Uncategorized

During the past few years, the IPO market has seen a variety of database companies hit the markets. And the latest one came this week: Couchbase. The company issued 8.3 million shares at $24 each, which was above the initial price range of $20 to $23. On the first day of trading, the shares jumped by 27%.

The origins of Couchbase go back to the development of an open source project, called CouchDB (its an acronym for cluster of unreliable commodity hardware). The developer of this technology was Damien Katz, who formerly worked at IBM. He started on this project in 2005 and launched the first stable version in 2010.

Startups like CouchOne and Membase saw the potential of this technology and began to build their own enhancements. These companies would also merge in 2011 to form Couchbase, which would combine core database technology with caching systems. The goal was to make the platform highly scalable and reliable for enterprise customers. 

The Technology

The relational database is the most dominant model. While it continues to be robust, the technology has had difficulties with the needs for handling millions of users and managing spikes in workloads. Relational databases are also far from cheap. 

“Relational databases were built for large and monolithic applications,” said Matt Cain, who is the CEO of Couchbase. “But today’s applications involve many datasets and heavy interactions.”

To address the problems, there has emerged a new model called the NoSQL database. It is based on a document model and can work effectively with any type of data, such as structured, unstructured, time series and so on. 

However, the first generation of NoSQL databases was not for mission-critical applications. Part of this was due to the underlying technology. But there was also the problem with the implementation because of the need to retrain database administrators.

As for Couchbase, the company has focused on applying enterprise-grade systems to its platform. Consider that it works in a myriad of configurations, whether for the cloud, hybrid, or on-premise environments. 

“At Couchbase, we are combining the best of both relational databases and the NoSQL model,” said Cain. 

A key factor for Couchbase is that the system has been built for seamless deployment. This means there is minimal downtime. As a result, the company has been successful in migrating mainframe and relational database implementations. For about 80% of its customers, the database is used as a source of truth or system of record for some or all of their business.

Market Opportunity

The database market is enormous. Keep in mind that it’s one of the largest undisrupted markets for enterprise software. 

According to IDC, the spending was about $42.9 billion last year and it is forecasted to hit $62.2 billion by 2024. One of the main catalysts for growth is the need for digital transformation, such as to offer mobile apps, manage edge systems and leverage Artificial Intelligence. Such technologies simply do not work well with relational databases.

“With digital transformation, it’s not just about creating new applications,” said Cain. “There is also a need for re-platforming existing ones.”

In other words, the future does look bright for Couchbase. The company has a robust technology that is used by many large customers and handles enormous workloads. More importantly, there is a real need for companies to transition to other approaches—and this may mean that the database market is finally at a critical point for major change.

Confluent, which develops database technologies, launched its IPO this week. The company issued 23 million shares at $36 a piece—above the $29-to-$33 price range—and the price rose 25% on the first day of trading. The market capitalization hit $11.5 billion. 

Jay Kreps, Jun Rao, and Neha Narkhede cofounded the company in 2014. Before this, they were software engineers at LinkedIn and had faced the tough challenges of scaling the IT infrastructure. One of the main issues was how to effectively process data in real time.

The cofounders searched for a solution but there was nothing that was viable. For the most part, the database technologies were mostly about storing information efficiently—not handling data in motion.

So the founders developed their own platform, which they called Kafka, and made it open source. From the start, it saw significant adoption. 

As of now, Kafka has a developer community of over 60,000 and the software is used by over 70% of the Fortune 500.  In other words, the cofounders were ideally positioned to capitalize on this growth with their own startup. “Confluent is part of a new wave of data companies that offer faster, more scalable solutions for the modern digital era,” said Jedidiah Yueh, who is the CEO of Delphix.

The Technology

An app like Uber or Lyft may seem simple. But the underlying technology is exceedingly complex, as it needs to manage enormous amounts of streaming data to connect customers. 

“In such situations, to provide a delightful customer experience, data needs to be analyzed in-motion even before it is saved in databases,” said Ashish Kakran, who is a Principal at Thomvest Ventures. “Such a feat was almost impossible in the past because data would need to be stored before analysis. This is where Confluent’s open-source Kafka stands out as it turns a sequential analysis process into a fluid one. It provides an event streaming platform and a rich set of APIs that developers can use to quickly be productive.”

The reliance on developer communities has been critical. It has allowed for a bottoms-up strategy for adoption. Let’s face it, selling a complex technology like Kafka directly to Chief Information Officers would likely be tough.

“Now as the same hard-to-sell products grow organically within an organization, they become hard-to-replace and their value becomes self-evident to CIOs,” said Kakran. “Companies like Confluent then step in with innovative business models to monetize their offerings and help organizations get the most out of innovative open-source tools.”

The opportunity for Confluent is still in the early stages. Keep in mind that it estimates the market at about $50 billion. That is, the need for real-time data spans many industries. For example, a retailer can use it for accurate inventory tracking so as to make sure customers get what they want when they want it. Or in manufacturing, a company can leverage IoT (Internet-of-Things) data to help with predictive maintenance, which can mean less downtime.

According to Kreps, in his letter to shareholders: “Today the data architecture of a company is as important in the company’s operations as the physical real estate, org chart, or any other blueprint for the business. This is the underpinning of a modern digital customer experience, and the key to harnessing software to drive intelligent, efficient operations. Companies that get this right will be the leaders in their industries in the decades ahead.”

The funding environment for AI startups remains robust—and some of the rounds have been substantial. Just recently Dataiku, which operates a machine learning platform, announced a $100 million Series D investment. 

AI truly represents a transformation in the tech world.  “AI is making software much more dynamic and improves as it understands user behavior,” said Gordon Ritter, who is the founder and General Partner at Emergence. 

OK then, what are VCs looking for when evaluating an AI deal? What should be in your pitch deck?

Well, to answer these questions, I talked to a variety of VCs. Here’s what they said:

Sri Chandrasekar, partner at Point72 Ventures: 

A slide on Why Now. What technology has recently been developed that has made solving this customer problem now possible. It might be “speech-to-text technology has gotten good enough to use in 95% of a call center’s communications.” Or “recent deep learning focused processors have made it possible to do computer vision on the camera instead of in the cloud.” It’s rare for people to identify a customer problem that nobody has heard about before–usually, what creates the potential for a large new company is that technology is now available to solve that problem in a new or differentiated way.

I also want to see a slide about solving the “Cold Start” problem. This matters most for the earliest stage companies, but AI companies need access to data to start training their algorithms. I like to see that they’ve thought about this problem and have a clear way to get access to enough data to build their business. The answer can be anything from buying data to partnering with an ancillary business to “faking it until they make it,” where they deliver the product or service with humans until they have enough data to build the AI model.

Mark Rostick, who is a Vice President and Senior Managing Director of Intel Capital:

When looking at a presentation of a potential AI deal, we look closely at the specific problem in AI/ML that they are solving and why solving that problem is important enough to build a company—not just a feature or tool. We also take a look at why the team is uniquely positioned to understand the problem they are trying to solve and how they are equipped to execute on it. The team must have line-of-sight to an economic model they can create that is capable of driving growth at “venture scale”.

Jake Saper, a partner at Emergence :

When evaluating companies that use AI to augment workers, I like to see charts that show the percentage of AI-generated suggestions that are taken by the user over time. For strong companies, this portion may start relatively low as the model is training and the UI is being tweaked. As both improve, you want to see the “coaching acceptance rate” improve to >75% and stay consistent.

Kenn So, a venture capitalist at Shasta Ventures:

There are a couple of slides I like to see:

#1: A high level architecture/diagram of how the data flows from source to training to AI predictions to the product that the users interact with. This helps brush away some of the AI pixie dust.

#2: Quantifying the value of the product or model for the user. For example, radiologists save 1 hour per day because the AI automates report writing. Only 10% of radiologists make adjusts to what the AI writes

#3: Defensibility of the data now or in the future. There are different ways to achieve this from proprietary data rights agreements to data network effects. In the end, ML is all about data. One thing to note is that data defensibility is just a minimum and not a sufficient condition of defensibility for AI companies.

Jeremy Kaufmann, a principal at Scale Venture Partners:

One of the most valuable metrics to show investors when pitching an AI company is how the accuracy rate of the underlying algorithm is improving over time. It’s important that investors see this improvement over time, particularly if humans are in the loop, as this analysis points to the fundamental solvability of the underlying problem. Investors are scared of the potentially asymptotic nature of AI algorithms (that they will never get good “enough”), so it’s very important to define “good enough” in a business context (what a business user will accept in terms of error rate) and then overlay this underlying expectation with the quantitative measure of how an algorithm is performing over time.