Starting a New Chapter for the Software Industry

Just came back from a venture capital conference hosted by Longworth Ventures in Boston, all in all it was a very good experience and having the opportunity to sit on a panel featuring fellow Irregulars Rod Boothby, Ismael Ghalimi, and Zolie Erdos was a highlight, especially considering that the panel was moderated by newly minted Irregular Andrew McAfee (aka “Probie”).

In my discussions with the Boston area venture capitalists I was left with the impression that this whole “2.0” thing is something that isn’t widely grasped outside of the Valley. At the risk of not making the best of friends with some VCs that I will actually have to sit on the other side of the table from and convince them to fund my new company, allow me to expand.

Of course I am making a generalization here because clearly there are VCs like Brad Feld and Fred Wilson that clearly get the impact of web 2.0 both on the consumer and enterprise sides of the industry and they aren’t located in Palo Alto, and to be fair I think there are quite a few VCs in the Valley that either don’t get it or don’t believe it’s meaningful to the degree that I do. However, what I am trying to suggest is that the volume of the conversation is a much more amplified where I live and work, which is reflected in the fact that all 4 panelists were from Palo Alto, and it would have been 5 if Ross Mayfield didn’t have to cancel at the last minute. Think about that for a minute, a panel discussion at a Boston VC conference and all the panelists are from the other side of the country.

I left SAP recently with a couple of observations that I am using to inform my decisions about where my new company goes:

1) Direct enterprise selling sucks, is highly inefficient, and makes you do unnatural things in your product strategy in order to drive higher deal sizes

2) Large enterprise software vendors are not the future. The whack-a-mole selling model where a handful of vendors (172 public companies in hardware and software) are fighting it out over approximately 50,000 business customers just can’t be the future! There are 38 million businesses in the U.S. alone that have less than 10 employees, there just has to be a way to grow our collective markets by appealing to these business users and I’m pretty confident in saying that it isn’t going to come from SAP, Oracle, or IBM.

3) The SOA-ification of big enterprise products has attacked a technical dimension, not an economic or business model one. In a somewhat bizarre turn of events, the historical strength of market leading business applications, the integrated suite approach, is being turned from an advantage into a liability. It’s like if you went into the Gap and wanted to buy a pair of jeans but because of the way the Gap was structured and how they merchandised, they would only sell you a pair of jeans, socks, underwear, shirt, jacket, and a hat… and when you told the salesperson that you just wanted a pair of jeans they said “but we are selling you a pair of jeans”.

4) Big enterprise software has historically been a product driven development process, not a user driven approach. Just look at the fact that all these companies have applications “products” and “modules” and “components” instead of something more logical, like “process portfolios” that map to what people actually do in their jobs. Furthermore, there is a big disconnect between what users do and what vendors build for the simple reason that no vendor except (more on that later) is selling to people who actually use their software, they are selling to people who buy and implement software for other people in the organization. Traditional enterprise vendors will tell you very convincingly that they are all about process, buuuuttt what do they actually sell you? Applications.

5) Lastly, and most importantly, there are no new big killer apps that are going to be built for today’s enterprise. Global business has spent the last 40 years automating every corporate function that is worth automating, and then they automated it again through “process reengineering” and once more when that didn’t work out quite like everyone thought. Today’s enterprise software market is about finding gaps and filling them, linking products in new ways, and leveraging more value out of IT investments that have already been made. I think the consumer side of the business is a fantastically rich opportunity landscape and if I were building a company today the company my team is building will straddle both worlds.

So getting back to the Boston area VCs, the reason I felt a disconnect with this group is that they were still thinking in the enterprise 1.0 mindset while fully acknowledging that most of the artifacts of the 1.0 version are antiquated. It takes a massive jump in mindset to believe that a company can exist by:

1) Building applications from other people’s stuff (OPS). There is a proliferation of web services being driven by all players in the market but they are largely looking at this as the mechanics for how to build and manage their own products. I look at the growing repository of web services and see a company that exist to package these services as components for new “meta-apps” or applications that users themselves are creating and publishing.

2) The “departmental sell” strategy is thrown around as essentially a downsized version of the traditional enterprise sell. In reality, the departmental sell is all about selling to people who use your software, actually use it in the sense of sitting down in front of it and really using it. This is one half of the substantial innovation that brought to the market, the other half being that they figured out how to do this with profitable unit economics.

2a) The closed loop process of seeing and hearing from actual users can inform a product development roadmap in ways that the traditional approach never could. As an added benefit, seeing in realtime what people are using, in terms of functionality, means that you can strip out the crap that didn’t work to your user’s satisfaction and wasn’t adopted. Wow.

3) Living on top of the idea that a software service can be built by and for it’s users featuring services that are built and supplied through other vendors, and sold through a channel(s) that doesn’t feature a bag carrying salesperson.

It’s an incredibly frightening thing to leave the safety of a very large company like SAP for a company that doesn’t qualify for a Cogenz beta account (Niall said he’s take care of us!), but I am comforted by the fundamental belief that the enterprise software industry is undergoing radical change that has little to do with technology alone. In other words, the company may not succeed but I know that the principles that we are building on put us in the very forefront of where this business is going and we will all be better off for it. My big concern at the moment, aside from delivering a product, is finding the right investor syndicate that not only believes in what we do but is capable of putting aside a lot of historical experience while we all learn how to do business at the beginning of a new chapter.

Technorati Tags: , , , , , , , , , ,

24 thoughts on Starting a New Chapter for the Software Industry

  1. Pingback Zoli's Blog
  2. Good post. If I may add one supporting comment. I believe it is incredibly important that the “web 2.0” application deliver measurable benefits to a closed loop process, whatever that process. If you have to wait and add a datastream or interface into another application, but that provides data that is vital to presenting this metric, then wait and put that in to your release. In another life I used to tell people “think about the dashboard or diagram you would want to see” and then build the numbers/application to deliver it.

  3. Jeff,

    Come’on theres many VC’s in Boston – and not all will show up at the Longworth event.

    Everything else I think is spot on, the model you refer to is much better at aligning incentives to improve the technology vs just selling another product. Good luck at the startups – they can be a wild ride and even if you don’ t make any money you will learn a thing or two. I look forward to hearing all about it.


  4. Karl,
    I tried to call out that I was in fact making a generalization, in retrospect I should have said that right up front. Having said that, i can’t think of a single “web 2.0” company (with some degree of brand) out of the Boston area, can you? Lot’s of infrastructure and traditional enterprise stuff… I am not alone in that observation:

    I don’t think it’s ignorance, wilful or otherwise, I just think that venture capital in Boston as a whole is in a funk.

    Thanks for the comment.

  5. Jeff – great post. I agree with what you are saying. The key thing in these so called assembled apps is going to be the (routinely recurring) problem of semantics of the interfaces/services. (This is why our PIC0 process too so long if you recall). Everyone solves the syntax issue and then “leaves it to the customer” to solve the semantic issue. So lets for a minute assume you are referring to soa-ized components from sap in bpp, then what you are going to need is to semantically map these to other components elsewhere from say Oracle. This is a very hard problem unless of course you pick all components from the same vendor. If you do the latter, then aren’t you simply building composite apps?The related issue to this is the implicit transactional behavior of the components that were written by different individuals, in different parts of the world.

    I believe the better you do in this mapping-of-meaning without adding too much of your own code / transformation logic, the simpler your overall solution will be, maintable/adjustable by business and not IT necessarily, and in that lies great value. So less would be more here.

    It is also interesting to think about the level of the mapping. If you map a phone number, great. Understanding delivery terms in a contract is an entirely different problem. Or loyalty of a customer based on more than just a loyalty score.

    Finally, it is important to think about the degree of automation that you will provide here.

    I will be watching you progress with great interest since I care about this topic! Good luck…

  6. Pingback Jeff Nolan and ‘Teqlo’ « TechnologyDriven
  7. Pingback Ouch, that hurts « HighContrast
  8. Pingback Hitchhiker’s Guide to 650 :: Lessons on Enterprise 2.0 From 1999 :: October :: 2006
  9. Pingback AccMan / Bigco software’s dog day over?
  10. Pingback AccMan / The company I keep
  11. Pingback Mission Creek / East Coast v West Coast VC
  12. Pingback Update on Trends and Information » Enterprise Software Landscape
  13. Pingback tecosystems
  14. So, the crux seems to revolve around building ready-made solutions using OPS and lowering COGS by not fielding a traditional sales force. It all sounds good. But getting people out of the “1.0 mindset” can’t happen until you *prove* that the approach – with or without 2.0 technology statements – avoids OPS pitfalls VC people can recite in their sleep.

    I work on commercial ERP apps and agree it still feels like a dinosaur age despite at least 3 iterations of “ERP apps are dead” declarations by analysts in the last 15 years. Investors are obviously more skeptical than analysts and won’t usually bite unless you have something other don’t and won’t have (patentable?) or you show you are a first mover and have a workable model. The latter is harder for enterprise apps because, as you point out, customers are already automated and saturated. But I think Investors will change mindsets surprisingly fast if you can demonstrate genuine resonance in a user base, especially those on the left edge of the long tail that you imply are already over-served.

  15. there is a major shift underway but we are stuck in the period of uncertainty at the moment, which is why enterprise software venture capital is much like a graveyard… very quiet and really unsettling.

  16. Pingback Venture Geek » Quickie: “2.0″ Really Has Meaning?
  17. Pingback Venture Chronicles
  18. Pingback Socialtext 2.0: Usability vs. Usefulness| Zoli’s Blog
  19. Pingback Lessons on Enterprise 2.0 From 1999 | Nosejob
  20. Pingback Lessons on Enterprise 2.0 From 1999 | Hitchhiker's Guide to 605

Comments are closed