alignment, cross-business application relationships cannot be captured within the site’s social graph and workers participating in the network will see duplicate customer profiles (one from the CRM system, the other from the warranty claims system).

Another important question regarding these types of integrations will be the role of security semantics managed by the originating business applications. Should the social network site enforce the same access control specified in the business application or should a different set of controls apply to this information? In other words, should revoking a worker’s view permission to a customer’s record automatically revoke this access to the entity’s shadowed social network profile? Replicating a single business application’s security semantics with the social network site may difficult. A more daunting task will be normalizing security semantics across multiple business applications feeding the social network site. As a result, compromises may be required, such as creating a new set of access controls within the social network site.

Many of these information aggregation challenges are not new. They share similar qualities to those being addressed by today’s enterprise search solutions (as well as data warehouse applications). Now, I am not talking about the simple keyword search, but rather search-based applications that aggregate information across multiple sources of business information. These applications often offer rich navigation methods for exploring the information and finding previously unseen relationships.

In short, I expect many of the disciplines involved in search-based application integration (as well as those involved in other types of data aggregation, such as ETL and data warehousing) will be important when striving to integrate business applications with enterprise social network sites.

Subscribers to Gartner for Technical Professionals (GTP) can read more about the similarities between search and enterprise social network sites in the report “The Post-2.0 Era: Social in the Context of My Work,” where the concept of a Social Online Workplace is introduced. If your company is a Gartner customer you may already be able to access this and other GTP reports. To see if you do, contact your company’s Gartner Membership Administrator. If you do not know who that is ask Gartner.

3 Comments »

Category: Uncategorized     Tags: , , ,

3 responses so far ↓

  • 1 Peter Wallqvist   April 12, 2012 at 4:54 am

    A very interesting post from our company’s point of view. Our product set, a tight integration between an enterprise search platform and an enterprise social networking tool, delivers a lot of the functionality mentioned.

    For instance, a user can ‘follow’ assets that are part of a particular project or matter and get updates of any changes or additions to these in their personalised ‘feed’.

    This integration between a universal search index of an organization’s assets and an internat social networking platform does indeed yield some very useful business features. Read more about these here:

    http://www.ravn.co.uk/products/ravn-enterprise-connect

  • 2 Samir Adams Ghosh   April 16, 2012 at 4:03 pm

    Great points Larry!

    It seems like the title of your post should be “Applications and the Enterprise Social Network”. Search is a reference when we only envision “application integration” to be an aggregation of a bunch of feeds. However, a Enterprise Social solution should deliver much more than this.

    Social Contextual Collaboration
    —————————————-
    Search is “after the fact” – digging around for what we’re looking for. Email forces us into this “archive and search” paradigm because even though an email sender knows his/her email is about “Order 2384295″, there’s no reliable way to ensure this context is conveyed to recipients.

    A good Social Network solution, on the other hand, should allow the user easily initiate (and thereby automatically associate) a conversation about an original order, so that conversation always has this context. This improves information retrieval and enterprise knowledge management as relevant conversations are automatically organized with the relevant business processes and records.

    For example, a customer service rep should be able to ask Order Fulfillment about the status of an order (Social allows this question to go unobtrusively to the Order Fulfillment department vs. spam email). The customer service rep should be able to post this question inside their CRM app “pinned” to the specific Account record, and the person from Order Fulfillment should be able to respond inside the Order Fulfillment system with the response automatically “pinned” to the corresponding Order record. Thus, the employees can collaborate inside the applications they already use and the relevant conversation and knowledge are organized both by Account in CRM and Order in ERP for easy future recall.

    Application Agnostic
    ————————–
    This means that the Enterprise Social solution must also be application agnostic. That is, using proprietary social features in each application only creates information silos in each application. Collaborative conversations happen across departments, geographies and applications. Your Enterprise Social solution must integrate with any application down to the record level while providing that common communication layer across all applications.

    Follow Record and ActionStreams
    ——————————————–
    Business needs more from “Frictionless Sharing” than Spotify’s flood of songs listened to by all friends. Workers need to prioritize. We need to select about which processes and records we receive notifications. For example, a Sales Manager may be particularly interested in following specific Accounts, or may want notifications only of status changes to Opportunities in Midwest > $100,000.

    For more info:
    http://www.qontext.com/products/overview/why-social-contextual-collaboration/

    Samir

  • 3 Larry Cannell   April 17, 2012 at 9:25 am

    Samir,

    I agree with everything you say about how collaborative capabilities should be available in multiple contexts (such as all the applications relevant to the worker) and the need to select which business application entities we follow. This aligns with my vision for the Social Online Workplace. The challenge for enterprise social software is to provide a worker-centric, yet social environment (and “work” extends beyond my social network and includes the applications I need to use).

    The aim of this particular blog post was to point out the similarities between social network application integration (in particular, event and entity aggregation) and enterprise search. As multiple applications are integrated with enterprise social network sites, this becomes a data integration challenge (similar to those enterprise search solutions have faced for years). This is a critical aspect often lost on those who only focus on the activity stream-based experiences.

    I disagree with your implication that search is bad. Activity streams capture a moment in time. These shared events can improve personal productivity (by proactively bring information forward) and can spark conversations that are exposed within multiple contexts (e.g., within an application, within a social network stream, and others).

    However, for a social network site to improve “information retrieval and enterprise knowledge management,” (your words) the information captured within these conversations and the information about business application entities should be captured within a social graph. Search is the most familiar experience people can draw on to explore a social graph and should be available from multiple contexts as well. There may be other analytic capabilities that can be used to leverage the social graph as well, such as recommendations, exposing “more like this” results next to an application record, or within an email thread.

    Larry