• July 27, 2017
  • Mobile technology has delivered a lot of useful functionality to vendors and their customers enabling the parties to be more frequently on the same page. But the screen size has an inherent drawback; it only shows a very narrow slice of a reality—typically one idea at a time. The problem is especially acute in two key areas, sales and service.

    There’s nothing worse than getting “help” from an app that misunderstands your problem or its role. It can be a source of pain, frustration, or even customer churn that a mobile app can’t deliver help that the customer values and we’d be kidding ourselves if we didn’t see that problem. The advent of good AI and machine learning have done a lot to help vendors to ensure that the sales offers they place on customers’ small screens are the ones that customers actually value.

    The same need exists in customer service—putting the right information before a customer but it might not need AI. However, because the screen will only support one idea at a time it’s critical that the help is, well, helpful. There’s less to wonder about in a help situation and therefore less need for analytics. What is needed is simply step-by-step help as a customer traverses a process.

    Helpshift, an emerging company in the customer service space, has been supporting service processes with an innovative approach to placing help at every step in a service process. This approach ensures the app’s help keeps up with the customer’s need.

    This week Helpshift did something really cool, the company announced integration with the Salesforce Service Cloud and made the integration available on the AppExchange. Importantly, Helpshift FAQ’s are mapped to and managed by the Salesforce knowledge base making it possible to manage Helpshift from the Salesforce dashboard. At the same time, when a mobile user initiates a chat in a Helpshift powered mobile app, it automatically creates a Salesforce case. The case contains all of the metadata and data from the user’s phone saving time and increasing accuracy and time to resolution.

    It’s not hard to create a first level integration with Salesforce and many vendors do this. What’s interesting about this approach is that Helpshift has created an integration that not only satisfies the end user in the moment, but it goes on to work within Salesforce to enable knowledge base updating over the life of the mobile app. From here you can imagine more functionality being added over time. For instance, how long before natural language processing becomes part of the interface?

    Customer self-service was once thought to be the peak of customer empowerment but we soon discovered that self-service systems were, in some cases, the same support systems used by agents but possibly with nicer front-ends. That didn’t work at least in part because there’s a lot of knowledge capital that agents have as employees that customers by definition don’t. So customer self-service had a rocky start and customers who used those systems could easily be discouraged.

    But today’s self help exemplified by Helpshift does away with old style self-service while really empowering customers and if an issue needs escalation, that’s a standard part of the solution. Conventional indirect service channels have been highly successful with better than 80 percent of customers checking them before calling a service center.

    That’s resulted in fewer calls but the ones that get through are often more complex and require the help of a real person. So don’t expect this or any service automation to replace the agent, the real benefits of the Helpshift integration will be two fold. For customers it will provide fast answers in a channel that might have lacked them before. For vendors, it might be another way to limit the high costs of customer service. Given people’s propensity to solve their own problems this should be a benefit to all parties.



    Published: 5 years ago

    Salesforce logoIt was gratifying for me to see the Salesforce announcement about the latest iteration of its SMB service desk product, because it is so in-line with my thinking as well as my book, Solve for the Customer (I know, it’s a shameless plug). While I happily acknowledge that I advise the company from time to time, there is no causal relationship between the book and product, but sometimes, correlation is just fine. This is one of those times when correlation yields validation in both directions.

    Of course there’s a press release and you can find it at because it is not my intention to regurgitate it here. I prefer to focus on one new function that draws my interest and shows the parallels I mentioned, Customer Health Monitor. Billed as a category-first among service providers, the monitor does what I’ve been advocating with minor exceptions. It tracks metrics about customers that a vendor thinks are important and reports on them thus providing alerts that help to prevent churn or attrition.

    FYI, Zuora, another company I advise recently bought FrontLeaf to do much the same from a different angle. This idea is gaining traction.

    This approach amounts to managing by exception. A small company can’t afford the labor or even subscribe to the systems involved in constant customer outreach and this tactic focuses on what evidence shows are customers that need an intervention, perhaps by a customer success manager. All good.

    Now for some nits that need to be worked out—not in the product but methodologically. The big, and for many, hidden issue is knowing what you don’t know i.e. how does a business know what things to measure? An obvious example in the press release is what happens when a customer calls support twice in a month. Is this a sign of trouble or frustration and possibly a churn signal? It could be and the point of an alert is to call for further investigation, which leads to interrogating other metrics to triangulate the situation.

    For example, new customers getting up to speed will likely call in more than established customers so it’s best to correlate frequency with other factors like seniority and possibly also products in use—did the customer just install the latest upgrade?

    There are many iterations of all this and the simple point is that any company will first want to identify all of the situations that need monitoring and develop accurate metrics for them. I call the situations Moments of Truth, things that both vendor and customer care about and that must be addressed, moments of truth. So we must know our moments of truth before the rest of this makes sense.

    We can safely assume we know some of our Moments of Truth but that’s no longer enough. We need to know all of them or we’ll be missing things we can help with and that’s bad because successfully negotiated Moments of Truth lead to bonding which leads to customer advocacy. We really can’t have too much bonding so we need processes that find all of the Moments of Truth and instruments them via tools like the health monitor.

    Discovering Moments of Truth is likely a task for a future product release and probably other products like community and analytics. Using our brains to find the low hanging fruit will do just fine for now but suffice it to say there’s more to be done.


    Published: 7 years ago