For enterprises to be truly revolutionary, data has to be leveraged correctly. But what happens when frontend units, such as sales, and backend ones, like engineering, do not have access to the same datasets?
DevRev, established in 2020, aims to bridge the gap between customer-centric and product-centric teams, in order to unlock more business value for companies. Manoj Agarwal, Co-Founder and President of DevRev, sat down with Frontier Enterprise to share the company’s enterprise vision, his previous experience as SVP of Engineering at Nutanix, and why developers are having a hard time overcoming systemic and technological challenges.
Could you tell me about DevRev’s global strategy?
In India, we have three offices. But our overall headquarters is in the San Francisco Bay Area. We started in Palo Alto. We have another office in Austin. Strategically, we have placed our offices very close to university campuses around the world. So you will see that Palo Alto is close to Stanford and Berkeley, while Austin is next to the University of Texas Austin. In India, obviously, Bangalore is the place where all technology companies have to be. There are lots of good schools around there. Our idea was not just development, but also to collaborate and conduct research with elite institutions – in India in particular that means two institutions: IIT Madras and IIT Bombay. We are in Europe as well, in Slovenia, and in Southeast Asia, in Singapore.
What sort of gap did you see in the market that prompted you to launch DevRev?
One of the things that we’re super proud of coming out of Nutanix was something called Net Promoter Score, or NPS. People talk a lot about the customer satisfaction score (CSAT), but a good NPS is an extremely hard metric to attain. How do you even achieve that state where you don’t even have to say the name of your company because your users are talking about you everywhere else? We looked at the best brands, and the one that I admired, which is Apple, has a net promoter score of +72, on a scale of -100 to +100.
At the speed at which we were growing every year, our net promoter score stayed consistently above +90 at Nutanix. And today, it is at +92. This is a number that is hard to even imagine, but there was obviously a lot of work that it required. You’re obviously talking to the customers, but inside the company, what are you doing? There are the product managers who are planning their roadmap, what to build, what not to build, and so on. And then you have a product engineering team who’s doing the execution of what to build, what to likely prioritise and work on. And for us it was, ‘could we really have this tight loop from the customers all the way to the development? And when they finish the work, can you really get that information out very quickly in front of the customers?’ That was super important for us. But there were no tools or systems that existed before that could achieve this. We did large scale automation, while also hiring a lot of people to bridge the data gaps, because you also need to get quality information when customers are really making too much noise. How do you really get a sense of what they’re saying and not saying? And how do we really feed that information to the product management team, so that even the planning becomes outside in as opposed to everything inside out? How do you ensure that the engineering execution is also not done in the silos without the customer input? To do that always required a lot of meetings, connecting many people, and consensus building.
For example, when you look at what’s happening today, the entire corporate world is built around having a front office and a back office. You could say that the front office is customer-centric, in some sense, but they hardly have the information about the product in real-time, and they don’t even have the tools and systems which give them that information on the product. Even knowing what exactly the customer is using, or what exactly their pain points are, or what exactly they are saying about the product – this information is hard to come by for them easily
In the back office, you have the software development, and product managers trying to really see both sides through and figure out if they can do something for cloud ops. The software development and the engineering folks are mostly product-centric. They don’t have much information about the customer. They just look up to the product manager, and take up whatever the product manager says – this is the customer centricity they can easily afford. There is practically nothing else, in terms of tools or systems that are being used to identify any information about the customer. So what happens is that there is a lot of manual tagging.
Even when you think about it, the way the prioritisation of customer pain point resolutions happens is that there are varying degrees of priorities and you have to identify the right priorities for the moment for the business. The key question is how do you prioritise current pain points based on the actual business need?
How do you move from that legacy way of prioritising things into that Now, Next, Later framework?
Humans can only think of so many things during a given period of time. When it comes to prioritising things you need to ask: ‘What am I really working on now? What should be my priority now? What are the next set of things that are prioritised for me?’ Everything else goes into the backlog, which goes into the Later category. When software developers do project management, they work on a Sprint, and a Sprint is completed in two weeks. We wanted to make sure that companies don’t lose out on most of the important constructs. So we do have Now, Next, Later in the sense that if you want to run your team that way, then you can have the entire roadmap that is also designed that way— Now, Next, Later. But then when you’re doing the engineering execution, where they want to know what you’re going to achieve in two weeks they can have that Sprint framework in which they also receive the entire customer information. Let’s say you have 100 items that you’re doing as part of the Sprint, even then you can go and sort them by customer impact.
Let’s say teams are very used to JIRA, for example. How do you integrate that process from the customer into the JIRA workflow?
If a company is using JIRA, it can continue to use Jira. However, what we have invested heavily in is Airdrop technology, which does a two-way sync with Jira. Every single workflow gets synced into the DevRev platform. For collaboration we put on top of it machine learning, AI, and a lot of value added features, however you also have the entire data, object by object with the workflows that gets mapped into DevRev.
The teams typically don’t have to right away move to DevRev, but only once they become very comfortable, or they can do it one team at a time. Today, when you go to Atlassian Jira, it’s mostly about projects. You think about a project when you go to Jira, rather than a product. Projects start, and then they end. A product has to live forever. There is no system that exists that can talk in the language of product and customers on the same platform, and that is what DevRev is doing.
DevRev is composed of these two words— Dev for development, which is proxy for product, and Rev for revenue, which is proxy for customer. So how do you bring these two important constituents, product and customers, on to the same platform because the entire existence of any company is built around products and customers? We wanted to make sure that we capture that in the new object model.
Which verticals do you see your solution creating a disruption?
We’re seeing a lot of adoption in B2B, but also in B2C. For example, consider product analytics, which analyses the way that the people use the product. When the customers have a problem, how can these companies quickly reach back to the customers? And when they reach back, can they really provide all the context of what exactly the user was doing in a more visual way? So think about it like a session recording, but very, lightweight to consume, highly-efficient, and high-performing. We are talking about ingesting and analysing hundreds of millions of sessions.
As part of our platform’s functionality, you can see how users are consuming an application, and more importantly, when any kind of failure happens and people complain about it on social media, or on the app stores, we can bring all that voice on one platform. We can cluster and classify that data, get the signals out of all that noise, and make it more actionable to the companies. We are seeing a lot of use cases now on the B2C front, but the BFSI sector is definitely onboard, with some of the largest banks in the world currently deploying our platform.
