Chat + AI Changing the Face of CRM

If tech news is part of your daily routine, at this point you’re relatively sure chatbots powered by artificial intelligence (AI) will dramatically change human life as we know it. Evangelists point to it’s ability to order flowers, chat with your favorite basketball team or even do away with all of your warm blooded friends. However, the reality is that most chatbots “suck”. As investors, when evaluating a new disruptive solution, we look at the incremental lift in value and customer experience. For example, in transportation opening an app on your phone and calling an Uber is an infinitely better experience than calling a cab company to wait on hold and maybe get picked up. This type of analysis yields abysmal results for many of the early applications of chatbots, including some or all of the ones mentioned above. However, the largest segment of enterprise software, namely customer relationship management (CRM), is uniquely positioned to benefit from this truly transformational facelift, and can perhaps be the saving grace for chatbot aficionados across the world. The lynchpin of the chat + AI transformation is mobile, a technology that has actually hurt the CRM experience during this last tech cycle.

While every other industry was seemingly getting a huge boost by the ubiquity of mobile, CRM applications (defined as top of funnel through retention, not just the Salesforce sales platform) declined in efficiency. CRM applications are only as good as the data that is entered into them (garbage in, garbage out), and mobile caused complete and utter chaos. The result is that the biggest and fastest growing segment of enterprise software has the lowest and most rapidly declining customer satisfaction score. To know CRM is to hate CRM in many ways (just pick one sales rep or customer representative at random and ask for yourself). While the industry is truly wrought with inefficiencies, let’s dig deeper on the two primary factors holding customer relationship management back from its full potential today.


CRM Data Leakage. Mobile has amplified the age old problem of data leakage where sales reps don’t input data into their CRM software (Salesforce, Microsoft Dynamics, etc). With mobile, the sales rep CRM experience simply became navigating a desktop platform via a smaller screen. With reps today spending more and more time on the road and not in front of a laptop, the result is even more data leakage. There are many reasons to explain this, but it really boils down to a terrible interface in conjunction with a set of CRM fields that aren’t dynamic to the sales opportunity.

Disconnected Customer Service. With the proliferation of mobile, consumers were able to research new products on the fly, call into customer support desks on a whim, and tweet at companies. Tickets surfaced via email, website forms, phone calls, social media, text message and others. Almost immediately, companies were met with a massive wave of new data coming from multiple channels. This led to a scramble and ultimately a system of large call centers managing customers as tickets, as opposed to relationships (like the traditional Salesforce CRM). Structuring customer service this way disallowed it from playing nice with the applications at different levels of the funnel (namely marketing automation and CRM). As we all know in our new world of digital commerce, customer retention is just as important as the initial sale, and disconnected applications throughout the funnel is most certainly leading to more churn.


As you analyze both of these key issues, the common denominators are interface and data collection & analysis. The two most prevalent trends today are the consumerization of the enterprise (read: chat interface / bots) and artificial intelligence collecting and learning from our data to streamline processes. Before jumping into how specifically these new technologies can solve the issues above, it’s worth a quick detour to understand why this is possible now given both AI and chat have been around for a long time.

A lot has changed since the Dartmouth Conferences (AI) and ELIZA (chatbots) made their debut. Notably, some of the more recent changes will drive the efficiencies in CRM that we need. First, there have simply been some major advances in AI driven primarily by our increasing compute power, including natural language processing and cognitive computing. Additionally, as previously mentioned, the growth of mobile devices has extended applications to beyond the desktop. Lastly, an abundance of open, web-based APIs has created a formula for connectivity that we’ve never seen before. Application interfaces are no longer static and users are not required to go to a specific web or mobile page. The interface, and more importantly the application capabilities, go to where the users live. A great example of this is slash commands in Slack where users can decide when to pull computers into conversations and when not to. Our end result is a mobile ecosystem that is served by a wealth of interconnected applications that are working together and powered by incredibly powerful AI. By leveraging these capabilities, we can make a lot of progress in each of the two issues outlined above — CRM Data Leakage and Disconnected Customer Service.


CRM Data Leakage. By utilizing this confluence of forces, we can reshape the experience by pulling data from sales reps via chat platforms into a reduced subset of standard Salesforce data fields via AI. This CRM can leverage context from multiple applications (Gmail, marketing automation, etc) in addition to signals from the sales / customer service rep to create a smarter and more streamlined experience. For example, in the venture business, we work with multiple different entities — portfolio companies, prospective investments, other VCs, and service providers (lawyers, investment bankers, etc). AI could be used here to dynamically create relevant fields — follow-up on a portfolio company board meeting versus get lawyers an answer on an outstanding term sheet. Additionally, these fields will not be stationary on a platform. Instead they will be surfaced proactively via the monthly flavor of chat platforms (Slack, Hipchat, Intercom, etc). This proactive pulling of data (previously it has always been push!) that is based on a smaller and more dynamic set of fields, completely transforms the CRM user experience and works seamlessly with the suite of applications an organization uses to manage the customer. Garbage in garbage out as it relates to data becomes less relevant to CRM due to automation and smart AI. Potential results include massively increased sales efficiency and accurate pipeline prediction.

Startups attacking the CRM Data Leakage problem will need to initially work alongside the legacy platforms. When they successfully capture the daily attention of sales reps via chat platforms, they themselves become the system of record. Troops, Yoke.io and Workato provide this functionality today, though it’s less clear if this will be their go to market.

Disconnected Customer Service. Marketing and sales functions have learned to work synergistically because the primary software applications are connected (e.g. Marketo > Salesforce), and more than ever, the physical teams are also working together. By utilizing a chat interface that captures multiple channels of communication while being informed by other relevant applications (Salesforce, Marketo, Sprinklr, Twilio, Talkdesk), an organization can start to sync customer service with sales and marketing. The chat interface is important because consumers today live there. The profile of a historical successful customer can be analyzed at the customer service level, and top of funnel campaigns on similar targets can be launched. Current customers at risk of churning or showing strong upsell signals can be easily passed back up the funnel to sales from customer service. Mecca for every single organization should be an interconnected sales funnel that pushes and pulls data across all stages of the customer journey. This will undoubtedly create a seamless customer experience, and complete optimization for the company selling product.

Startups looking to fix the Disconnected Customer Service problem will be the internal platform that powers the text based communication with the customer, and also the internal team. Full integrations throughout the funnel are table stakes! Should they make it a focus, Slack, Intercom and Hipchat are positioned well to attack this market.


Chat is no longer just a person-to-person connection. Chat is now a person-to-person connection with software selectively eavesdropping (by invite) and also joining in to inform the conversation participants when appropriate. All industries will be affected in some capacity by chat and AI. Unlike the mobile revolution where CRM was left behind, this industry is going to be completely uprooted. The incumbents see the wave coming, but in an industry that is so heavily dependent on the interface and system of record, the opportunity and wedge for startups is extraordinary.