The AI Paradox

Nobel Prize-winning Economist Robert Solow once quipped that “You can see the computer age everywhere but in the productivity statistics”. This famous observation has become known as the Productivity Paradox – the conundrum that economists face when trying to explain how is it that productivity in the US slowed down just as investments in information technology exploded in the 1980s and 1990s.

Today, we face a similar paradox, the AI Paradox. Just as corporations are investing billions of dollars in AI infrastructure, the actual impact and absorption of AI in the enterprise seems relatively muted.

Earlier this month, the MIT Technology Review published an excellent article entitled “This is why AI has yet to reshape most businesses”.  Some of the themes rhymed with our recent post about the potential for Applied AI companies, but also the real adoption barriers. It is striking that, although AI has been a central area of focus for may enterprises for years, actual adoption has lagged the hype. PWC recently conducted a survey of 1000 enterprises that are currently implementing or investigating AI and learned that only 20% plan for enterprise-wide adoption in 2019. Many projects are thus still stuck in pilot purgatory.

This dichotomy between promise and reality points to the power of Applied AI startup companies that focus intensely on the path to the adoption, and absorption, of AI into the enterprise.  Internally, we are calling this type of startup “AAA” grade – Absorbable, Applied AI.

Here’s why the approach of AAA startups matters so much in driving AI adoption::

  • First, as the article points out, the “initial payoff is often modest” of AI projects. Thus, we think it is important for Applied AI companies to “manage expectations along the way, as many Applied AI use cases falter based on overselling the potential and customers expecting too much when in reality improvements come incrementally over time”.
  • Second, the article mentions how if “companies don’t stop and build connections between such systems, then machine learning will work on just some of their data”, which strikes us as an approach a company focused solely on developing narrower, application-specific solutions can take as they can often derive insights gleaned from data that reflects “the real-world variance and dimensionality of the problem space” and can work aggressively with “input from domain experts to identify the logical gaps that exist and how to fill those gaps”.
  • Third, as “Ninety percent of the work is actually data extraction, cleansing, normalizing, wrangling”, this provides an advantage to third-party Applied AI companies, as opposed to internal DIY AI efforts, as the third party application provider–which are typically cloud-based platforms that scale across customers—is able to leverage this data wrangling cost across a wide swath of enterprises.  
  • Finally, the article points to several aspects that impede adoption including how end users need to be “attuned to how AI works and where its blind spots are” and how “In order for them to trust its judgments, they needed to have input into how it would work” which is why we feel it is critical for Applied AI companies to combine domain expertise with technical expertise, carefully collect the domain-specific requirements and use cases, use those requirements to “instantiate the models into a broader application that fits into customer’s broadly defined workflows”, and make transparency and explainability a core feature of the application, not an afterthought.  

If AI startups can apply these AAA techniques, they are going to continue to achieve superior customer traction as compared to their competitors. And, in doing, attract plenty of attention from AI-focused venture capital firms like Flybridge!

Applied AI: Beyond The Algorithms

One of the primary areas of focus for Flybridge over the years has been to be the first institutional investor behind companies looking to transform the enterprise technology landscape with modern software.  Given the explosion in the volume of data being generated globally, this theme has led to investments in companies such as MongoDB (databases) and Nasuni (storage) that operate at the data infrastructure layer of the enterprise tech stack.

More recently, we have been investing in further advances in data management, analytics, machine learning, and artificial intelligence.  While the potential for artificial intelligence has been written about extensively, what is less well understood is that the algorithms and underlying tools are only a fraction of the value and are unlikely to be a source of long-term differentiation.  Fully realizing the power of AI requires a deep understanding of the domain and the specific workflows that AI will seek to improve and optimize. In other words, the application layer of AI ultimately drives the business value. And we believe the window of opportunity for the AI application layer is now.

This evolution from platform to applications is not an uncommon one: when a new technology platform is immature and not well understood, there is a lot of room for innovation at the underlying technology layer, but as the platforms mature the application layer is where value accrues.  For example, in the PC era, once Windows and its associated tools were well developed, apps accumulated a huge amount of value; and similarly in the early Internet era, once the browser, server, and app server infrastructure were well established, apps became super valuable.

This “Applied AI” investment thesis resulted in four new investments by Flybridge in 2018: Aiera, which is using AI to drive fundamental equity analysis; Kebotix, which is using AI to discover and create advanced chemicals and materials; Looka, which is an AI-powered graphic design platform; and Proscia, an AI-powered digital pathology solution.  We also made follow-on investments in our existing, scaling, Applied AI portfolio companies Bitsight, Bowery Farming, Datalogue, and DataXu.  

Given the breadth of this emerging portfolio, we thought it would be helpful to expand on what we look for in Applied AI companies and some of the keys to success in building a company in this exciting field.  We believe that it is important to:

  1. Go Old School. Opportunities for Applied AI companies lie outside of the markets targeted by traditional web-scale companies.  As Andrew Ng recently observed, “a lot of the stories to be told next year [2019] will be in AI applications outside the software industry”.  Many of these markets, such as Real Estate, Finance, Healthcare, Oil & Gas, Agriculture, Manufacturing, and Logistics, have the advantages of being A) extremely large, B) where innovative AI driven approaches can drive massive levels of improvement versus the status quo, C) where potential customers may not be able to access AI talent on their own, such that build versus buy is less attractive, and D) not being an area of focus of the Googles and Baidus of the world where their massive troves of data can be a source of significant competitive advantage.  
  2. Combine Talents. The most successful teams in Applied AI will have a unique combination of an understanding of the domain and the technical capabilities to realize the vision.  Even more specifically, it is doubly helpful if one of the founders was formerly a practitioner in the field. For example, Dawson Whitfield, the founder of LogoJoy, was previously a top-notch graphic designer himself; Ken Sena, was a top-ranked equity analyst before founding Aiera, and Kebotix co-founder Professor Alan Aspuru-Guzik holds a Ph.D. in Chemistry and is a leader in the field of computational chemistry. In other words, AI experts will do better when teamed up with someone that comes from the field in which they are seeking to operate.
  3. Drive continued technical innovation.  Given how quickly the field is advancing, a deeply technical co-founder who is up to speed on, and willing to continually learn about the latest advances in AI, and see the application of new approaches to the problems their company is seeking to solve is essential.  Whether it is “few-shot” learning approaches, ensemble models, GANs, CNNs, transfer learning, explainability, and a myriad of other developing techniques, knowing and understanding the strengths, weaknesses, and applicability of different approaches is critical. We often see the domain expert mentioned in point 2) renting or borrowing their AI expertise in the form of advisors and part-time experts, but this approach is not good enough given the need to have a tight feedback loop between market-driven customer needs and the AI-driven technology insights and art of the possible.
  4. Create Data Network Effects. The most successful companies will have a clear understanding and angle on how to start and continue to spin the data network effects flywheel.  Generally, this requires having access to initial datasets that can begin the model building process, and a well thought out and focused strategy on how to increase the quantities of data available for analysis.  The initial data sets, which Proscia refers to as “inorganic data”, might be acquired, and are used to overcome the cold start problem of training a new model from scratch. In contrast, “organic data” that comes from the ongoing use of the platform can help hone and refine the algorithms over time.  Taken together, this means the cost of data should decline over time because organic data is typically free (or even negative if you can get users to pay you for the service delivered while the data is collected). In the pursuit of data, it is important to remember that the sheer volume is not always inherently better. Yes, size matters, but quality matters more.  The data should reflect the real-world variance and dimensionality of the problem space and a data strategy should incorporate input from domain experts to identify the logical gaps that exist and how to fill those gaps. Further, when assessing a technical team (per point 3), we believe it is important that they know how to build an AI infrastructure that can be monitored for changing performance and updated accordingly as the scale and scope of the datasets increase.  For example, when Aiera first started making buy-sell calls on stocks, they only did so on 16 companies based on a model that analyzed 10,000 documents a week from 300 data sources. Today, they cover nearly 2,000 securities with a model that analyzes 500,000 documents a week from 22,000 data sources. Perhaps not surprisingly, the accuracy, breadth, and duration of their buy-sell calls increased significantly over this time.
  5. Absorb The Algorithm. The specific algorithms and AI techniques themselves are not the sources of defensible value so the most successful AI companies will instantiate the models into a broader application that fits into customer’s broadly defined workflows.  We call this “Absorbable AI”, which means customers can incorporate the AI into their business and realize the benefits of the operational insights. Successful AI Applications need to not only explain why the model is generating certain results (and, importantly, explainability also helps understand and highlight bias such as gender-based or racial bias), but also integrate into the customer’s business in a logical and systematic way.  It’s also important to manage expectations along the way, as many Applied AI use cases falter based on overselling the potential and customers expecting too much when in reality improvements come incrementally over time. These application and process skills are often found coming from the more traditional application development space in fields such as UX, visualization, workflow/BPM, and integrations.
  6. Craft the Business Model. Thinking through the business model of a company is critically important.  Depending on the domain, the openness of customers to new approaches, their willingness to pay for innovation, and the scale of the level of AI being incorporated, the best way to realize the value of an Applied AI company could be by selling an application that makes human work more efficient and accurate–an end-to-end automation stack that replaces humans–or it could be by selling a complete product.  For example, our indoor farming portfolio company, Bowery, leverages a significant amount of AI to drive efficiency and quality in their operation, but they decided the best way to realize the value of that AI was to sell incredibly tasty, locally grown, pesticide-free green vegetables versus selling an AI-powered Farm Operating System to other growers. A similar example would be Tesla, where the vision is the sell a complete, AI-powered, autonomous vehicle, as opposed to say Cruise, which chose to realize the value in selling the application (and the company) to other automobile producers.  

Points 5 and 6 can be better visualized in the following matrix:

Slide1

With the continued explosion of data availability, advances in AI techniques, and the accessibility and performance of computing (GPU) cycles, we believe the trend of AI as the next great application enabler will continue for some time, and we look forward to finding more Applied AI companies with passionate domain experts and technical founders to invest behind in the coming year.

Thanks to my partner Jeff Bussgang, our advisor Harini Suresh, David West of Proscia, and Bryan Healey of Aiera for their input and insights in developing these thoughts.

Reflections on MongoDB’s IPO

MongoDB_Gray_Logo_small

A month ago our portfolio company MongoDB went public. The stock priced at $24.00 per share and closed the first day of trading at $32.07 per share. While an IPO is merely a financing event in the trajectory of a successful company and, in the words of the company’s super talented CEO, Dev Ittycheria, “NOT an end but rather a new beginning for MongoDB”, it’s still an important milestone and one worth celebrating and reflecting upon.

My association with MongoDB has been an absolute pleasure, having served on the board since Flybridge led a $3.4M financing in the company at a $12M post-money valuation ($0.66 per share) in October of 2009. In that month, MongoDB had 2,563 downloads for its nascent product. Today, it has had over 30 million. When we announced the investment, I wrote about what led to our decision to support the company: a phenomenal team, a large market with trends in the company’s favor, and a great product that customers loved. What strikes me upon re-reading that post 8 years later, is how simple, but true, that analysis was.

Since the beginning, under the leadership of co-founders Eliot Horowitz and Dwight Merriman, MongoDB has had one of the most talented, creative and driven technical teams I have had the pleasure of working with. Further, throughout the company-building process, the company benefited tremendously from the advice and guidance of Kevin Ryan, the company’s third co-founder and Chairman. Our market thesis was that the database market was large, growing (at the time, $30B in annual revenue; in 2016, $44.6B) and trending towards more special purpose solutions rather than the legacy, one-size fits all relational database model. This market insight has played out broadly. In addition to MongoDB, alternative datastores like Hadoop and its derivatives in the analytics market have thrived in the market. Finally, on the product front, MongoDB has continued to be loved by developers for it’s simplicity, flexibility, scalability and the fact it can run in any environment from the company’s database as a service offering, Atlas, to the cloud, on-premise or in hybrid environments.

What we did not write about publicly, but discussed in our internal analysis, were three additional observations:

  • MongoDB is a classic disruption story. When we were conducting diligence on the company, many of our friends and experts with deep expertise in the relational database market were quick to point out all the product’s shortcomings and the features it lacked. These objections failed to recognize that user’s desires to (1) develop software in a more agile, iterative manner; (2) deploy databases in horizontally scalable cloud architectures; and (3) utilize a product that was easy to access and allowed for immediate productivity gains all created benefits that more than compensated for the product’s supposed shortcomings at the time. Today, this “nice toy” of a database, as one of these experts called it, sees 30% of it’s new paying customers come from applications migrating off of relational databases (the other 70% comes from net new applications).
  • Developers are the new King of IT. When we first invested, it becoming apparent that the proliferation of software applications across all enterprises coupled with the rise of of the cloud, and more distributed architectures, was making the developer the new “King of IT”. This allowed, and continues to allow, MongoDB to go-to-market with a very developer-centric approach and then leverage this grassroots adoption into paying customers over time. I have written much about this developer adoption strategy in general, and as it applied to subsequent investments we made in companies such as Firebase, Stackdriver and Crashlytics (all successful investments and interestingly, all now owned by Google), but the approach of building a passionate user base prior to selling into a large enterprise has proven to be a successful one.
  • Land and Expand is a powerful business model. The flip side of a grassroots adoption first model is that when you do land a paying customer, you often land them for relatively small dollars. But, with a recurring revenue model and a product that delivers on its promises, over time these small customers renew and expand and this can build a large and growing recurring revenue base as shown in the chart below. In the case of MongoDB, this led to annual revenues that grew from $40.8M in FY 2015 (Jan) to $101.4M in FY 2017; in almost 300 customers that now spend in excess of $100K per year, up from 110 in early 2015; net ARR expansion rate of over 120% for each of the last ten quarters; and, annual cohorts that show, in the case of 2013 for example, 4.1x expansion over 4 years (i.e $5.3 million in FY13 grew to $22.1 million in FY17).

Of course getting the investment thesis right only matters if the company is able to execute. And the team at MongoDB has executed exceptionally well. Along the way, the founders were joined by Dev Ittycheria, who to no one’s surprise given his track record of success, has proven to be a remarkable, strategic, focused and results oriented, leader. He is also an exceptional recruiter and under his guidance the company has added well over 400 employees, including Michael Gordon, an extremely adept CFO (who after the IPO process is in need of a good night’s sleep), Carlos Delatorre, CRO who has built a world-class sales team, and Megan Eisenberg, CMO who has the unique talent in a marketing executive of being to drive both high level corporate marketing and a demand generation machine. Under Eliot’s leadership as CTO the company has also built a world-class, deeply technical, enterprise software team in New York City (which many thought was not possible, but it turns out to be a distinct advantage), including Cailin Nelson, SVP Cloud Engineering, Dan Passette, SVP Core Engineering, and Richard Kreuter, SVP Field Engineering. A hat tip to all of these executives, plus the 800+ other employees at MongoDB, on what they have built together.

Mongo celebration

It has been my distinct pleasure to work with such a talented team for the last 8 years. But, again, the IPO is just a financing. The company feels like it’s just getting started in its quest to disrupt this massive database market. I look forward to remaining on the company’s Board and continuing the journey for many years to come, building an important, anchor public technology company in New York City.

Fired Up By a Flybridge Family Reunion

crashfirebase-photo

Last week there was an interesting piece of news in the tech world when Firebase/Google acquired the Fabric product line and team from Twitter. Over 20+ years in the venture industry and hundreds of companies, this was a first for me: two companies we had invested in merged post their respective acquisitions by larger players. While it was unusual, nothing could make me happier than to see Crashlytics, which was acquired by Twitter in 2013, and Firebase, which was acquired by Google in 2014, join forces and continue the missions they held from their founding, and our original seed investments, of improving the lives and effectiveness of mobile developers. Huge congrats to Andrew, James, Jeff and Wayne and many thanks to both Twitter and Google for their support of both companies and allowing them to deliver on their common goals.

Rewind the clock to 2011. At the time my core investment focus was on developer-driven cloud platforms and an insight that companies that went to market with products that delighted developers could achieve significant adoption and break down many of the barriers seen by companies that focused on more traditional enterprise sales models. Within this broader theme (that also led to our investments in companies such as MongoDB, Stormpath and Apiary), it was clear at that time that the mobile developer was the new rock-star and thought leader and that most new application development spend was for mobile apps. But while mobile developers were leading the way, it was still too hard and technically challenging to quickly and easily get high quality apps into the market. Over the next year, this thesis led us to make seed investments behind two phenomenal teams. Both companies started out focused on very different markets – Crashlytics on crash reporting and Firebase on a platform to allow developers to easily build serverless back-end platforms. But both had a common goal of creating innovative technologies to help developers create amazing apps.

It’s always interesting to look back on your investment successes to see what if any common traits they shared. For Firebase and Crashlytics, there were many:

  • Founded by young, passionate entrepreneurs[i] who had strong technology backgrounds, startup experience and an innate understanding of their target customer who were
  • Creating platforms that addressed large and expanding markets with a
  • Special “developer first” approach to working with the developer community that led to rapid adoption of their platforms which led them to
  • Blow away their seed round metrics within less than a year and in turn raise Series A Rounds led by their initial investors and ultimately to being
  • Acquired by leading technology companies less than a year after Series A rounds and these
  • Acquirers provided significant incremental resources, let the teams run independently and continue to innovate under their own brands post the acquisitions

I could not be more proud of both of these teams. With millions of apps and hundreds of thousands of developers now using their technologies, what each has built is exceptional and I am 100% sure, will only get better as they join forces.

[i] It should also be noted in this time of anti-immigrant sentiment that in each company one of founders was born outside the US

Full-Stack Analytics: The Next Wave of Opportunity in Big Data

This post orginally appeared on KDnuggets in May of 2014 and came out of a panel discussion at Analytics Week in Boston that was moderated by Gregory Piatetsky of KDnuggets.  On the panel, I was asked to discuss where we see investment opportunities in the Big Data landscape and this post will expand on my comments.  The lens through which I make these observations is from our role as a seed and early stage venture capital investor, which means we are looking at where market opportunities will develop over the next 3-5 years, not necessarily where the market is today.

Over the past few years, billions of dollars of venture capital funding has flowed into Big Data infrastructure companies that help organizations store, manage and analyze unprecedented levels of data.  The recipients of this capital include Hadoop vendors such as Cloudera, HortonWorks and MapR; NoSQL database providers such as MongoDB (a Flybridge portfolio company where I sit on the board), DataStax and Couchbase; BI Tools, SQL on Hadoop and Analytical framework vendors such as Pentaho, Japsersoft Datameer and Hadapt.  Further, the large incumbent vendors such as Oracle, IBM, SAP, HP, EMC, Tibco and Informatica are plowing significant R&D and M&A resources into Big Data infrastructure.  The private companies are attracting capital and the larger companies are dedicating resources to this market given an overall market that is both large, ($18B in spending in 2013 by one estimate) and growing quickly (to $45B by 2016, or a CAGR of 35% by the same estimate) as shown in the chart below: 

BigDataMarketForecast2013

While significant investment and revenue dollars are flowing into the Big Data infrastructure market today, on a forward looking basis, we believe the winners in these markets have largely been identified and well-capitalized and that opportunities for new companies looking to take advantage of these Big Data trends lie elsewhere, specifically in what we at Flybridge call Full-Stack Analytics companies.   A Full-Stack analytics company can be defined as follows:

  1. They marry all the advances and innovation developing in the infrastructure layer from the vendors noted above to
  2. A proprietary analytics, algorithmic and machine learning layer to
  3. Derive unique, and actionable insights from the data to solve real business problems in a way that
  4. Benefits from significant data "network effects" such that the quality of their insights and solutions improve in a non-linear fashion over time as they amass more data and insights. 

A Full-Stack Analytics platform is depicted graphically below:

Full Stack Analytics Graph

Two points from the above criteria that are especially worth calling out are the concepts of actionable insights and data network effects.  On the former, one of the recurring themes we hear from CIOs and LIne of Business Heads at large companies is that they are drowning is data, but suffering from a paucity of insights that change decisions they make.  As a result, it is critical to boil the data down into something that can be acted upon in a reasonable time frame to either help companies generate more revenue, serve their customers better or operate more efficiently.  On the latter, one of the most important opportunities for Full-Stack analytics companies is to use machine learning techniques (an area my partner, Jeff Bussgang, has written about) to develop a set of insights that improve over time as more data is analyzed across more customers – in effect, learning the business context with greater data exposure to drive better insights and, therefore, better decisions.  This provides not only an increasingly more compelling solution but also allows the company to develop competitive barriers that get harder to surmount over time.  In other words, this approach creates a network effect where the more data you ingest, the more learning ensues which leads to better decisions and opportunities to ingest yet even more data.

In the Flybridge Capital portfolio, we have supported, among others, Full-Stack Analytics companies such as DataXu, whose Full-Stack Analytics programmatic advertising platform makes billions of decisions a day to enable large online advertisers to manage their marketing resources more effectively; ZestFinance, whose Full-Stack Analytics underwriting platform parses through 1000s of data points to identify the most attractive consumers on a risk-adjusted basis for its consumer lending platform; and Predilytics, whose Full-Stack Analytics platform learns from millions of data points to help healthcare organizations attract, retain and provide higher quality care to their existing and prospective members. 

Each company demonstrates important criteria for success as a Full-Stack Analytics company:

  1. identify a large market opportunity with an abundance of data;
  2. assemble a team with unique domain insights into this market and how data can drive differentiated decisions and have the requisite combination of technical skills to develop and;
  3. manage a massively scalable learning platform that is self-reinforcing.

If your company can follow this recipe for success, you will find your future as a Full-Stack Analytics provider to be very bright!

A return to the clouds

I was at the GigaOm Structure conference last week which provided me with an opportunity to reflect on trends in the cloud computing world and how my thoughts on the market opportunities and trends have changed since I last wrote about this market in depth almost three years ago.

Before looking ahead, it is worth revisiting that old post, which can be found here, as not surprisingly i got some things right and missed some things.  Observations that in hindsight I got right include:

  1. The advent of cloud computing is a fundamental architectural shift that will create numerous opportunities up and down the enterprise IT landscape.  This has, and continues, to come to pass in spades.
  2. Competing as a company in the public Infrastructure as a Service (IaaS) market is a fool's errand.  One only needs to look at the pace of innovation at AWS, which is awe inspiring, and their dominant market share, to realize that this too was correct.
  3. The cloud will be a huge enabler of innovation.  This in hindsight, is the most fundamental impact and has changed how entrepreneurs think about starting and building companies, not to mention how venture capitalists fund them.
  4. Latching on to developer communities, which I also discussed in depth here, has also proven to be a successful go-to-market strategy for many companies.
  5. Some of the specific ideas I discussed such as tools to better manage cloud environments (NewRelic, OpsCode come to mind) and SaaS applications for vertical markets (Jive, SuccessFactors) have also proven to be fruitful market opportunities for talented entrepreneurs.

Observations that, with the benefit of hindsight, I missed include:

  1. Enabling enterprise data centers to be more cloud like has not proven to be as fertile an area as I suspected.  Yes, VMware is seeing success with their software defined data center strategy, but as of yet none of the private cloud  platforms has broken out of the fray with significant revenue momentum.
  2. Bridging the cloud enterprise boundary has also not really developed into a significant market opportunity.
  3. The impact of mobility on cloud adoption.  No use case has driven cloud adoption the way the explosion of mobile devices has, and in hindsight, even three years ago this trend should have been identified more explicitly.
  4. The impact cloud scale data centers would have on networking technologies.  One only needs to looks at the >1,000 people that attended the recent Open Networking Summit to realize that the trend to Software Defined Networking and OpenFlow is gaining steam, although in my defense this was less evident in 2009.

So where does this leave entrepreneurs and investors on a go forward basis as we collectively look for opportunity in the cloud computing market?  A few thoughts:

  1. The public cloud is going to win.  Due to a combination of innovation at large companies such as Amazon, the OpenStack initiative that will enable Amazon's competitors and the fact that the vast majority of young companies developing innovative new solutions will be focused on the public cloud, it is going to be tough for private clouds other than at the largest of companies (JP Morgan and Zynga come to mind, or perhaps some highly regulated industries) to compete with the public cloud on a cost, performance or features basis.  This will, in turn, increasingly drive usage to the public cloud, in turn furthering the pace of innovation and differentiation for the public cloud creating a virtuous cycle.
  2. Driving efficiency will remain critical.  As cloud adoption continues apace, driving the efficiency of these large scale systems will be of paramount importance for cloud operators and their customers.  This means there is significant running room for companies not only that effect the labor side of the equation by driving efficiency in DevOps but also across other operating costs, especially power consumption (both in aggregate and in terms power usage effectiveness, or PUE.  As the CIO of Zynga said at Structure, "we need more innovation on power").
  3. Performance and response time demands from consumers will continue to increase.  The threshold of patience from any consumer of cloud services, whether they be delivered over the web or a mobile device, continues to decline rapidly.  Technologies that offer greater performance, or a path to developing applications with greater performance, to address this trend will see great demand.
  4. Leveraging developers to drive adoption remains critical.  As fundamental infrastructure architectural decisions get delegated into the cloud, developers increasingly will make critical application level decisions, so tools that increase their efficiency and/or offer new functionality that speed application development have the potential to see rapid uptake in the market.
  5. Look for ways to leverage the cloud to bridge data and applications onto mobile devices.  Just as Dropbox did for consumer file access, there are a myriad of applications and use cases that can harness the combined power of mobile and the cloud to provide unique user experiences.

I am sure another 3 years down the road, some of these insights will also prove to be off base, but let me know what you think or other opportunities I should be adding to the list!

 

Driving developer adoption

One of the investment themes we are pursuing here at Flybridge within our Dynamic Computing (which my partner David Aronoff covered in detail here) is the rise of developer driven business models.

Ten years ago architectural standards in large or mid sized companies were set by the CIO or Architecture review boards and the choices were fairly limited (Solaris, Linux or Windows for the OS; Oracle, DB2 or mySQL for the database; Weblogic, Websphere or jBoss for the app server, Java or .Net for languages, etc).  In today's world, on the other hand, the choices have expanded by a couple of orders of magnitude with the rise of open APIS (what Ross Mason calls the ninja slicing of the application stack) open source, virtualization, cloud computing and dynamic scripting languages.  This has empowered the developer to experiment, prototype and ultimately drive the adoption of new technologies in ways that were not previously practical.

As a result many companies are seeing great success in infiltrating large customers not with a top down approach marked by expensive sales people, but rather on with a bottoms up, developer led approach that demonstrates value on a project by project basis before the CIO even knows what is happening.  So what are the keys to success in driving this grassroots adoption?

The most important point is the simple one: your product needs to work and have a clear value proposition.  In other words the age old better, faster, cheaper still applies.  But assuming that, here are ten concrete suggestions for getting the developer driven business model to work for your company:

  1. Rapid time to value: Developers are always stretched and have more to do than time to do it.  Your product needs to show value fast.  Traditionally this meant less than 30 minutes to be up and running, although today this is compressing even further (one of our companies has a 5 minute register, download, install, integrate and run goal).
  2. Free to start: Have a fully functional, high value, free product that allows experimentation.  If your developer customer needs to pay to do anything useful, they wont adopt.
  3. Superior documentation and support: Related to point one, your documentation needs to be clear, easily searched and any support issues need to be addressed quickly and thoroughly.  Get everyone in your company into your support forums providing high value assistance.
  4. Leverage existing communities of interest: As an early stage company it is hard to create a movement on your own.  Tap into existing communities with momentum and where your potential customers can be found.  NewRelic, which out of the gates latched on to the growing Ruby community, has always struck me as one company that did this well.
  5. Buy your engineers plane tickets (or at least subway passes): Find where your developer customers congregate and get your team into the community to attend meet-ups, show off your product and explain what it does and why it solves real problems.  Overtime, generic meet-ups can become more targeted around your specific solution as we have seen withour portfolio company 10gen, and the highly successful events they put together for the mongoDB community.
  6. Facilitate word of mouth: Get your customers to speak on your behalf, either live or via their own blogs.  Share presentations and use cases via slideshare.  Contribute positively to HackerNews & Stackoverflow.  Work twitter.
  7. Have a CEO that can code: Street cred matters both externally and internally and respect in the community will build if the leader of your company has walked in the shoes of your customers.  If you are a CEO and are not a hard-core developer, fix some bugs, work on the front end code, and resolve support issues. 
  8. Don't market:  Developers are smart and cynical.  Slick marketing wont work.  Speak their language, solve their problems, add value first.  That gives your credibility to expand usage and ultimately get paid.
  9. Have a clear pricing model:  Historically people think developers don't pay for anything.  This is true, but their bosses will.  If you are adding value, people will want to pay (and in fact a company with out clear revenue model may actually see slower uptake as Sean Ellis wrote about here), but they will need to know on what basis they will be charged so they can understand the longer-term implications of working with you.
  10. Help your early adopters sell internally: Once you have some champions at a customer, especially in larger organizations they may need to convince their bosses that what they want to do is the right idea.  Be ready to arm them with information on security, performance, integration points and other similar implementations to make this easy for them.  

This is my top ten list, would love to hear other thoughts on what works, or does not work, for you.

Quick reminder of the challenges (and opportunities) in enterprise IT

Last week I did several reference calls on a new project we were looking at in the enterprise IT market.  In each case I was speaking with a senior IT executive at a large financial institution.  The calls were not to the company's existing customers, but rather to people I just felt would be thoughtful on the company's market opportunity.

As it turned out, each had actually evaluated the company's product and they started out the conversation by saying how impressed they were with the company's technical approach, how it was really innovative and superior to other solutions, and how much they liked the company's management team. So far, so good.

Then I asked the obvious next question: so did you buy the solution?  In each case, the answer was no. The why behind the no, however, is what is illuminating on some of the recurring  challenges for traditional start up enterprise IT companies.  Although they used different words, the consistent themes were:

  1. We went with our incumbent vendor.  We know their solution is not as good, but it is good enough and it is already integrated into our infrastructure and we know how to work with them.
  2. This project, while important, was not high enough on our priority list to get everyone's attention.  Given tight budgets, the only new initiatives we are doing relate to driving revenue or addressing pressing regulatory changes in our business.
  3. The company wanted to do a paid pilot with us to prove how much superior their solution was, but given a pilot required standing up new iT infrastructure and integrating into some of our core systems, this was a significant undertaking and not something we were willing to do given point two above.  Further, the pilot was required to demonstrate the business case which left us in a Catch-22 situation.

Many of the takeaways from this conversation are the same as I wrote about two years ago in my long enterprise IT post, but I felt they are worth repeating in the light of this conversation.  First, if your product/market focus requires a traditional enterprise IT approach, which for me means high touch direct sales with a product that requires some level of IT support and integration, you better make sure that:

  1. Your "superior solution" is significantly superior.  As in 10 times better, not 50% better.
  2. You are focused on a company's top three strategic initiatives, because nothing else is going to get funded in amounts or in a timeframe that will make you happy.

As you have likely figured out, these two things are hard.  In the fast paced technology world, sustainable 10x product differentiation is uncommon and fitting that with the ever-shifting strategic priorities of large organizations is even harder.

The alternative, of course, is to think creatively about ways to reduce the friction in the adoption process which is why we continue to be attracted to open-source, freemium and SaaS business models.  Instead of the conversation I had above, imagine if the customer was able to download/ sign up for the company's product -  without consuming any IT resources or getting into a lengthy procurement cycle – and then use the product and see its value.  If the product was easy to use and understand, did not require deep integration or if it did, the integration came "pre-built" via some partnerships, the potential customer would have been able to develop its business case and pre-qualify themselves as a relevant customer.  Thus they would consume minimal resources from the vendor until they raised their hand and said, I am ready to buy.  While this is by no means easy, if you have the right focus across your company in terms of product management, development and sales and marketing, it is likely an easier approach than trying to find that magical 10x better product that meets your customers top strategic priorities in an era of constrained budgets and shifting priorities.


Welcoming 10gen to the Flybridge portfolio

While I try to avoid promoting Flybridge portfolio companies on this blog, from time to time I will write about our new investments in the hope it sheds some light into how we are thinking about opportunities. 

Today it was announced that we recently invested, along with Union Square Ventures, in a New York City based company called 10gen. 10gen was founded by Dwight Merriman, Kevin Ryan and Eliot Horowitz and they are an open source software company that developed the non-relational database, MongoDB. 

We invested in 10gen for a few reasons:

  1. Strong founding team: Dwight was the co-founder and CTO at DoubleClick, Kevin was the President and CEO of DoubleClick and Eliot was a lead software architect at DoubleClick.  As a result, the team knows a few things about both building large scale web infrastructure and growing and scaling businesses.  We were also fortunate, in a prior life, to have been investors in DoubleClick so we also had the benefit of knowing the team well.  
  2. A belief in the "NoSQL" database market: while the name is a bit of a misnomer, we believe that databases are getting more specialized over time as users realize that the one size fits all relational database model does not often fit what they are trying to achieve with their applications.  Specifically, a schema-free, non relational document oriented database such as MongoDB is particularly well suited to web applications where scalability, performance and flexibility will be highly valued. Virtually all the web developers we spoke to in our diligence were using, either in production or in their labs, NoSQL databases with very positive results.  We expect this trend to only increase as more applications are deployed in Cloud environments as these databases are particularly well suited to that architecture.  
  3. MongoDB is a leading solution in this emerging market: while MongoDB was only recently introduced into the market after a few years of development, the product has been downloaded by tens of thousand developers and is in production at companies such as SourceForge, Business Insider and Disqus.

While there are other reasons we, in short, liked the team, the market opportunity and the company's specific solution.  We look forward to working with this great team to take advantage of this exciting opportunity.  If you have an interest in downloading the product, please click here, or joining the team, please click here.

The Rebirth of Enterprise IT

(This post pre-dates my own blog and first appeared on my partner Jeff's Seeing Both Sides.  While it is now dated – the first post was from September of 2007 – many of the thoughts still ring true so I thought I would include it here).

When Nicholas Carr wrote his now-famous Harvard Business Review article over four years ago, “IT Doesn’t Matter”, the most damning claim to our industry was that IT had become a commodity input – irrelevant as a source for strategic advantage. Many pundits, from Larry Ellison on down, began pontificating on the maturation, consolidation and eventual death of the enterprise software business – at least for companies whose names are not IBM, Microsoft, Oracle, SAP or Symantec.

The general thesis goes something like the following: 1) corporate IT departments are looking to reduce, not increase their number of vendors and are therefore not inclined to work with start-ups; 2) customers no longer are pursuing best of breed strategies, but instead want integrated suites to simplify deployment and operations; 3) the sales and marketing costs of large enterprise software solutions are extremely high and drive a need for significant investments that are beyond the capabilities of many early stage companies; 4) the overall rate of growth of the software industry as a whole has slowed and there are few areas for innovation. Common analogies used by these pundits include the maturation and consolidation of the automobile and railroad industries in the early to mid 1900s. Pretty depressing stuff.

In the last six years, many venture capitalists are submitting their own vote on this debate with their feet, as the percent of funding dollars to software companies has declined from 25% of all venture disbursements in 2001 to 19% in the first half of 2007. Anecdotally, when you walk the halls of VCs around Sand Hill Road and Route 128, you hear a similar refrain: “We’re diversifying away from software… we are experimenting with consumer-driven business models… we like Web 2.0/new media plays”.

So where does that leave a talented entrepreneur (or VC, for that matter) with deep experience in this now passé field? While challenges remain, we submit that there remain numerous glimmers of hope in the enterprise software market – and certainly the recent reopening of the IPO market and the more robust M&A environment has brought some of these to light. If you look at some of these recent successes, themes and strategies emerge that entrepreneurs can adopt to drive the creation of successful companies:

  • Innovate to drive efficiency. For many times over the last decade, enterprise software companies positioned themselves as automating certain functional departments of corporations. First it was manufacturing, then financials, supply chain, sales, marketing etc. If this is your view of the enterprise software environment, then by and large Larry Ellison is right – there is little room for new categories and innovation. That said, if you spend time with the average CIO, you will hear a different story. In today’s “post-bubble” environment, CIOs have seen their staff and capital budgets cut back, but the demands on their organizations from business executives have continued to increase as companies seek to have a more flexible and cost-effective IT organization to support their business plans. CIOs have gotten their much sought-after “seat at the table”, but with that seat comes the pressure of accountability to deliver bottom-line results. Compounding this challenge of doing more with less is the sheer magnitude of the accumulated applications and technologies that have been deployed by enterprises over the last 20 years. The number of lines of code, disparate pieces of software, and points of integration has exploded exponentially. As a result, there remains a robust opportunity for focused vendors to drive innovative technology into enterprises to drive efficiency in IT operations. The bar, however, is quite high. If you can’t drive a 5 to 10 times reduction in key metrics, the status quo will prevail. A recent success story is Bladelogic, which went public in July of 2007 and trades at 13 times trailing twelve moths revenue, primarily due to the company’s success in automating data center operations, a key means to drive efficiency in IT operations. Opsware, which HP just agreed to acquire for $1.65 billion, is another example and also demonstrates there is a relatively healthy M&A market, as these innovative companies fill key product gaps for large acquirers, such as IBM, Microsoft, Oracle, HP and EMC, as well as mid-sized public companies such as BMC, CA and Symantec.
  • Wrap your software in commodity hardware. One of the complaints you will often hear from IT departments about working with a new vendor is the challenge of integrating their solution into their already complex environments. The mundane, manual tasks of requisitioning and provisioning the necessary hardware to run, or even pilot, the shiny new piece of software slows the path to adoption. As a result, a number of innovative software companies don’t appear at first blush to be software companies at all. Instead they sell pre-provisioned, plug and run, hardware appliances. Companies that adopt this model are not only able to leverage Moore’s law to drive performance, but also can ship their customers a unit that can be slotted into a rack and up and running in hours, not days. This allows customers to trial the solution and see the benefits immediately, mitigating the long sales cycles that plague many traditional enterprise solutions. Further, the appliance approach tends to lead to easier adoption by channels that are better suited to selling hardware than complex software. This appliance strategy was seen initially in the security software industry, but has since spread to other areas such as storage back up solutions from companies such as Data Domain, which recently went public and currently commands a $1.4 billion market capitalization on trailing twelve months revenue of $76 million.
  • Dominate a niche. Start-ups are often caught in a quandary. To raise money and hire the best people, they need to convince VCs, employees and other supporters of the company of a big vision and the opportunity to capture a billion dollar market. To do so, however, they run the risk of going too broad, too quickly and losing the laser focused approach that allows young start-ups to win against large, incumbent vendors. A better strategy is to instead think about climbing a staircase. You know you want to reach the next floor, but you don’t do that by trying to jump up 13 stairs all at once. Ask yourself, “What can I uniquely do today for a customer that solves a real problem and also provides a link to doing more things for those customers in the future?” In today’s age of rapid development, componentized software and offshore resources, software code is relatively easy and cheap to write, and is no longer the “barrier to entry” and source of competitive advantage it was ten or twenty years ago. Instead, what matters to customers (and potential acquirers) is the deep, domain-specific knowledge instantiated in that software. For an early stage company to build this knowledge, they need to be incredibly focused in a given domain and make sure they have people on their team who understand a customer’s business better than the customer does themselves. Unica, a recently public $80 million in revenue marketing automation company in Boston is a good example of this. When they first got going, they had the best data mining tools for marketing analysts on the planet. Not a huge market, but one that valued innovation and provided a logical steppingstone to campaign management, lead generation, planning and the other marketing tools that the company sells today.
  • Explore SaaS (software-as-a-service). If the key barrier to success for early stage enterprise software companies is excessive sales and marketing costs, adopting a software-as-a-service model may be the right approach. This is more than just selling your software on a subscription versus perpetual license basis. Instead, SaaS is all about making it easy for customers to understand, try and, ultimately, gain value from your software. In 5 minutes and for no up front cost, I can become a user of Salesforce.com. Within the 30 day trial period, I can self-qualify and decide if it is the right solution for me and worth the on-going subscription cost. Most importantly, I can potentially do this without consuming a single dollar of their sales and marketing spend. None of the airplane trips, four-legged sales calls, custom demos, proofs of concept or lengthy contract negotiations that lead to the 6 to 12 month sales cycle that costs a traditional software firm 75% of their new license revenue in a given quarter.
  • Consider Open Source. Open-Source is not about free software, but rather products that have seen, or have the potential to see, widespread grassroots customer adoption. A passionate end-user community has the benefit of driving a development cycle that quickly surfaces key product requirements and needed bug fixes. Further, the grassroots adoption of the product provides a ready installed base of early adopters who will promote the product across their enterprise, purchase professional services and acquire more feature rich versions of the product. Like SaaS, this is a way to mitigate high sales and marketing costs. When My SQL looks for customers for the enterprise version of their open-source database, they have to look no further than the estimated 11 million active installations of their software or the 750,000 plus people that subscribe to their email newsletter. RedHat’s version of Linux, Jboss’s version of the application server and Sugar CRM are three other well-known open source success stories, but other opportunities abound.

Enterprise software entrepreneurship and investing is certainly not for the faint of heart, but when pursued with some combination of the strategies above, we believe interesting opportunities remain for innovative companies to make their mark in the world and have a positive impact. Contrary to the claims of many, it is still possible to build these companies in a relatively capital efficient manner. Sticking to some of the examples cited above, it is illuminating to note that Bladelogic raised $29 million of venture capital before its IPO, Data Domain $41 million, Unica $11 million, Red Hat $16 million and Jboss (pre-acquisition) $10 million. Only Salesforce.com raised a lot of capital – $64 million – although almost 75% of that came in their last round when one would assume there was evidence the model was beginning to work.

In the end, we believe the analogy to the automotive industry is flawed. The manufacture and distribution of cars is fundamentally different from the software industry. In auto industry, there are tremendous benefits of scale, the underlying platform (tires, chassis, internal combustion engine, frame and skin) has remained the same for decades, and there is little room for small players to access end-users. Software, on the other hand, is a digital good and an information business. Innovation is limited only by the creativity of the author. Small teams can be extraordinarily productive – often times more so than larger teams and organizations. The underlying platform and architecture has changed several times in the last 30 years, and there is no physical product to distribute, thus end-users can be accessed much more directly. Is there a benefit to the incumbency and distribution might of IBM, Oracle or EMC? Absolutely. Does that mean there is no place for creativity, innovation and entrepreneurship in this industry? Absolutely not.