One of the eternal problems in a call center is getting an enquiry routed to the right agent. This is doubly true in a mobile world that demands conversational support in near real time. Add to this the fact that most customers seeking support already failed to find an answer to their inquiry using an FAQ, a community, or other self-services. In this situation customers expect an answer within few minutes, if not seconds. On top of this, the ability to make engagements with the company easy, efficient, and ideally joyful, is becoming more and more a distinguishing factor for companies. Customer experience is the result of engagements, and for humans the experience gained from the most recent engagement tends to have a higher influence than older ones. Consequently, a positive customer experience matters, not only during marketing- and sales, but even more so in situations that require active help of the company that sold the product or service. So, getting a solution to an issue must be as easy and as human as possible. The challenge is that every support organization needs to live and work with limited resources – human as well as technical ones.

It is Like a Good Game of Soccer

11 players and a ball. There is a goal keeper, are defenders, midfielders and attackers who play as a team against their opposition, trained by their coach and guided by the captain. There is a core team, and some players may be assigned to different roles, even within one game. Depending on the opposition team, the coach and the captain change player assignments, tactics and roles. Together they play the ball with the objective of scoring. Like a soccer team call center agents are organized around their strengths and like soccer players the agents are – or should be – carefully trained and assigned to roles on their field: the queues with which incidents are managed. Like the soccer players, agents can be assigned to different positions on the field: different queues. The incident is the ball and the scoring is resolving the incident. With all its intrinsics, soccer is a simple game. It has a few rules that dictate the do’s and don’ts. Following these rules, considering available players and their strengths, is the table stakes. Yet, this doesn’t make for a game that the spectators may find exciting – nor might it lead to scoring. There is an art to it, the art of how and to which player to move the ball in any given situation. This is the same in a call center environment. Yet, in some call centers, when receiving the incident, the team stops and asks the spectator where to start. The user is asked to qualify it and to route it into the right queue. This is a task (s)he is obviously ill prepared for. Doing so would lead to a heavily disrupted game that no one would like to watch, not even talking of enjoying. Scores would be rare and far in between. In a call center environment, this leads to unnecessary load. Load that is caused by re-routing the incident, delays in handovers and resolution, and ultimately frustration on both ends: The end user and the service agent. In brief, it is causing a poor customer experience, following another poor experience. There must be a better way. And there is. Back to our in-app support world. Why not using the incident itself to automatically identifying the right queue and smartly routing it there? This is like the soccer players using the ball’s momentum and the given situation to proceed to score. There are two basic possibilities to achieve this smart routing. One I would name the classic approach, the other one the text mining approach.

The Classic Approach

Automatic routing to the right queue can easily be achieved by using metadata that is provided by the app as part of the incident report. Incidents are tagged based on this metadata. The tags are used to route the incident to an appropriate queue or agent. A configurable automation does both, tagging and routing, without explicit intervention of the app user, who is already in distress enough as there is an issue that he or she couldn’t resolve. This is a proven approach that requires an elaborate design of metadata and its collection within the app and the ability to build rule sets on top of this metadata in the service back end. With sufficient data and advanced analytics tools this approach can also be used to offer pre-emptive support, which further improves the customer experience by avoiding the negative experience in the first instance. This approach, however, finds its limits in the ability of instrumenting an app so that enough situations can be identified from the metadata for appropriate tagging. Not in the least because instrumenting has an adverse effect on the app.

The Text Mining Approach

With more and more service center software running in the cloud, massive improvements in AI, Natural Language Understanding (NLU) and machine learning technologies, additional information can be used: The description that gets submitted by the user as part of the incident report. Albeit the users do not always provide accurate information, this is regularly enough to augment and improve the tagging, therefore getting a better routing and ultimately faster resolution. Given that in many systems the first line service agents can resolve incidents by themselves consulting a knowledge base the AI might be able to resolve the issue before escalating it to an agent.

Smart Routing is the Future

This paves the way. Currently leading systems smartly queue and route incidents based upon an elaborate architecture of metadata and tags that are automatically given based upon the metadata. This already tremendously helps companies in providing good service. The next step is including the descriptions given by the users into the mix and to use it to suggest solutions to agents until confidence is high enough to have the software independently suggest solutions to the users or to route the incidents. NLU in combination with machine learning / deep learning are nearly at this stage. In either case, smart routing is reducing the friction in the support process and therefore improving customer- and service agent experience. The result of this concentration on positive experience is a measurable benefit for the business – because of adding value for the customer.  

In a mobile world, where the smartphone has become the command center of our lives support needs to be offered from directly inside the app, using in-app messaging. This way the advantages of being able to send relevant contextual information about the state of the app to the service agent and the ability to engage in a service conversation via a conversational UI can get brought to full advantage. The user is identified, relevant information has been gathered, which the service agent can use right away. This leads to capabilities that a genuine mobile in-app support system needs to have on top of generic help center functionality:

1. In-App FAQ that gets pushed out to the phone and is available in an offline scenario
2. Collation of meta data about the phone, user and the incident that created the support call, along with the ability to send that to the customer service center
3. In-App messaging/conversational UI in combination with push notifications
4. Automation to properly route incoming issues and to increase the issue resolution efficiency
5. An ability to integrate into CRM- or other systems
6. An ability to selectively and proactively engage with users, to e.g. support onboarding or push notifications about special situations to relevant parts of the user community.

It is possible to find vendors that deliver parts or all of this in order to deliver a mobile service experience. Platforms like G2Crowd, but also traditional analyst companies like Forrester and Gartner give some leads. Gartner lists Salesforce, Pegasystems, Oracle, Microsoft, Zendesk as leaders in customer engagement centers, with SAP being the only Challenger and Lithium the only Visionary. None of these companies gets high marks for in-app support, though. So, we are apparently looking for specialist vendors. Concentrating on the bigger medium and enterprise end of the market I’ll briefly look over the following companies, which offer parts or all of the desired solution.

• Apptentive
• Freshdesk/Hotline
• Helpshift
• Intercom
• LivePerson
• Salesforce Service Cloud and com
• Zendesk

Salesforce and Zendesk are part of this list as in-app service specialists see them in sales opportunities as competitors.

Vendors

Apptentive

Apptentive focuses on mobile customer engagement to foster a better way to communicate. The solution offers in-app messaging, rating prompts, notifications and surveys. There are no FAQ’s, though. The company got founded in 2011. The solution integrates into Zendesk, Uservoice, and Desk.com, which can be used to build a more complete in-app customer service solution in combination with Apptentive. Further integrations should be possible using the company’s API. Apptentive gathers meta-data from the mobile apps and presents it via its reporting solution. This data then can be exported as CSV and imported into a CRM system to enrich it and to provide for segmentation. However, there does not seem to be a campaigning functionality inside Apptentive itself.

Freshdesk/Hotline

Hotline.io, the former Konotor, is the mobile in-app support platform of Freshdesk. Konotor got founded in 2012 and was acquired by Freshdesk in November 2015. It is a native in-app support platform that fully caters for the criteria mentioned above. Hotline positions itself more like an engagement platform than a customer service platform. Being a part of the Freshdesk group their focus is more on the upper end of the SMB arena. Their clients mainly consist of startups and bigger corporations (as the smaller ones often do not have mobile apps, according to Srikrishnan Ganesan who heads up Hotline). Freshdesk itself seems to be sales-driven (which is not a bad thing). The company recently won the 2017 CRM Watchlist award, which is an impact award given by Paul Greenberg, which gives high visibility. Hotline has built-in integrations e.g. with Zendesk, and Slack and, of course, with Freshdesk itself. Using APIs it is possible to integrate with other CRM- and customer service systems. With the help of the Freshdesk platform it turns into a multi channel support system. The campaign module that is built into the back end allows segmenting customers based upon meta-data, app usage, and other criteria and allows for simple campaign automation.

Helpshift

Helpshift seems to be the leading mobile in-app support platform with a base of about 1.3 billion installations. The company got founded in 2011 and pioneered the idea of in-app support, providing a native platform that fully caters for the criteria above. The company positions itself as revolutionizing the customer support experience on mobile – with a strong focus on customer service. The target of Helpshift is enterprises and bigger call centers, and their number of installations proves that they are up to it. In 2014 Helpshift won the CRM Idol competition and the company is probably a bit of a hidden gem, which may partly be attributed to their enterprise focus. Helpshift has a deep integration into the Salesforce Service Cloud and I personally expect integration into MS Dynamics looming, as Microsoft Outlook Mobile is a key customer and Microsoft itself a leading investor into the company. Being API-driven integrations into other CRM systems are easily possible. The backend itself is multi-channel capable. The Helpshift campaign module is geared for service rather than marketing and leverages meta-data for segmentation, easily allowing for the facilitation of onboarding or recovering abandoned baskets, and getting over the early retention barrier.

Intercom

Intercom positions itself as a platform with support for live chat, marketing and support, with all part solutions being seamlessly integrated via their platform. The company has a strong mobile focus. A lot of their customer base seems to deploy the solution as a customer service solution. As Helpshift, the company was founded in 2011. Their solutions cover the criteria above, with the exception seeming to be the in-app FAQ. FAQ’s seem to be published through the web, instead. Their analytics offer the nice feature of identifying candidates for improved or new knowledge base articles, based upon searches. The solution offers multi-channel support and integrates into a good number of communications channels, and into Salesforce, to have a strong support system. They have deliberate support for onboarding processes, helping customers to get over the early retention barrier. Using data collected from the devices a segmentation and hence a targeted communication with users is possible, for example the onboarding support.

LivePerson

LivePerson offers a mobile in-app support solution through its LiveEngage platform. The company sets out to drive conversations and loyalty through the platform, and does not position it as a support- but an engagement system. LiveEngage is an ‘upgrade’ from the company’s earlier LivePerson solution, which seems to have caused some grief for customers. This ‘upgrade’ seems to resemble a pivot into the mobile in-app business. They are targeting enterprises although they also have startup customers. LivePerson is in the chat business for around 20 years. Focusing on messaging they do offer a chat platform that can be integrated into apps and web sites. They do not offer knowledge base support of customers but rely on custom integrations into CRM systems. Their key asset is their long standing history along with a platform that can turn chat into action, albeit more with a focus on cross-selling. Their current challenge is their need to manage their pivot and some customers that lost functionality that was important for them.

Salesforce

Salesforce offers two customer service solutions, the Salesforce Service Cloud and Desk.com, addressing different customer sizes. Desk.com is geared towards smaller organizations where the Service Cloud is enterprise scale. Both solutions offer multi channel support capabilities. Gartner Group, in their 2016 Magic Quadrant for Customer Engagement Centers states that “mobile chat” (their term for in-app messaging) is best handled by partner solutions that are available via the AppExchange market place.

Desk.com

Desk.com is Salesforces “Out-of-the-Box helpdesk for small businesses”. Being a member of the Salesforce family the solution has considerable grunt and exposure behind it. Desk.com is not a native mobile in-app support platform but a general support platform that allows for embedding support functionality inside mobile apps via an SDK. The SDK allows for the creation of tickets from within the mobile app and for knowledge base search with the knowledge base not being inside the app itself. In contrast to other providers Desk.com does not cater for conversational in-app support. The main support experience is provided outside the app. Being a member of the Salesforce family integration into other family members is a given, since they share the same platform. This includes a connection to the Sales Cloud that facilitates the conversion of support cases to sales opportunities. But at its root Desk.com is a helpdesk solution that got extended to cover mobile apps.

Service Cloud

The Service Cloud is the Salesforce customer service offering for enterprises. As such it offers a full fledged service solution that also extends into mobile devices via e.g. ticketing live chat, knowledge bases and support of communities. All this takes advantage of the powerful Salesforce platform. There, however, is no direct in-app support, which makes a good case for the integration of Helpshift.

Zendesk

Zendesk is a leading cloud based customer service platform, offering multi channel support. Zendesk features as a leader in the Gartner 2016 Magic Quadrant for Customer Engagement Centers. Zendesk is not a mobile native. Their chat widget integrates into web pages and the company does not offer in-app chat. Instead the company offers solutions that hook into existing messaging apps like Facebook Messenger or Whatsapp. They offer knowledge bases, support of communities, and portals. Zendesk has integrations into many other systems, including MS Dynamics and Salesforce. The support for messaging apps, besides Zendesk being competitively met by mobile in-app vendors, is the main reason for them appearing here.

What It Means

The in-app service market is a specialist market. If a strong support for mobile apps using a conversational interface and an in-app experience are required, not many players with the strength to support larger enterprises are left. This can be seen in the table below, which covers above solutions and is enriched by some generic help center functionality. Mobile In-App Support Provider Overview Zendesk and Salesforce both offer partner solutions to fulfill in-app service requirements. By themselves they are not in-app support players. Apptentive focuses on rating systems and surveys and is not a full-fledged player as well.. Intercom positions itself more as an engagement platform although customers seem to deploy it as a support platform, too. Without an own knowledge base capability it, however, misses an important capability. This leaves us with three vendors: Freshdesk, Helpshift, and LivePerson, with LviePerson currently pivoting from being a web based chat provider to an in-app service solution provider, which seems to cause existing customers some grief. The solution is also missing an in-app FAQ. Helpshift and Freshdesk as the front-runners are running very different strategies. For Freshdesk in-app support is one of their capabilities that they acquired over time and made part of their solution set. Helpshift totally focuses on in-app support, believing that mobile will be, if not already is, the dominant channel, which needs specialized support functionality to offer maximum service and experience.

Following some of my posts on AI in customer service environments I got contacted by Daniel Doctor from Tenacity who invited me for a chat with Ron Davis, the founder and CEO of the company. Which I had. And it was an interesting conversation. In my articles I spent a lot of time focusing on how AI, machine learning and chatbots can help improving both, the customers’ and the service agents’ experiences by making sure that all relevant data is collated and available, reducing wait times for customers, being able to already suggest good solutions to both, customers and agents, and so on. The objective is at all times to have the customer get a good solution as frictionless as possible and to enable the service agent to concentrate on the hard jobs. The idea behind this approach is that it reduces customer irritation by having the answer faster and improves the agent situation by making the work more attractive. After all, who of us loves dull, boring and repetitive work. Not many, I bet – certainly not I. Of course, this is only half of the truth. Service agents, like all employees also react strongly on who they work with, who they work for, whether they have the right tools at hand to get their job done, how their stress levels are, whether their private lives are untroubled, whether they have enough sleep, and so on. Additionally, the more interesting situation of the dull jobs being taken care of by the machine creates stress, as the customers tend to already have an elevated level of frustration that was caused by the unsuccessful solution attempts they already went through. In this situation the best-suited agent is also a different one than the one best suited for first customer contact. It is the controller as opposed to the empathizer. This, to some extent, is outside the scope of employment, but it still is a good idea for employers to equip their employees with the education, tools and techniques to deal with stress and to create a healthy social environment. This is where Tenacity comes into the picture. The value proposition of Tenacity is to “reduce attrition and absenteeism with an employee engagement solution that will give your call center agents a better quality of life”. The philosophy behind it is that happy employees create happy customers. Tenacity is tackling this with a solution set that bases on three levers:

• Stress management
• Creation of a community of co-workers
• Creating a sense of meaningful work

And Ron claims that Tenacity has reduced attrition by 1/5 to 1/3 in every(!) call center they are in. Which is an outstanding achievement. They have done this by delivering a combination of wellness content, social network and online community that one can imagine as similar to what Reddit looks like. Now, if you are looking at supporting and promoting the reduction of attrition and absenteeism as mainly an HR function, then you are probably right. However, as all this is about call center performance it is also very much an operational topic. It starts with persons – call center agents – not simply accepting resilience content and other things that are pushed into their faces. Instead this needs to be personalized and be made part of a system. Tenacity wants to be this system by tackling the above three levers. Agents log in to the cloud based system and are presented with a Reddit-like community where they can interact with colleages and are presented with a dashboard that gives them an overview of their achievements in terms of health goals. The whole system shows gamification elements to encourage usage and allows not only to earn points, but to also meaningfully redeem them. Companies offer the agents some time in the application and they can voluntarily spend additional time in there. The system is backed by a decision tree based AI that learns from interactions in the team and gets data feeds from performance management-, WFM-, and HR systems that are uploaded into the system. Ron tells me that they are looking into moving towards deep learning once their data sets are big enough, which they aren’t yet. Being a young system there is no direct integration into other systems yet. With this data the system then can suggest a number of things, like where team-building exercises should happen, or how teams could be organized to improve the overall performance, or how an individual agent can improve the own situation.

My Take

Tenacity wants to re-humanize the call center by using data. This sounds like an oxymoron but depending on which data gets used this seems to be a good way. Their system currently uses organizational-, WFM-, and PFM data, in addition to the behavioral data that it collects. Considering that a person’s motivation level also depends on whom she works with and for, this is clearly a good start and there are plenty of success videos on the site. The idea of combining resilience content with community, backed by data to allow decision-making based upon facts could be a winning strategy. Where I am somewhat of split minds is the merits vs. disadvantages of having the system separate. While this reduces distractions it also reduces the potential benefits because the system just might be a click too far away. Knowing that Tenacity is a young company with a new system and limited resources I do think that integration is key. Integration into the above-mentioned systems, in particular WFM and PFM, but also with a Learning Management System and the customer service systems. I do also think that an integration with the actual call center software would be beneficial to the agents if used for helping them determining stress levels and potentially doing some load balancing. Tenacity brings an interesting concept into the market but needs to make its stand against the call centre solution-, performance- and workforce management system providers who all have the objective of making the call centre more effective and efficient, albeit with different approaches. So far signs are positive for them, but as usual only time will tell.