Recent studies, one by Sentiment360 and the other by FreshMinds, concluded that social media monitoring tools aren’t very accurate in automatically measuring sentiment, much less influence. The insight isn’t a new one and speaks to the now well-known issue of whether social media engagement is scalable. Consequently, we now see social media monitoring companies combining with text analytics companies to bundle their service offerings to increase their ability to monitor a customer’s activity and online influence, tracking that information to the workflows of marketing, sales, customer support, or operations in near real-time.
How well hybrid analytics companies, combining social media monitoring with text analytics, can deliver on the automation promise, and scalability, in managing the customer experience remains in question. For example, in attempting to convey the limits of the marketing promise, Maria Ogneva of Attensity360 in The “Right” Degree of Automation recently offered the following distinctions, between process automation, response automation, and pre-response automation.
Process automation involves developing rules to use in decisions about the flow of information. Response automation involves using automated and “canned” responses to customer questions, generally a “no no” in social media unless tied to an information update rather than a marketing message. It is worth noting, as Maria’s colleague Michelle de Haaff does, that response automation also includes automatically determining which social media messages merit engagement and which ones don’t. The whole SCRM discussion needs independent research on these new hybrid tool sets to assess their degree of accuracy over the existing automated sentiment analysis tools.
I don’t think it is too soon though to assert, following Mark Tamis recent point, that the importance of collaboration across the Enterprise and its ecosystem is crucial to SCRM. It isn’t as simple as training people to collaborate, as some imply. Rather, a learning organization and the culture that goes with it are crucial preconditions for employing analytics effectively in SCRM, especially if business processes and work practices are to deliver customer experiences seamlessly.
The thoughts Maria shared about pre-response automation are key to our discussion here.
Somewhere in between process and response automation there exists another kind of automation. It’s a hybrid of sorts, let’s call it pre-response automation. What in the world is pre-response automation? Well, I did just make up the term, but bear with me – let’s see if we can make it catch on. Your system reads, understands and distributes social media messages in step 1. Then taking it a step further, it looks up a potential answer from either within your FAQ or an external user forum, and queues it up as a potential answer for the person who should be sending this message. This way, you as the company rep, get to send a message that’s automated and personalized at the same time. The thing you are automating is the research that would take you time to look up – time you would’ve spent on a menial task that could be spent on engaging and humanizing your responses. Imagine how many more customers you could talk to then! As long as you are putting human touches on all of your messages, using automation to help you write the straightforward response is A-OK. Of course this only works for fairly straightforward cases, nothing custom or complex. Then there’s no shortcut around research (my emphasis).
Maria’s distinctions about how to apply text analytics in fine tuning social media monitoring to engage the customer experience are well put. In fact, as Lior Arussy recently noted, the more Social CRM advocates promise automation as a feasible choice for meeting the scalability challenge of social media for businesses, the more their consulting strategy mimics traditional IT consulting where the technology, though claiming to only provide a part of the solution, is actually assumed by clients to provide THE solution. As a result, crucial organizational and cultural challenges too often go unaddressed.
…we should not rush to embrace new technologies, when we lack the substance to initiate the customer engagement. A fan club on facebook or constant tweeting will not disguise inferior customer experiences. In fact it will only magnify the problem and distribute it to millions of potential new customers.
At the core of social CRM success must be not the tools but the organizational readiness to act. Both through executives’ readiness to listen and commitment to act combined with design and delivery of superior, differentiating experiences.
In his comment to Lior’s post, Marc Mandel observed that ” in my experience the fault about trying to substitute a tool for a truly appropriate organizational solution is neither the exclusive domain of the buyer or the seller, but often a shared culpability.” To her credit, Maria Ogneva of Attensity360 straightforwardly notes that analytics and monitoring tools cannot substitute for a business strategy.
How can we keep the people and culture challenges in organizational focus while deploying analytics in SCRM? As Christian Finn, Microsoft’s Director for Collaboration and Enterprise Social Computing, recently noted regarding Microsoft’s use of Sharepoint 2010, “Solve a Problem, Don’t Deploy a Technology”. To get more specific, ready the organization to solve bumps in the customer experience in a seamless way first. A good customer experience can be delivered without SCRM technology, as the video below by Jaffe Juice makes clear in relationship to an experience with Starbucks and Foursquare.
Frequently a seamless customer experience will need delivering without SCRM since the customer’s job demands application of a barely repeatable process. Or, as Sig characterizes barely repeatable processes over at Thingamy,” The activities that employees spend most of their time on every day”.
In other words, Brands Don’t Talk to Customers, Employees Do. Organize for collaboration accordingly and emphasize empathy as well as information sharing.
Posted by Larry R. Irons
Share this post…