Name: Dave Stephens
Location: Half Moon Bay, California
I am an enterprise software guy – someone who has a passion for building software for businesses. When other kids we’re playing video games I was playing around with a Word processor. What a geek!
I spent close to a decade building and growing Oracle’s Procurement Applications. These applications help the world’s largest companies buy the goods and services they need to run their businesses.
In 1998, I delivered Oracle’s first applications product to run solely on Apache and JServ, iProcurement. In 2000, I built Oracle’s first multi-tenant SaaS solution, Oracle Exchange. From 2002 onwards, as Vice President, I directed Oracle’s procurement investments and ran product strategy, product development, and quality assurance across 7 products: Exchange, Purchasing, iProcurement, iSupplier Portal, Sourcing, Services Procurement, and Procurement Contracts. The product line proved very successful, growing 7 times faster than market rates in my last year. But something was missing – I wanted to innovate faster and delight customers more often. So I left in November 2005.
Through early 2006 I began to follow the open source movement. That’s when it clicked – the more I learned about open source the more I became convinced that it represented the future of enterprise software. The reason is simple: open source products work better and cost less.
At the same time I saw a glaring hole – all Procurement software vendors were fighting over business at the world’s largest companies. As a result, their costs were going up and their software was growing more and more complex & unusable. No one seemed to be building a simple, quick, and affordable solution that would work not just for the big guys but instead for companies of all sizes.
So I co-founded Coupa, a company that uses commercial open source software to deliver a web-based purchasing system for small and mid-sized businesses. Our mission is to bring the power of effective procurement “to the masses”. Along the way, we discovered Amazon AWS for our SaaS-infrastructure and never looked back. Oh, and it’s a fun place to work! Check us out at http://www.coupa.com.
9 responses so far ↓
James Arnold // 01/27/07 5:33 PM at 5:33 pm |
Dave,
I came across your article about the different ways to price software and I wanted to ask you a question. I have recently been asked to give pricing for a software solution that I will be providing for a company of 10,000 people. I was wondering if you have tips on what the typical pricing and support packages (if any) for solutions for a corporation of this size. What is the typical range for something like that?
Thanks.
Dave Stephens // 01/27/07 10:42 PM at 10:42 pm |
Hi James,
Your question is intruiging but too open-ended for me to give a worthwhile response. Pricing is usually set by forces beyond any 1 company’s control (unless you are an Oracle, SAP, Microsoft, etc). Early on in new markets, companies can price software based on “value” but that goes by the wayside the moment competition comes in.. Email me at dave at coupa dot com with some specifics and I’d be happy to weigh in further..
Best, Dave
Will Donovan // 09/23/07 5:41 PM at 5:41 pm |
Hi Dave – I’ve been reading some of your stuff since I came across you in Manoj Ranaweera’s blogreel. I am really interested in your transition to enterprise-class open source software.
Manoj and I both operate in the accounts payable automation field – I am very interested to hear about some of your work and/or experience in open source enterprise solutions for invoicing and EIPP.
Would love to hear some of your thoughts – We almost exclusively work with corporate developers, but I’ve noticed a serious turn towards, at least, open source “thinking,” among many of our developer clients. Is this an air to the future?
Take care,
Will
Will Donovan // 09/23/07 5:43 PM at 5:43 pm |
PS – Manoj got us really hooked on the blog model – Here’s the blog we’re operating – http://www.paystreamvoices.com
Dave Stephens // 09/23/07 5:55 PM at 5:55 pm |
i think there’s an undeniably large amount of inefficiency in corporate development – and a questionable motive to build great software since most of what the customer uses is seen only after a sale. open source changes this – but often in so doing it takes away some of the financial incentive associated with building great products too. so i don’t have a black-and-white viewpoint on whether the world is “turning to open source”. it is and it isn’t – but at a minimum open source strategies are keeping closed source vendors honest in the market.
Will Donovan // 09/23/07 5:59 PM at 5:59 pm |
Interesting analysis – I am of the opinion that ultimately the open source philosophy will come to dominate the whole of software development, whether enterprise or not, even if actual open-source applications do not.
For example, many Google products are designed in-house, but then have enormous areas available for user-generated adaptation. That’s just one example, but I think it speak to the future of (at least) web applications. I wonder if you think, for example, if EIPP solutions could ever go the same way.
josephmartins // 12/2/07 4:57 AM at 4:57 am |
Hi Dave,
I know this is an old post, but I thought I should comment anyway.
What data do you have to support your claim that “open source products work better and cost less”?
The cost of licensing aside, software is software. Organizations still incur the costs of deployment, ongoing support, upgrade/change management, security, provisioning, migration, etc. In addition, if an application has been [or will be] altered or customized, one must include the ongoing development costs necessary to create and/or maintain the branch.
The question then becomes How much should an organization tackle internally, and how much should it outsource, based on cost and competency? This naturally leads to other questions about the type and amount of resources available.
Now, regarding opensource apps “working better”. What are your metrics? How do you define “better”? In my experience, opensource [like commercial] is a mixed bag – from well conceived designs with abundant support (e.g. products such as Linux and Truespace), to poorly designed and/or incomplete projects with virtually nonexistent support (take your pick).
I can’t say I’m any more satisfied with opensource than I have been with commercial applications – aside from the relief that I haven’t had to pay for sub-standard products [when they're opensource].
I’m not going to claim that commercial products trump opensource in cost or quality…there is no clear cut answer. But I will suggest that organizations proceed cautiously and do some homework before making assumptions about quality and cost.
nexusxyz // 10/9/08 3:57 AM at 3:57 am |
I quite like the Coupa concept but I have some observations:
• Firstly, there is interest in the ‘on-demand’ model but not from the perspective that proponents of these solutions think. What companies seem to want is an internalised on-demand capability – a hybrid model will probably emerge as large companies internalise the on-demand services model and smaller companies access services from a host (third party);
• Secondly, the Oracle offering based on the current generation of technologies is a challenge – only large organisations have depth or resources to set it up and then continually fund support;
• Thirdly – 70% to 80% of eProcurement initiatives fail to achieve their objectives in terms ROI and ERP/Finance/IT centric procurement initiatives are generally doomed to fail;
• Fourthly – solutions based on multiple modules are increasingly hard to integrate and maintain and the SOA is an unattainable holy grail with our current generation of technologies as the number of integration points rises – the SOA should be internalised in a fully integrated application (or applications) with a pure Internet architecture – there are a few out of these with this capability and it changes the TCO/TAC or software and finally,
• Finally, the above points point to the need for a new way of building applications – having humans involved in coding is laborious and is a significant cost element of application development. This is a failed model – application generation of software products from an engineering design is the way to go – easier to maintain a properties file rather than millions of lines of code. This can be done today.
Dave Stephens // 01/15/09 10:58 PM at 10:58 pm |
joseph: in theory open source apps should work better because of the transparency. it’s an assertion from my standpoint, not a proof & you’re right to call that out.
You must be logged in to post a comment.