Why I love and fear AWS

The AWS launch of Amazon Connect (see techcrunch article) got me thinking about the current state of play in SaaS. Amazon Connect is a call center in a box, the same tech it uses in-house for their current platform. With that release, companies like Talkdesk and others have much to fear. While I see partnerships with companies like zendesk, salesforce and freshdesk to integrate voice with chat and email, I also firmly believe that it is just a matter of time before AWS continues to extend outward and deploy their own chat/email customer support system to go after their partners. Trust me, it will happen.

I fully acknowledge and love AWS for the opportunity to fund so many amazing founders who are fully leveraging the power of the cloud platform and services. What I also greatly fear is that Amazon and AWS have proven that they are amazing at taking markets that become hyper competitive and just blowing them up overnight with the lowest cost and good enough offering. AWS has also proven that it will continue to move upstream in the stack from the pure infrastructure layer to the application layer.

Here are a few examples:

  1. Amazon Quicksight (launched 10/15) –  fast, easy to use business analytics at 1/10 the cost of traditional BI Solutions
  2. Amazon Chime (launched 2/17) – frustration-free online meetings with exceptional audio and video quality – companies like gotomeeting (Citrix) made a smart move selling to LogMeIn
  3. Amazon Workdocs (1/15) – fully managed, secure enterprise storage and sharing service, users can comment on files, share, etc – box, dropbox watch out

Read More

Apps sell infrastructure

Pivotal just announced it did over $270mm of revenue in 2016 from Cloud Foundry helping large companies with digital transformation. That’s some nice growth from the $115mm the year earlier.

The initial Pivotal Cloud Foundry sales pitch was that it gave big companies a way to build new applications that run in a public cloud (rented space on Amazon (AMZN, +0.47%) Web Services, Google (GOOG, +0.34%) Cloud Platform or Microsoft (MSFT, +0.62%) Azure) or private cloud (flexible infrastructure that runs in a company’s own data center.

The need for faster, better software deployment resonated with older companies facing competition from smaller, newer rivals that already use cloud computing. You could argue, for example, that Hilton (HLT, +0.05%) and Hyatt (H, -0.47%) hotels should worry more about Airbnb (AIRBNB) than about each other.

This is yet another sign how large companies are embracing cloud technologies and microservices to be more agile. At the end of the day, it’s not about buying Cloud Foundry because of infrastructure savings, its the ability to quickly and scalably deploy new applications quicker to meet business needs. That’s the bet Pivotal made many years ago, and it’s paying off.

Remember if you are selling infrastructure – stop, sell apps to the heads of business who have a huge sense of urgency to get things done. Most of them also have pretty sizable budgets as well. The byproduct of all of this is saving money but that is not what moves the needle.

Read More

Developer love vs revenue Going from Seed to Series A

Great blog post by CockroachDB on open source business models and their plans to make money:

If you’re serious about building a company around open source software, you must walk a narrow path: introduce paid features too soon, and risk curtailing adoption. Introduce paid features too late, and risk encouraging economic free riders. Stray too far in either direction, and your efforts will ultimately continue only as unpaid open source contributions.

I would say same goes for any developer-focused company whether OSS or some other hybrid free/premium model. It is truly an art form when it comes to striking that steady balance between developer and community love versus generating revenue and potentially alienating those who supported you.

This is also an important question as it relates to fundraising for dev-focused startups. Introduce your pricing page too soon and that is the metric that Series A investors will track religiously. Bet the farm on developer love and metrics only and you may never get enough traction to get to that next round.

From what I have seen in our portfolio, goal #1 is always to build an amazing community, focus on developer love and track the metrics and tweak. Without the developers, you have no customers.

Read More

The Googlization of IT

Today I took a sales team from a portfolio company to meet with a couple of senior IT executives at a major retail company.  Towards the end of the meeting, it started to become quite clear to me the effect that Google and the web has had on IT to date and where it was going. In an oversimplified way, it seems that there have been 3 distinct phases to how the web and Google have impacted the enterprise, first starting at the app layer and increasingly diving deeper into the core infrastructure.

Phase 1 – Consumerization of IT – all internal corporate users are consumers first and then employees second.  we have all seen how consumers have gotten used to using browsers and SAAS-based applications and how successful startups have been able to provide web-based applications that users can pull into the enterprise environment starting at a department level rather than having to go out and sell and push technology into enterprises.

Phase 2 – Rise of open source – I would call Phase 2 the rise of open source software over the last 10 years – most of which is hardcore infrastructure type software such as databases, virtualization software, and the like.  IT folks leveraged the web and Google not just for applications but also to download core software to help run their internal operations.

Phase 3 – Googlization of IT – have as much of your infrastructure as you can run like Google's – distributed, commodity-based, and in the cloud on a private basis.

Phase 1 and 2 are ongoing and Phase 3 is where I see a few of the more forward-thinking IT departments I have met with over the last few months going.  I am not just talking about Google Apps (like email, etc) but about how companies can run their infrastructure internally like Google.  If Google can deliver a number of highly scalable web-based apps by clustering commodity servers, then how can enterprises do the same for themselves.  This is not about getting sucked into buzzwords on the cloud but really understanding the cost savings and performance benefits a company can get from transitioning some of their infrastructure to a Google-like model. 

One company in my portfolio that is leading the charge in the data warehousing space is Greenplum.  A customer can buy our data warehouse, cluster commodity servers like Google, and get petabyte scale and much better performance for less than the cost of maintenance of many existing solutions on the market today.  In addition, large global companies can have these nodes accessible to anyone anywhere in what we call the Enterprise Data Cloud.  One of our large customers said that data was a strategic weapon and that he wanted to make the cost of a running a new query zero.  In today's world and without the enterprise data cloud initiative I can tell you that running new queries in a global organization is an expensive and time consuming task of replicating data, creating data marts, running the processes, etc that can take months to get going and days to run reports.  Another company in which I am an angel investor is called Eucalyptus Systems whose tagline is your hardware, your data, your cloud. Eucalyptus is an open-source system for implementing on-premise private and hybrid clouds using the hardware and software infrastructure that is in place, without modification.  Eucalyptus adds capabilities such as end-user customization, self-service provisioning, and legacy application support to data center virtualization features, making IT customer service easier, more fully featured, and less expensive.  Yes there are public clouds like Amazon EC2 which is now also offering virtualized private clouds.  But the reality is that many large IT organizations want to control their own data, find ways to make it more easily accessible to everyone, significantly reduce infrastructure costs, and be able to launch new apps or services quickly and cheaply.  This is where I believe many IT organizations will be headed in the next 5 to 10 years creating private and hybrid clouds for existing and new applications, a phase which I call the Googlization of IT.

Data wars heating up – Microsoft buys DATAllegro

As I have written in previous posts, what you do with data will be one of the next battlegrounds on the web.  Knowing that they had some limitations with SQL Server, Microsoft announced its acquisition of DATAllegro (full disclosure: my fund is an investor in competitor Greenplum) to enter the data warehousing market.  Enterprise volumes across the board are ramping up quickly and this clearly gives Microsoft an opportunity to capture that market.  Being an investor in Greenplum, I always like to see healthy exits of competitors as many believe it will trigger further consolidation.  When a competitor is acquired, the first reaction from many is often asking themselves why it wasn’t them and fear about competing with a juggernaut, but my perspective is quite different as it usually opens new opportunities.  As I have written before, many acquisitions fail and companies are usually so distracted for the first 6-12 months trying to integrate operationally and technically, that this gives others in the market a nice window to continue executing on their business plan.  So I tip my hat to DATAllegro and look forward to an exciting 12-18 months ahead as the data wars are clearly heating up now. 

Picks and shovels for the web

We have had quite a resurgence in the web market during the last few years.  A number of great companies have come out of nowhere to become household names, and it seems that everyday we are inundated with news on another slew of new web startups going after the consumer.  And yes, looking for the next YouTube or Facebook or Myspace is exciting.  Depsite all of that, the one area is that is not discussed much is the boring infrastructure market where companies sell the picks and shovels to allow these startups to run their operations.  And what could be more boring than talking about a database or data warehouse?  Anyway, I am glad that Don Clark of the Wall Street Journal wrote a nice article on a new breed of startups going after the database market.  Shamelessly, I would like to add that he has a nice writeup on Greenplum (full disclosure: my fund is an investor and i am a board member).

Granted, the opportunity to make money selling picks and shovels during this web resurgence is definitely much harder as developers typically go for free and cheap software and hardware to launch their new companies.  That being said, every click that we make is being stored somewhere and the companies who can better analyze this data to better monetize their sites will be the winners in the next phase of the web.  This is where Greenplum comes into play.  The company is not only playing off of the data volume and analysis trend but also the move towards commiditization.  As Don mentions in his article, the secret sauce is that our customers can deploy massive data warehouses using our software which is built on top of the open source database Postgres and deploy it on commodity boxes.  The benefit is not only in terms of cost but also in significant performance increases over the competition.  As per Don’s article today:

One user is iCrossing Inc., of Scottsdale, Ariz., which provides analytical services to companies that operate Web sites. Analyzing a day’s worth of some types of data once took 20 to 22 hours, said Tony Wasson, the company’s vice president of engineering. With Greenplum’s technology, and some modifications to its own software, the job now takes about an hour, he said.

Anyway, it is nice to see the mainstream press finally getting the fact that data and analytics matter. Yes, plumbing is boring, but without cost effective platforms which can scale and perform under heavy stress, we won’t be able to reach the full peak of monetization on the web.