We’ll see how advanced analytics, drawing on multiple data sources, enables INOVVO’s mobile carrier customers to provide mobile users with faster, more reliable, and relevant services.
To learn more about how INOVVO uses big data to make major impacts on mobile services, please join me in welcoming Joseph Khalil, President and CEO of INOVVO in Reston, Virginia. The discussion is moderated by me, Dana Gardner, Principal Analyst at Interarbor Solutions.
Here are some excerpts:
Gardner: User experience and quality of service are so essential nowadays. What has been the challenge for you to gain an integrated and comprehensive view of subscribers and networks that they’re on in order to uphold that expectation for user experience and quality?
Khalil: As you mentioned in your intro, we cater to the mobile telco industry. Our customers are mobile operators who have customers in North America, Europe, and the Asia-Pacific region. There are a lot of privacy concerns when you start talking about customer data, and we’re very sensitive to that.
The challenge is to handle the tremendous volume of data generated by the wireless networks and still adhere to all privacy guidelines. This means we have to deploy our solutions within the firewalls of network operators. This is a big-data solution, and as you know, big data requires a lot of hardware and a big infrastructure.
So our challenge is how we can deploy big data with a small hardware footprint and high storage capacity and performance. That’s what we’ve been working on over the last few years. We have a very compelling offer that we’ve been delivering to our customers for the past five years. We’re leveraging HPE Vertica for our storage technology, and it has allowed us to meet very stringent deployment requirements. HPE has been and still is a great technology partner for us.
Gardner: Tell us a little bit more about how you do that in terms of gathering that data, making sure that you adhere to privacy concerns, and at the same time, because velocity, as we know, is so important, quickly deliver analytics back. How does that work?
Khalil: We deal with a large number of records that are generated daily within the network. This is data coming from deep packet inspection probes. Almost every operator we talk to has them deployed, because they want to understand the user experience on their networks.
These probes capture large volume of clickstream data. Then, they relay it to us almost in a near real-time fashion. This is the velocity component. We leverage open-source technologies that we adapted to our needs that allow us to deal with the influx of streaming data.
We’re now in discussion with HPE about their Kafka offering, which deals with streaming data and scalability issues and seems to complement our current solution and enhances our ability to deal with the velocity and volume issues. Then, our challenge is not just dealing with the data velocity, but also how to access the data and render reports in few seconds.
One of our offering is a care product that’s used by care organizations. They want to know what their customers did the last hour on the network. So there’s a near real-time urgency to have this data streamed, loaded, processed, and available for reporting. That’s what our platforms offers.
Gardner: Joseph, given that you’re global in nature and that there are so many distribution points for the gathering of data, do you bring this all into a single data center? Do you use cloud or other on-demand elements? How do you manage the centralization of that data?
Khalil: We don’t have cloud deployments to date, even though our technology allows for it. We could deploy our software in the cloud, but again, due to privacy concerns with customers’ data, we end up deploying our solutions in-network within the operators’ firewalls.
One of the big advantages of our solution is that we can choose to host it locally on customers’ premises. We typically store data for up to 13 months. So our customers can go and see the performance of everything that’s happened on the network for the last 13 months.
We store the data at different levels — hourly, daily, weekly, monthly — but to answer your question, we deploy on-site, and that’s where all the data is centralized.
Gardner: Let’s look at why this is so important to your customer, the mobile carrier, the mobile operator. What is it that helps their business and benefits their business by having this data and having that speed of analysis?
Khalil: Our customer care module, the Subscriber Analytix Care, is used by care agents. These are the individuals that respond to 611 calls from customers complaining about issues with their devices, coverage, or whatever the case may be.
When they’re on the phone with a customer and they put in a phone number to investigate, they want to be able to get the report to render in under five seconds. They don’t want to have the customer waiting while the tool is churning trying to retrieve the care dashboard. They want to hit “go,” and have the information come on their screen. They want to be able to quickly determine if there’s an issue or not. Is there a network issue, is it a device issue, whatever the case may be?
So we give them that speed and simplicity, because the data we are collecting is very complex, and we take all the complexity away. We have our own proprietary data analysis and modeling techniques, and it happens on-the-fly as the data is going through the system. So when the care agent loads that screen, it’s right there at a glance. They can quickly determine what the case may be that’s impacting the customer.
Our care module has been demonstrated to reduce the average call handle time, the time care personnel spend with the customer on the phone. For big operators, you could imagine how many calls they get every day. Shaving a few minutes off each call can amount to a lot of savings in terms of dollars for them.
Gardner: So in a sense, there’s a force-multiplier by having this analysis. Not only do you head off the problems and fix them before they become evident, which includes better user experience, they’re happier as a customer. They stay on the network. But then, when there are problems, you can empower those people who are solving the problem, who are dealing with that customer directly to have the right information in hand.
Khalil: Exactly. They have everything. We give them all the tools that are available to them to quickly determine on the fly how to resolve the issue that the customer is having. That’s why speed is very important for a module like care.
For our marketing module, speed is important, but not as critical as care, because now you don’t have a customer waiting on the line for you to run your report to see how subscribers are using the network or how they’re using their devices. We still produce reports fairly quickly in few seconds, which is also what the platform can offer for marketing.
Gardner: So those are some of the immediate and tactical benefits, but I should think that, over time, as you aggregate this data, there is a strategic benefit, where you can predict what demands are going to be on your networks and/or what services will be more in demand than others, perhaps market by market, region by region. How does that work? How do you provide that strategic level of analysis as well?
Khalil: This is on the marketing side of our platform, Subscriber Analytix Marketing. It’s used by the CMO organizations, by marketing analysts, to understand how subscribers are using the services. For example, an operator will have different rate plans or tariff plans. They have different devices, tablets, different offerings, different applications that they’re promoting.
How are customers using all these services? Before the advent of deep packet inspection probes and before the advent of big data, operators were blind to how customers are using the services offered by the network. Traditional tools couldn’t get anywhere near handling the amount of data that’s generated by the services.
Today, we can look at this data and synthesize it for them, so they can easily look at it, slice and dice it along many dimensions such as, age, gender, device type, location, time, you name it. Marketing analysts can then use these dimensions to ask very detailed questions about usage on the network. Based on that, they can target specific customers with specific offers that match their specific needs.
Gardner: Of course, in a highly competitive environment, where there are multiple carriers vying for that mobile account, the one that’s first to market with those programs can have a significant advantage.
Khalil: Exactly. Operators are competing now based on the services they offer and their related costs. Back 10-15 years ago, radio coverage footprint and voice plans were the driving factors. Today, it’s the data services offered and their associated rate plans.
Gardner: Joseph, let’s learn a little bit more about INOVVO. You recently completed purchase of comScore’s wireless solutions division. Tell us a bit about how you’ve grown as a company, both organically and through acquisition, and maybe the breadth of your services beyond what we’ve already described?
Khalil: INOVVO is a new company. We started in May 2015, but the business is very mature. My senior managers and I have been in this business since 2005. We started the Subscriber Analytix product line back in 2005. Then, comScore acquired us in 2010, and we stayed with them for about 5 years, until this past May.
At that time, comScore decided that they wanted to focus more on their core business and they decided to divest the Subscriber Analytix group. My senior management and I executed a management buyout, and that’s how we started INOVVO.
However, comScore is still a key partner for us. A key component of our product is a dictionary for categorizing and classifying websites, devices, and mobile apps. That’s produced by comScore, and comScore is known in this industry as the gold standard for these types of categorizations .
We have exclusive licensing rights to use the dictionary in our platform. So we have a very close partnership with comScore. Today, as far as the services that INOVVO offers, we have a Subscriber Analytix product line, which is for care, marketing, and network.
We talked about care and marketing, we also have a network module. This is for engineers and network planners. We help engineers understand the utilization of their network elements and help them plan and forecast what the utilization is going to be in the near future, given current trends, and help them stay ahead of the curve. Our tool allows them to anticipate when existing network elements exhaust their current capacity.
Gardner: And given that platform and technology providers like HPE are enabling you to handle streaming real-time highly voluminous amounts of data, where do you see your services going next?
It appears to me that more than just mobile devices will be on these networks. Perhaps we’re moving towards the Internet of Things (IoT). We’re looking more towards people replacing other networks with their mobile network for entertainment and other aspects of their personal and business lives. At that packet level, where you examine this traffic, it seems to me that you can offer more services to more people in the fairly near future.
Khalil: IoT is big and it’s showing up on everybody’s radar. We have two paths that we’re pursuing on our roadmap. There is the technology component, and that’s why HPE is a key partner for us. We believe in all their big data components that they offer. And the other component for us is the data-science component and data analysis.
The innovation is going to be in the type of modeling techniques that are going to be used to help, in our case, our customers, the mobile operators. Moving down the road, there could be other beneficiaries of that data, for example companies that are deploying the sensors that are generating the data.
I’m sure they want some feedback on all that data that their sensors are generating. We have all the building blocks now to keep expanding what we have and start getting into those advanced analytics, advanced methodologies, and predictive modeling. These are the areas, and this is where we see really our core expertise, because we understand this data.
Today you see a lot of platforms showing up that say, “Give me your data and I’ll show you nice looking reports.” But there is a key component that is missing and that is the domain expertise in understanding the data. This is our core expertise.
Gardner: Before we finish up, I’d like to ask you about lessons learned that you might share with others. For those organizations that are grappling with the need for near real-time analytics with massive amounts of data, having tremendous amount of data available to them, maybe it’s on a network, maybe it’s in a different environment, do you have any 20/20 hindsight that you might offer on how to make the best use of big data and monetize it?
Khalil: There is a lot of confusion in the industry today about big data. What is big data and what do I need for big data? You hear the terms Hadoop. “I have deployed a Hadoop cluster. So I have solved my big data needs.” You ask people what’s their big-data strategy, and they say they have deployed Hadoop. Well, then. what are you doing with Hadoop? How are you accessing the data? How are you reporting on the data?
My advice is that it’s a new field and you need to consider not just the Hadoop storage layer but the other analytical layers that complements it. Everybody is excited about big data. Everybody wants to really have strategy to use big data, and there are multiple components to it. We offer a key component. We don’t pitch ourselves to our customers and say, “We are your big data solution for everything you have.”
There is an underlying framework that they have to deploy, and Hadoop is one of them. then comes our piece. It sits on top of the data hosting infrastructure and feeds from all the different data types, because in our industry, typical operators have hundreds if not thousands of data silos that exist in their organization.
So you need framework to really host the various data sources, and Hadoop could be one of them. Then, you need a higher-level reporting layer, an analytical layer, that really can start combining these data silos and making sense of it and bringing value to the organization. So it’s a complete strategy of how to handle big data.
Gardner: And that analytics layer that’s what HPE Vertica is doing for you.
Khalil: Exactly. HPE is a key component of what do we do in our analytical layer. There are misconceptions. When we go talk to our customers, They say, “Oh, you’re using your Vertica platform to replicate our big data store,” and we say that we’re not. The big data store is a lower level, and we’re an analytical layer. We’re not going to keep everything. We’re going to look at all your data, throw away a lot of it, just keep what you really need, and then synthesize it to be modeled and reported on.
You may also be interested in:
- DevOps by design–A practical guide to effectively ushering DevOps into any organization
- Need for Fast Analytics in Healthcare Spurs Sogeti Converged Solutions Partnership Model
- HPE’s composable infrastructure sets stage for hybrid market brokering role
- Nottingham Trent University Elevates Big Data’s role to Improving Student Retention in Higher Education
- Forrester analyst Kurt Bittner on the inevitability of DevOps
- Agile on fire: IT enters the new era of ‘continuous’ everything
- Big data enables top user experiences and extreme personalization for Intuit TurboTax
- Feedback loops: The confluence of DevOps and big data
- Spirent leverages big data to keep user experience quality a winning factor for telcos
- Powerful reporting from YP’s data warehouse helps SMBs deliver the best ad campaigns
- IoT brings on development demands that DevOps manages best, say experts
- Big data generates new insights into what’s happening in the world’s tropical ecosystems
- DevOps and security, a match made in heaven
- How Sprint employs orchestration and automation to bring IT into DevOps readiness
- How fast analytics changes the game and expands the market for big data value