Peter Yared wrote recently a BusinessWeek guest blog post called “Failure of Commercial Open Source Software.” Not surprisingly his post caused a lot of angry replies from people who work for COSS companies. “The emperor is not naked” they argued.
I believe that the COSS emperor is openly naked. And the discussion shouldn’t be whether COSS is a complete or a partial failure just because there are few successful exits that Peter neglected to mention. At the end of the day Peter’s comment that “selling software is miserable” is true. Every sales rep involved in selling COSS would agree (I’m interviewing many of them now). Selling COSS is no easier than selling any other form of software.
Any company using the word “open” should be able to explain the true cost of delivery (this is one of Peter’s points). And there is an obvious litmus test of openness of COSS companies: One that I would call “open pricing.” COSS companies should openly publish their price list and clearly mark what’s free and open and what’s paid and closed. Otherwise OSS is just a bait-and-switch to a familiar proprietary software tactic of customer lock-in. This is what OSS was supposed to get rid of in the first place.
Let’s take a look at some of COSS companies in the Business Intelligence space. The bait and switch is in a full swing here:
We announced GoodData pricing earlier today and I would actually argue that we are a more open company than any of companies listed above. Our customers know exactly what service they get and how much it will cost.
We stick to our company motto: GoodData = BI – BS. And at there is a lot of BS going on in COSS space. It may actually be its biggest failure.
Full disclosure: I have been a big believer in open source since we opensourced NetBeans more than 10 years ago.
A long time ago I came to the conclusion that “independent industry analyst” was an oxymoron. But the willingness to sell independence for cash reached a new low with TDWI’s New SaaS Business Intelligence Portal. Please visit the link and see if there is any trace of independence left…
Moore’s Law states that computer system performance/price ratio will double every two years. And that was very much my expectation when GoodData started using Amazon Web Services almost 2 years ago. But I had to wait until today to see Moore’s Law at work: Amazon announced 15% drop of EC2 prices. The price of the small Linux instance was constant at $0.10 per hour for the last two years – now it will be $0.085.
15% in 2 years – not exactly the exponential growth in the performance/price curve that I expected. And I started to wonder why. Here are my two explanations – I believe the second one is more likely:
AWS prices were set way too low to attract developers two years ago. Moore’s Law helped the price to catch up with the real cost of running the cloud.
AWS is a monopoly and Moore’s Law does not apply.
What? Cloud and monopoly? Isn’t utility computing a perfect example of fiercely competitive commodity where the price curve is shaped only by demand/supply? What would Nick Carr say? Unfortunately not. As much as we read about different cloud providers, AWS is the only real provider of “infrastructure as a service” in town. If you don’t want to be locked-in to proprietary Python or .Net libraries there is not that much choice.
Until we will see performance/price of AWS double every two years, we should still wonder about monopolistic pricing.
Back in the old good days of enterprise software, we did not need to worry about our customers. We delivered bits on DVDs – it was up to the customers to struggle with installation, integration, management, customization and other aspects of software operations. We collected all the cash upfront, took another 25% in annual maintenance. Throwing software over the wall … that’s how we did it. Sometimes almost literally…
I now live in the SaaS world. My customers only pay us if we deliver a service level consistent with our SLAs. We are responsible for deployment, security, upgrades and so on. We operate software for our customers and we deliver it as service.
But there now seems to be a new way how to “throw software over the wall” again. Many software companies have repackaged their software as Amazon Machine Image (AMI) and relabeled them as SaaS or Cloud Computing. It’s so simple, it’s so clever: Dear customer, here is the image of our database, server, analytical engine, ETL tool, integration bus, dashboard etc. All you need it is go to AWS, get an account and start those AMIs. Scaling, integration, upgrades is your worry again. Welcome back to the world of enterprise software…
AMI is the new DVD and this approach to cloud computing is the worst thing that could happen to SaaS. And SaaS in my vocabulary is still Software as a Service…
Terry Pratchett once wrote that “Gravity is a habit that is hard to shake off”. We could make a similar comment about the financials of SaaS BI companies. As much as startups in this field would like to shake off their bad economics, reality always catches up. We’re seeing one after another SaaS BI startup to go out of business. Back in June it was LucidEra and earlier this week Blink Logic ceased operations. But anybody who only briefly looked at Blink Logic’s finances (it was a public company) shouldn’t be surprised by this event.
Why do so many of the attempts to marry BI and SaaS fail? The problem is that Saas BI sounds simple … simple enough to take an existing BI asset (integration engine, open source analytical engine, columnar database, dashboarding, even domain expertise & consulting) and just host it! All it takes is VMware or an AWS account, web server and Flash or JavaScript. Some people call this a paradigm shift, I call it window dressing. LucidEra was essentially restarted Broadbase, BlinkLogic was once called DataJungle, PivotLink recently changed their name from SeaTab, Cloud9 Analytics has a secret history as Certive, Success Metrics morphed into Birst. I could go on…
Why do SaaS BI companies have bad economics? It’s an attractive market – one of the last few open spaces in software. BI requires dealing with lots of data, lots of compute power and many users. SaaS + BI seems obvious. But truthfully, it’s such a difficult opportunity that it requires a new approach, yet everybody is taking shortcuts. SaaS BI isn’t just hosted BI just as email is not just better faxing, wikis are not just simplified Microsoft Word. Some time ago I wrote a case study on how my former company, NetBeans, was able to successfully compete against giants like Symantec, Borland or IBM, this case study is very relevant to our SaaS BI discussion.
The SaaS BI paradigm shift needs to be truly transformational in order to be successful – something that will get BI above the 9% adoption flatline it’s been at for years. Not everybody gets this. One of the best analysts in this space Boris Evelson wrote a blog post earlier this week where he focuses on differentiation of SaaS BI startups. His first question is: VC backing. Is the firm backed by a VC with good track record in information management space? But LucidEra was very well funded by leading VCs. The correct question that Boris should have asked is: Are the backers of the company funding innovation? Do they understand that it takes three years to become an overnight success?
It’s not a shock to state that cloud computing will disrupt the business model of commercial software. But how it will affect the open source movement?
The rise of open source is clearly linked to the rise of the web. Buy a commodity piece of hardware, download source code of any of the thousands of open source projects and start to “scratch your own itch”. My Linux box will communicate with your Linux box as long as we stick to some minimal set of protocols. The web is loosely coupled and software can be developed independently in a bazaar style.
It’s not quite as straightforward in the cloud. Clouds are also composed of thousands of commodity PCs, but the cloud operator manages the overall architecture and deployment – power supply, cooling, hypervisors, security, networks and so on. We don’t rely on minimal set of protocols in the cloud. On the contrary the cloud is defined by fairly complex, high level APIs. Even though the actual cloud OS may come from the open source domain, the tightly coupled nature of the cloud prevents users from modifying the cloud software.
There’s a lot of talk today about setting up private clouds with an Open Source Cloud OS, but the idea of private clouds is simply a delusion. Since the owner of private cloud has to purchase all required HW upfront, private clouds don’t provide the main benefit of cloud computing: elasticity. Other people will claim that clouds are not compatible with the open source movement or call it outright ‘stupidity’.
I see two possible solutions to this problem:
Benevolent dictator: Leading cloud providers (Amazon, Google, MSFT) will open-source their complete stack. This means that they would let the community to inspect the code, fix bugs, suggest improvements and define a clear roadmap similar to the Linux roadmap. This will also require a role of benevolent dictator to manage the evolution of the cloud. Given the level of investment required to build and operate the cloud I don’t believe that this is likely scenario. The new PC: The open source community accepts the cloud as the new HW/OS platform. Instead of building apps on top of x86 platforms (Wintel, Mac…), open source applications would be built on top of Amazon Web Services or Google AppEngine APIs. And these apps would handle the portability of data so that data doesn’t get locked in the cloud.
At the end of the day, cloud computing equals utility and utility creates stability. And a stable set of APIs, protocols and standards is a great place for open source to flourish. The best open source projects grew on top of stable standards: MySQL/SQL, Linux/x86, Firefox/http/HTML. I wonder what will be the most important OSS that will grow on top of the cloud…
I am not happy to see LucidEra disappearing. It is not a good sign for the SaaS BI market in general and the startups in our space specifically. And I still believe Rob Ashe (IBM/Cognos) was wrong when he said that “BI doesn’t lend itself to SaaS”.
There are some fundamental differences between first generation SaaS BI providers and cloud-based platforms like Good Data. Some of them are technological while others are simply common sense:
We use Amazon Web Services to host our multitenant platform and so we have minimal fixed and very low variable costs.
We are true believers in Steve Blank’s Four Steps to the Epiphany, and the idea of spending over $20M before validating our go-to-market strategy is foreign to us.
Cookie-cutter pre-built analytics apps are should be the STARTING POINT for customers to try – not the conclusion of an enterprise sales process.
LucidEra was probably too expensive for small companies and too limited for large ones. And this is why we offer plain-vanilla NetSuite analytics for free.
I am sure we will see the era of success of on-demand analytics. The most useful analogy here is the disruptive business model of low cost airlines – it did not disappear after the demise of People Express airlines either…
PS. Good Data Offers Safe Harbor to LucidEra Customers (link)
Business Intelligence projects are famous for low success rates, high costs and time overruns. The economics of BI are visibly broken, and have been for years. Yet BI remains the #1 technology priority according to Gartner. We could paraphrase Lee Iacocca and say: People want economical Business Intelligence solutions and they will pay ANY price to get it.
Nobody argues with the need for more Business Intelligence; BI is one of the few remaining IT initiatives that can make companies more competitive. But only the largest companies can live with the costs or the high failure rates. BI is a luxury.
I believe that the bad economics of BI are rooted in the IT department/BI vendor duopoly on BI infrastructure. This post focuses on IT’s inability to deliver efficient BI projects; I will write about the BI industry in my next blog:
There are three fundamental reasons why IT departments in their current form fail to deliver economical BI solutions:
1) They don’t understand elastic scale
IT departments are good at scaling: adding more and more hardware and software but scaling makes sense for tasks that are highly predictable. Given the ad hoc nature of BI we not only need to increase the compute power when we need it for a complex queries but we also need to be able to decrease the compute power when it’s not needed to keep the costs down. Elastic is more important than scalable. And this precisely why internal BI solutions will always be either too expensive or too slow for complex queries…
2) They try to control BI with a single version of the truth
While the volatility of business environment is increasing the IT departments are trying to button up the business knowledge (data, metadata, processes) into a top-down, inflexible and lengthy process that should produce a single version of truth. The problem is that the underlying business is changing so rapidly that by the time this is done the resulting analysis and reports are not correct anymore and the BI project becomes shelfware.
3) They cannot measure success of BI
“If you can’t measure it, it’s not worth doing!” is one of the selling point of BI but it is difficult to measure the success of BI projects. IT delivers on initiatives that are quantifiable (throughput, response time, performance, data sizes) and since the data size is one of the few easily measured aspects of BI it is the only metric where IT can claim success. This is why we often read about terabyte and petabyte datawarehouses. But it is a small portion of the BI market (2%) and they happen to be places where data goes to die.
The analyst firm The 451 Group asks technology companies “Who else do you see in your accounts?” Being “seen in the account” is perceived as a sign of market presence and ability to execute. And the opposite is true as well. Not being seen in accounts is sign of weakness and lack of market penetration. It’s also a proxy for the longevity question: “Will they even survive if they are not on anybody’s radar screen.”
My perspective is completely different. I believe that being seen in the accounts of large competitors is a sign of confusion and a complete waste of time and money. Startups are best when they disrupt existing markets, not attack them head on. Any sufficiently disruptive technology should be first deployed in a market segment that is seen as secondary or completely irrelevant by the big guys.
Established companies often compete on feature/functionality depth — delivering more features at an ever diminishing rate of value to customers to extract more money from them. Clearly not an interesting place to be for a young company.
I would like to promise here to our large competitors: You will not see Good Data in your accounts if:
your customers believe in a single version of truth
you deal with BI and data warehousing “experts” who attend TDWI seminars
Inmon Vs. Kimball matters to you
your projects are measured in months or six figure dollar numbers
you engage in star-versus-snowflake schema debates
your product offers 30 ways to format a decimal number
producing 1,000 different reports a day is one of your product claims
I could go on and on. Simply put – every time we read that competitor XYZ doesn’t see us in their accounts, we consider it a small victory. We don’t want to be seen in your accounts. At least not until we are ready…
Systinet’s founding CTO and my friend Anne Thomas Manes pronounced the demise of SOA a few weeks ago. Honestly, SOA lost its meaning for me on the day when good, old Solaris became the “SOA operating system”. But is SOA dead or not? I don’t believe so but I think that Anne and others are looking for SOA in the wrong places. Here is why:
Part of our Systinet SOA pitch was this truism: “SOA is not something you can buy”. We believed that SOA didn’t come in a box and companies have to invest time and money to build it. And maybe this is the crux of the problem. What if the act of building internal service blueprint is beyond the capabilities and budgets of the individual customers? Go to the SOA mailing list and try to understand how to build your own SOA and you can spend the rest of your life reading the discussions and related blogs and comments.
My point is that IT departments will always spend most of their budgets keeping the lights on and there is not enough money left for a complete architectural redesign. And even if they decide to throw more money at it they will still not get it right because of lack of internal expertise, lack of vision and simply because it is too hard to rebuild systems that somehow “work”. Every company seems to have a set of requirements that none of the commercial products can ever satisfy and as a result the existing internal architectures are usually completely proprietary. And sediments of bad architectural decisions are nearly impossible to peel off…
Maybe it’s time to forget about this SOA delusion and look someplace else. For companies like Google, Amazon, Workday and others (including my company – Good Data) SOA is not only “yet another IT initiative” but the key differentiator that allows them to deliver a flexible and extensible set of services. And the only way IT departments will be able to “buy SOA” is to use services from the companies in the cloud. The role of proprietary internal architectures will diminish over time as companies move to an increasing number of on-demand services – and that is probably what Anne wanted to say when she declared SOA dead…
I spoke at the Entrepreneur Country forum at Institute Of Directors in London earlier today. The event was organized by Ariadne Capital (Disclosure: I am one of Ariadne investors). Here is the feedback from the audience via Twitter and below is my keynote:
Roman is founder and CEO of GoodData. He had the crazy idea back in 2007 that cloud computing could finally make good on the broken promises of generations of Business Intelligence Software. Roman has been a tech entrepreneur for almost 20 years. He was Founder and CEO of NetBeans (acquired ... Continue reading »
