JASON REDLUS: Will you give us an overview of Coveo and some examples of how the company helps its clients?
LOUIS TETU: Coveo is an Insight Solutions technology company. We serve customer service organizations as well as engineering and sales organizations within enterprises. We provide technology that enables users to gain great insights into the vast and diverse amounts of information that reside within, outside, and around organizations. These insights enable our clients to serve customers better, sell more products, and foster innovation.
There is a fundamental theme we hear at almost every company. Companies are looking to gain more answers while facing uncertain situations. For example, consider the area of customer service. Increasingly, products are getting more complex, and customers expect more knowledge on the part of their organizations to solve their issues. Coveo’s Insight Solutions pull together all the information residing within the various technology tools and organizational silos, thus providing the users with a greater ability to take action. When troubleshooting a complex customer issue, there could be ramifications into engineering and product information. The best answer to a specific problem might be found in the customer community. The ability to pull together all that information and make it available and highly conversational for the customer service agents, or for customers themselves on a self-service basis, helps drive efficiency and effectiveness within customer service. It saves costs and increases both effectiveness and customer satisfaction at the same time.
Can you give us an example of a customer that has been positively affected by Coveo’s solutions?
Sure. CA Technologies uses Coveo as the backbone for pulling together information within their customer service organizations, and they have expanded it enterprise-wide. There are about 13,000 users, and we index 74 systems. There are some older legacy systems there, and we can’t completely integrate all of those. The old way of doing things included integrating. Back in 2000, if we talked to the average CIO, their strategy was consolidation. They wanted to eliminate systems into a centralized, perfectly integrated system that will do everything. The outcome of that attempt is CRM, but the reality is that information sources keep proliferating. At CA Technologies, where we index 74 different systems, we pull together everything into one single hub. A search can be launched across all of those systems, securely and according to their own credentials and privileges. They can start performing mash-ups and assemble pieces of information together. It allows them to solve customer issues more quickly, especially with over 13,000 users and 1,600 of them working in the call center.
From an external perspective, Coveo also changed CA.com. We serve about 150,000 external authenticated users who use the CA.com website at a rate of approximately 1,000 core users per minute globally. We assemble a personalized experience for every one of them. When customers enter the site, it already knows what they bought from the company, their open tickets, when they were there last, and so on. We can make the experience far more relevant and create call deflection because customers are able to find what they need for themselves quickly and efficiently, which they prefer.
Will you tell us about the major components of the Coveo solution?
Coveo essentially has two major components. One of them is called the enterprise indexing technology, and it reaches out to any content residing within the organization, including CRM content, technical reference, call center information, infrastructure content, engineering databases, as well as social media and customer community content. It reaches out to the content and pulls it onto a common unified secure index and normalizes the information. That’s the first component. The second one is what we call the Coveo role-based Insight Consoles. This component essentially provides the user, such as the call center agent or the customer, with the ability to gain more comprehensive knowledge and context, link the information together, and identify key people to contact so they can gain knowledge and the capability to take the appropriate action. These Consoles basically glue the information together from the index, regardless of source or format. It makes the data highly available, highly conversational, very personalized, and relevant for the user. As a practical matter, the ability to do that enables the customer to come online and immediately see and converse with information. It’s more relevant to the specific products they bought, specific issues they logged, specific symptoms they have, or the questions they might ask. It’s relevant to the pattern of how they have historically consumed information. This is all factored into the relevance of the information we deliver.
How long does it take a client to deploy the Coveo solution?
That is one of the nice things about this kind of technology. It’s somewhat like Google or Yahoo, which are totally noninvasive for the web. If you put up your new website tomorrow with a server running behind it, Google and Yahoo are going to crawl and basically pull the information into the common index. It’s the same paradigm. We can come into an organization quickly, providing we have access to the systems, and basically connect, crawl those systems, and pull the information. It also depends on the size of the data stores and the volumes of information. We have had implementations from 50 up to 700 different systems, with hundreds of millions of documents. In most organizations, we can come into a customer service environment and upload within a matter of a few days or weeks by harvesting the existing systems already there. Typically, there will be Microsoft SharePoint, a CRM system like Salesforce.com, or maybe even a ticketing system such as BMC Remedy. There could be a couple of knowledge bases and databases with email content or document content. For us to crawl the content and bring it together in a common index, it would be a matter of a few days or weeks, depending on the volume. The nice thing is that it creates zero disruption because we’re just harvesting existing information. Then, we design the consoles, using our console components and widgets, for every role. Typically, our customers have a customer service organization that is split into two or three different roles, usually the entry level and then advanced levels and roles when a case gets escalated. We are able to configure consoles for every one of those roles and alternatively configure the consoles for the self-service module in order to drive call deflection. With a self-service experience on the web for users trying to solve a problem or get an answer to a question, the company will be more likely to drive call deflection.
Regarding the data you’re indexing, does a limitation exist based on whether the data is structured or not?
There is no limitation. Structured information, by definition, obviously has a structure. So, that can be database records and other information typically found in the ODBC world, or DB II or SQL databases. We index those. Unstructured pieces of information are things like documents, communications, e-mails, chats, customer comments, call transcripts, or social feeds within communities, such as data from Twitter for example. The way we make sense of unstructured information is through the use of advanced text analytics. Essentially, we crawl those systems and extract key entities and themes. Key entities or themed entities are names of people, customers, and also products, product codes, component names of codes, and such. Themes are things such as oil, petrol, and gas. Using those parameters, we start making sense of the content. We start understanding the meaning of the unstructured content, and we can then correlate it. For instance, we look at a symptom expressed in a call center transcript and specifically correlate that symptom against all the history of tickets the company has ever solved. We can find out about a certain type of product, go back into the engineering database, and start finding engineering projects or certain features that actually correlate with the theme. Recently, we had a customer where this was implemented, and the number of tickets escalated by the customer support department to the engineering department dropped by 70 percent, because the original customer service agent was able to connect the dots between engineering information and customer symptoms.
When deploying Coveo, is a human intermediary required to package the results for the end user? Or does Coveo drive the solutions itself?
I’ll preface this answer by saying that in an organization, there is data, information, and people. Data is numbers, organizing tables, and such. Information is essentially data that is organized in a somewhat meaningful way to aid decision making. In order to solve customer issues, innovate products, or know the best things to sell to a certain customer, you need knowledge. Many organizations confuse knowledge and information. That is why knowledge management is something residing in the IT department or something that organizations throw into a database. By definition, knowledge is not information. It’s the human capacity to take effective action when facing an uncertain situation. Customer service is a classic example of that. In customer service, 80 percent of the interactions are reasonably certain. For example, you call T-Mobile because you’ve lost your password. That is reasonably predictable because it’s a highly repetitive issue. Although 80 percent of the calls are about repetitive issues, they only represent about 30 to 40 percent of the cost, and they are not the source of customer dissatisfaction. However, 60 to 80 percent of the cost of call centers comes from dealing with uncertain situations, when it takes significantly longer to solve an issue. The customer could become upset, and it could be a source of customer turnover.
You need knowledge and a better human capacity. At the end of the day, you’re dealing with people. Those people could be agents or customers who need better knowledge so problems can be solved. Insight is the ability to gain knowledge. In order to gain knowledge to solve a problem, you need to do two things. You need to link all the key relevant information together and do that in the context of your own knowledge. You don’t need a system to tell you what you already know. As a user, you already have certain knowledge, and you just need to complement it. You need to be able to converse with information and link the key information. The second thing you need is to identify key people. People with knowledge are walking around an organization, and they can help solve the problem at hand. In the instance of a call center with a thousand agents, there are agents on the phone with customers and other colleagues somewhere who have solved similar problems. You need to identify them.
Going back to your question, this is a new generation of systems and user interfaces with a highly conversational feature. These inside consoles are somewhat similar to mash-ups of information you see on the web that can be performed using indexed information. Essentially, we don’t need any human intervention to organize the information. We simply provide the agent with a much better interface to converse with multiple sources of information, so they can search simultaneously within many systems. We consolidate information from multiple systems and correlate information. We provide the ability to converse. The data is present; there is an immense amount of knowledge and information residing within an organization. Usually, individual customers and agents just can’t tap into it.
Will you further explain the link between knowledge and economic value?
For some companies, knowledge is the greatest asset they have. In an organization, there are hard assets, such as desks, computers, and phones. There is knowledge, which is the human capital, the assembled know-how, information, and so on. This is the biggest asset, and there is evidence of that in the stock market. There is a gap between enterprise value and book value of companies because the book value represents the hard assets minus depreciation. For example, assume you owned Coca-Cola and decided to destroy all the plants, equipment, trucks, and hard assets they have, but you still owned the brand and people. Would you be able to raise money at a bank and rebuild the infrastructure? I believe you would. If knowledge is an asset, then the return on that knowledge is ultimately linked to the ability to enable people to access that collective knowledge more efficiently. It either drives a return, or otherwise, that asset sits in a corner and the organization fails to leverage it.
As another example, assume you and I are two agents in two customer service organizations. Yesterday, I was called by a customer, I solved their problem, and it’s all logged in my ticket. Today, you’re sitting in another office in a different part of the world and you get a call from a customer with a similar issue, but you don’t have the ability to tap into the work I did yesterday. What is the impact on the organization? They are basically reinventing the wheel, so the impact is dollars. By failing to leverage the collective vat of knowledge, the organization is asking people to recreate knowledge each day. The currency of customer service is time. It can translate into a lot of money. The impact of bad customer service is a greater investment in sales and marketing because of the need to regain lost customers or rebuild a poor reputation. The strategic and economic impact of managing knowledge is very significant.
How do companies justify the level of investment or cost needed to manage knowledge?
I mentioned that time is the currency of the customer service organization. Fundamentally, time has a direct impact on the cost of an organization. If agents spend too much time solving issues, it has a direct impact on cost. If calls are logged that should have been answered by self-service, that is another direct source of cost. More indirect costs are incurred by prolonging the time spent in customer service, including the impact on satisfaction, reputation, and customer churn. The ability of an organization to effectively resolve customer issues is very important in terms of time.
The way our customers justify the cost is fairly simple. If you break down the anatomy of a customer service call, there are three core activities. Identify the problem or the question and define it. Research the solution. Implement the solution. The rest is logistics, such as tracking the call, wrapping up the call, or potentially escalating the call to someone else. Customer service leaders can start by measuring the activity performed in their customer service organizations. I expect that most would find that that than 50 percent of their call center bandwidth is associated with gaining insight and knowledge, which we define as knowledge-driven activity.
From an ROI perspective, it’s straightforward. We can cut those knowledge-driven activities in half. And if 50 percent of the call center time is spent on those activities and the company spends $100 million per year on the call center, we can guarantee that $50 million of the budget is associated with people spending time trying to figure out answers to problems. If we can slash that time in half, we can save the company $25 million per year.
Are there applications of this solution for the health care industry?
We just won a significant situation in the healthcare sector. It’s actually a customer service situation within call centers. In this case, citizens are calling a government agency’s health line. Essentially, the person answering the call has 21 different systems from which they pull data because it can be anything from symptoms to medication to treatments to patient information. It has ramifications everywhere. The ability to tap into an environment of highly diversified information in order to quickly find answers and solutions is fundamental. That’s what Insight is all about, and it’s very powerful.
Normally when partnering with client organizations, who is your main contact at the company?
It can be the head of customer experience or head of engineering. Increasingly, it’s the COO or CEO because they are starting to understand that their ability to leverage collective knowledge is deficient. Our solutions can fix that.
Using a baseball analogy to analyze the adoption curve of this type of technology, what inning are we in?
I think we’re in the first inning of a long game. I think this is the launch of a new paradigm, and it’s not something we have previously seen in the consumer world. Looking at Google and Yahoo and the search ability across the web, these are all ways to get information quickly. That’s the tactical aspect. Fundamentally, you’re empowering people to become smarter, tap into the world’s collective knowledge more effectively, and get a return on the collective knowledge. The question becomes what are the constraints within enterprises that prevent this paradigm shift from occurring. Why are we still boxing people in front of a genius screen or SAP screen and using a preconceived workflow instead of SQL queries? It’s primarily because an enterprise’s content is diversified versus the content on the web, which is reasonably homogeneous. This paradigm shift is providing every single stakeholder and user of this type of technology with a greater ability to gain knowledge and insight. Insight is about gaining knowledge, identifying people, and linking key relevant
information in order to solve problems. The ability to do that will create huge economic value.
What is your opinion about trends that will impact customer service organizations in the future?
At a very high level from a pure company perspective, the common theme across the industry is a drive to become better at lowering costs. The margin pressures on customer service are ever increasing. Firstly, customer service is viewed as a cost center. Companies are often asked to do more with less. At the other end is a customer audience that operates in an increasingly proliferate and democratic world. Customers have access to a lot of information, which creates a more competent customer who expects companies to be more competent. Additionally, it creates a customer with higher expectations around the availability of relevant information to solve issues. There are user pressures for more insight agility, possibly fueled by their own consumer experiences. If they’re unhappy with the service, they can turn to other providers that have better differentiation on service.
Those factors drive a lot of requirements for self-service on the part of the customer. Self-service has been around for a long time, at least since the web has been pervasive. Relevant self-service is something that will be a major trend. The question companies need to address externally is how to make the self-service experience one that is relevant to your current situation. Internally, it drives a requirement for more and better knowledge and accessibility on the part of customer service agents. In turn, that drives the confluence between sales, service, and engineering. Enabling that triangle is critical. Those departments typically operate in a silo, or a waterfall. With the waterfall analogy, engineering builds a product and shoves it over to sales. Then, the sales department shoves it over to service. And there is little interaction. Now, these three are really coming together with the consolidation of knowledge across the organization between engineering, sales, and service. It’s critical to solving customer issues faster and ensuring product issues get resolved faster.
At a macro level, the rate of change and the variety of information sources are increasing. Between the years 2000 and 2010, the average number of information sources a company had to deal with was multiplied by roughly two-and-a-half times. In 2000, assume a customer service environment had five to ten sources of information that were relevant to answering customer demands and solving issues. Ten years later, they now have 25 sources. That is the reality today. I was with the Executive VP of Engineering of a large consumer electronics company recently. Even though they have engineering systems and PLM systems internally, when he throws a new product into the market, within 24 hours, he has already heard about the design flaws via Twitter. That story speaks volumes about the importance of distilling insights from the consolidated view, from all enterprise information plus all social media data. We don’t know what the landscape will look like three years from now. No one can predict the next paradigm shift. Adapting to change and being able to quickly reassemble the information people need is the new paradigm. That is opposed to the old paradigm of IT personnel sitting around the table discussing the information needs for the next five years.
Louis Têtu is Chairman and Chief Executive Officer of Coveo. Coveo transforms companies’ ability to gain insight from diverse and overwhelming amounts of unstructured and structured data, whether it exists behind the firewall or in social media. Greater insight enables more effective and efficient customer service, more relevant customer experiences, increased sales and shorter sales cycles, faster innovation for better product development and ultimately, increased profitability. Coveo customers include Lockheed Martin, PepsiCo, Verizon, GEICO, CA Technologies, T-Mobile, Terumo Medical, IBM Netezza and Children’s Hospital of Boston.
Prior to Coveo, Mr. Têtu co-founded Taleo Corporation [NASDAQ: TLEO], the leading international provider of on-demand Internet software for talent and human capital management, where he held the position of Chief Executive Officer and Chairman of the Board of Directors from the company’s inception in 1999 through 2007. In 2004, Taleo was recognized the 11th fastest growing technology company in the United States within the Deloitte Technology Fast 500.
Prior to Taleo, Mr. Têtu was President of Baan SCS, the supply-chain management solutions group of Baan, a global enterprise software company with more than 5,000 employees. This followed Baan’s acquisition of Berclain Group inc., which Mr. Têtu co-founded in 1989 and where he served as president until 1996. Mr. Têtu is an Engineering graduate from Laval University of Canada in 1985 and in 1997 was honoured by Laval for his outstanding social contributions and business achievements. He also received the 2006 Ernst & Young Entrepreneur of The Year award in the Technology and Communication category.
Outside of his professional career, Mr. Têtu is a private equity investor involved in technology and infrastructure projects within emerging countries, as well as a commercially licensed helicopter pilot. He lives in Quebec with his wife and their three children.
Managing Member and Founder Argyle Executive Forum Jason is Argyle Executive Forum’s managing member and founder. Argyle Executive Forum is a professional services firm that convenes and connects business leaders from highly targeted business-to-business communities for strategic collaboration and business development.
Over 40,000 executives participate in one or several of Argyle Executive Forum’s communities, with over 700 new members joining every month. Prior to forming Argyle Executive Forum, Jason launched the private-equity business effort for Capital IQ. Capital IQ was acquired by Standard & Poor’s in 2004. Prior to Capital IQ, Jason was an investment banker, focused on middle-market M&A and LBO transactions. He holds a Bachelor of Science from Cornell and an MBA from Harvard Business School.