Web as platform-don’t forget the enterprise

As I have written before, most of the talk about this next generation web has focused around consumer applications.  I, however, have always believed that we should not forget the enterprise.  This resurgence of web-based and loosely coupled applications has been driven by consumer-based innovation but there are many pockets of opportunities for the enterprise to take the best of the open web.  As you look at the adoption of technologies in the enterprise much of it has been driven by a push-pull mentality where a vendor tries to sell enterprises something they don’t necessarily need.  On the other hand, with the growth of the web and broadband over the last five years, it has made it easier for vendors to leverage the pull-push mentality where a single user begins using a service or downloads some code, hacks away on it, and then pulls it into the enterprise.   All of this make sense-consumers are web-savvy, broadband is everywhere making it an enjoyable experience, web-based services vs. client applications are driving growth in communications, sharing, storing, and collaboration—these same consumers also work at enterprises and "pull" some of their best practices and learning from the consumer world into their everyday working world. Let me give you an example.

A friend of mine heads up IT architecture at a large health care organization.  One of the big initiatives is to reorient the company to focus on the consumer (sounds like they hired too many consultants).  What that means for IT is how to do they integrate thousands of different databases to figure out all of the information about a particular doctor?  Sure, some of this is an exercise in enterprise data integration but you have to remember that probably 75% of the real information is in the form of unstructured notes about the particular doctor.  Think about how much data gets put into CRM systems which is not structured.  So naturally he asked me about what was happening in the consumer blogosphere, about tagging technologies, about RSS and turning every application into a publishing system, and how he could potentially integrate this into his enterprise.  The pain was large enough that he was looking for new and better ways, think loosely coupled ways to solve his problem.

This is just one example but you could easily think about that customer pain and extract it to a number of enterprises.  Data and application integration continues to rank either #1 or #2 in every CIO spending survey but going for the expensive $1mm plus point to point integration methodologies is not the way to go.  We just have to be creative and think about new ways to unleash the massive amounts of data in the enterprise to make the workers more productive.  Think of the easy-to-use technology used in search, RSS as the new publish subscribe, and loosely coupled applications as a new wave to hit the enterprise in the next few years.  Obviously all of the buzzword du jour technologies are just enabling technologies and it is incumbent upon the startup to find the problem, figure out the market potential, and understand how to sell it.  It is quite early in the process but this next-generation web will have a huge impact in the enterprise as well as in the consumer space.  It will just take some time  because while many of the startup companies I speak with understand the opportunity in the enterprise, they are rightly focusing on the market opportunity with consumers first.  Many of these entrepreneurs do not want to be seduced by the big dollar figure type deals that are out there knowing that it costs a lot of money to sell to the big boys and a whole different kind of support infrastructure.  In addition, most enterprises are not ready for it yet, but trust me, the early adopters are already out there trying to figure out how to use wikis, RSS, and other successful consumer technologies in their shops.  This means it is a good time to be looking so if you are an entrepreneur bringing some of these new technologies into the enterprise, let me know as I would like to speak with you to learn more.

Published by Ed Sim

founder boldstart ventures, over 20 years experience seeding and leading first rounds in enterprise startups, @boldstartvc, googlization of IT, SaaS 3.0, security, smart data; cherish family time + enjoy lacrosse + hockey

17 comments on “Web as platform-don’t forget the enterprise”

  1. Just to counter what you saying: in recent years I have found the enterprise less willing to bet on a small application or solution provider – from their point-of-view it carries too much risk. Personally, from what I’ve seen, the big change in enterprise software is the rise of open-source solutions combined with services companies. All the RSS/search/web 2.0 stuff is just technology – not a solution. I’m not saying there won’t be a hit or two – but it’s much harder for a startup in the enterprise space at this point.

  2. Your comments about incorporating consumer technologies are very interesting. Our company has already started the process of using Wikis to handle all of our system development documentation. All process flows, design docs, etc. are accessed through our Wiki providing us a cheap and easy way to track changes and to centralize the data.
    The problem we’ve found with incorporating these technologies into our business is finding the proper use, and making sure there is adequate security for the information.

  3. Damian-we are pretty much on the same page-first technologies don’t solve problems. Secondly, I view the bottom up movement of open source and SAAS as part of the “web as platform” mantra and new way of reaching enterprises. Here are a couple posts related to that:

    http://www.beyondvc.com/2005/04/welcome_greenpl.html

    http://www.beyondvc.com/2004/11/_product_innova.html#comments

    As far as enterprises buying from small companies, it is much more difficult but still doable with the right model for selling and delivering the service/product.

  4. Yeah – we’re definitely on the same page. I think Greenplum and Pentaho, along with SugerCRM and the like, of how open source is moving up the stack – still has a ways to go, but it’s getting there. ASP models, then, can be just an extension of this idea – so instead of making money from the services (or maybe in addition to), there is additional revenue from the hosting.

    I just get a bit tired of AJAX/wiki/tags/web 2.0 talk – and the business propositions for most of these companies are becoming more and more strained.

  5. Shhhh! The longer most entrepreneurs are drawn to the siren song of the consumer Web, the longer the rest of us can focus on the enterprise market, which is usually an order of magnitude bigger.

    You’re going to ruin it for us!

    All kidding aside, I can already see the beginnings of the push into the enterprise with the use of things like wikis for lightweight collaboration. I can’t wait to see what’s next.

  6. Ed, From the very beginning, the strategy of PubSub has been to focus first on the consumer space as an environment in which refine and prove our technology by providing a service and then later to pursue the enterprise technology (licensing) market.

    The intention is to ensure that we can reduce the technology risk to any potential enterprise customer and thus shorten sales cycles. As noted by LisaLiu2669 in the comments above, enterprises have learned to be very suspicious of small companies making big promises. (At PubSub, we make some *very* big promises!) This often means that the cost of selling to enterprises is increased drastically by demands for trials, benchmarks, and other expensive and time-consuming “proofs.”

    My theory is that if you can demonstrate an ability to satisfy millions of customers with high performance and low down-time then it should be possible to focus the discussion with an enterprise more on how the technology would be used in the enterprise and less on why the supplier should be trusted to have built a production-ready system.

    Of course, the transition from services on the web to technology licensing in enterprises is a very difficult one to pull off. For the web, you need no explicit sales force yet for the enterprise market, you typically need experienced, expensive sales folk with long Roldex’s and reputations with the enterprises they will be selling to. You also need an engineering support team that will typically be called on to do work drastically different from that called for by those running the service. In essence, pursuing the enterprise market is much like building a whole new company within the original — a challenging management task that calls for disciplined management.

    The alternative to starting on the web first is to approach the enterprise market directly and to hope to fall back to the online services market if the effort fails. However, I have seen no one succeed on that path. One reason for this is that enterprise software, is often optimized for management, reliability, etc. but simply can’t handle the Internet Scale performance requirements of the web. Another reason is that a company which is pursuing the enterprise space will have come to rely on an expectation of “big sales”, commissions for sales people, “elephant hunting”, etc. It is hard for such a once-enterprise-focused organization to get used to “squeezing water from a stone” — which is the Internet way — rather than the “big pay day” approach of enterprise sales.

    The enterprise market is certainly an important opportunity for many companies. However, folk must be very careful in chosing if and when to approach the market.

    bob wyman

  7. Ed, I agree that the enterprise is not to be overlooked. However, as an entrepreneur that has a platform that is equally applicable in both an enterprise and consumer environment, I agree with Bob and warn other entrepreneurs that are interested in the enterprise space to be careful. I have met with executives from prominent fortune 500 companies that are interested in our technology. Most companies had not even heard of Web 2.0, let alone some of the more popular technologies such as AJAX and rich clients. There is an education component involved. This is not cheap. Also, if you are dealing with large firms, you have to be able to fit into existing IT environments, budgets, and vendor qualifications unless you are a hosted service. On the flipside, if you are hosting a service, you have to ensure security, privacy, scalability, and reliability. This is particularly salient if you are hosting any sensitive business data, or data subject to regulatory constraints. Enterprises will not tolerate the incremental, hiccup-laden, services that consumers are willing to shoulder. In business time is money. The challenge for Web 2.0 entrepreneurs in the enterprise is to come up with a holistic solution that solves a specific business pain, not a set of technologies, or tools. IT has almost no pull these days. Budgets are tight and they no longer have room to experiment. Purchases are driven by measurable impact and have to be justified. All of this equates to long sales cycles for businesses. These types of long and process-intensive sales are not trivial for a start-up company to handle. In short, I would advise people to make sure that they are better funded than other Web 2.0 companies if they want to move in that direction. The rewards are high, but so is the level of investment required to get to a point where you can achieve real value.

  8. Bob and Hooman,

    I am glad I could get the conversation started. Great posts by both of you. I couldn’t agree with you more. We are still in the early phases of this revival and I just wanted to stir the pot to say that eventually this will move to the enterprise. And yes as I mention in my post, it is still early and requires a whole different infrastructure, expense plan, and go-to-market strategy. BTW, I hope any entrepreneur talking to an enterprise does not pitch Web 2.0 or other buzzwords because what they care about is solving their problem, not buzzwords.

    Ed

  9. Well, it is easy enough to list many ways in which the Web is important. And, there is a lot of evidence that the current importance was not easy to see five, 10, 15, or 20 years ago, or back L. Kleinrock’s work.

    Still, can’t just jump and say that the Web, RSS, Wiki, AJAX, etc. should be important for ‘enterprises’. Sure, it’s difficult to say they are not important, but for progress, at least for one step, need to have in mind ideas for the usual suspects: (A) What problem is being solved? (B) Who has the problem? (C) Why should they be willing to pay for the solution? (D) How many such target customers paying how much, when?

    The way my parents, professors, etc. taught me, finding the problem — where it hurts — is the easy part. Finding the solution is harder. Finding a pair — problem and good solution for it — that can make money is harder, still, but nearly essential to making money in a new business.

    So, for a good pair, should we start with the problem or the solution? E.g., should we start with Web, RSS, Wiki, AJAX and then say, “This should solve some big problems for enterprises, too!”. Well, my reaction is: There is a famous recipe for rabbit stew that starts out, “First catch a rabbit.” Then, my recipe for an application of technology would start out, “First find an application.”. Or, as people in research know, one of the most important steps to good new results is to pick a good problem. The solution remains the more difficult part, but a good problem is important, nearly essential.

    Actually, my view is that in nearly any large organization moving quickly and spending a lot of money, if can be a fly on the wall during meetings for a few days, then can find many cases of ‘pain’ (candidate rabbits). Get a long list, the longer and more detailed, the better. Sure, will only attempt to solve one of these problems, but for good problem selection, a good ‘rabbit’, likely need a LONG list. Then, climb down off the wall and head for a research library and look for what might be promising to help solve one of these problems. Sure, will likely need some software, but, naw, don’t expect to get a valuable pair where the solution is just routine software — want something more valuable than routine software, and can be sure that few software people will head for the research library and, if they do, will mostly just stay in the computer science collection which has nearly none of the best material in ‘information technology’! Likely won’t find the exact solution needed on the library shelves, but likely for at least one of the problems can find some “shoulders of giants” to stand on. How this could be so is curious, but it’s so! Then, maybe put feet up, pop open a cold can of diet soda, think about the list of problems, think about the contents of the shelves, and maybe have a new idea that does really well on one of the problems. Since we will likely be exploiting Moore’s law, the TCP/IP stack, RDBMS, etc., keep this in mind. Then, get a list of these pairs — problems and solutions. Do some first-cut investigations until the most promising pair seems clear. If this is a really promising pair — helps to have a good problem, some tall giants, and some bright new ideas — then proceed with prototype software, etc.

    But, in this, what is the promise of the Web, RSS, Wiki, AJAX? I’d say, “Beyond what is routine and well understood, not much.” Sure, I very much like the Web. Then, if something changes at a Web site, maybe I’d like to know about it. One way to know is to keep checking the Web site — in parts of computing this approach has been called ‘polling’ and is inefficient. The solution to the inefficiency of polling is ‘notification’ — RSS! Nice enough. Notification is fundamental — there should be many uses. Okay. But, that’s still not a rabbit.

    Some progress over RSS is the recent R. Ozzie work on SSE. Still, SSE can look like a solution looking for a problem, but Ozzie DID start with a problem and THEN outlined a solution and THEN, for some of the more important low level plumbing, did SSE. I don’t really like SSE — believe it is too close to some old ideas from Notes and, those, from ‘transaction processing’ where we have to be really really careful never to lose anything ever (the Web is different). SSE should have some value in the application (a kind of distributed calendar maintenance) Ozzie has in mind, and it should pop up elsewhere. SSE is not a rabbit; still need a RABBIT, a real live rabbit.

    Uh, so far, to use SSE for Ozzie’s calendars, need some ‘sharing with security’ — AGAIN! Would be too bad if the calendar application had to do its own! So, would want to ‘factor’ out the ‘sharing with security’ and make it a general purpose facility. Uh, ‘sharing with security’ very much does promise to be a common need in ‘networked’ and ‘distributed’ applications and work. So, get to write another draft standard? Or, maybe get a patent?

    So, looking for a rabbit, without being a fly on the wall in the meeting rooms for a few days, what pain might an enterprise have?

    Broadly the category that occurs first to me is advertising. Or, there is a remark that “The Web is the Internet hijacked by the advertising industry.”! Apparently so far, Google is mostly making their money just from advertising.

    Okay, try to imagine walking in the shoes of someone in an enterprise trying to do some advertising. Job with a lot of pain! Do some big fancy ad (spend a lot of time and money), run it on major TV channels (spend more time and much more money), try to get some decently accurate measurements of what ‘impact’ the ad had (lots of luck), remember that 50% of the ad dollars are wasted but don’t know which 50%, realize that likely under 1% of the viewers who watch the ad are in any sense candidate customers, and start to wonder how the world got into such a mess.

    So, realize that in sowing expensive seed, 99 44/100% of it falls on stony ground. Bummer. Instead, then, a candidate ‘problem’ is how to aim the seed at the fertile soil! So, want ‘targeted’ ads! How ‘targeted’? Sure, right down to the guy ready to buy.

    Then remember that the enterprise doesn’t really want ‘ads’ at all; instead, mostly what they want are just sales. Ads have just been the way to use newspapers, magazines, billboards, radio, TV, sky writing, and Google ad words to stimulate the sales. But, with Moore’s law, TCP/IP, broadband Internet, the Web, etc. — which are a long way from newspapers, etc. — are ads now the best way to stimulate sales or could there be other ways?

    Ads have a problem that is fundamental in nearly everything else in life — a lot of talking and no listening at all! So, one way to improve might be to get the targeted customer talking. Or, start to build a ‘dialog’, ‘conversation’, ‘relationship’ with the customer. Even if the enterprise end is plainly nearly all automated via software, still might be effective.

    So, if selling highly customizable products, then might want a customer dialog. Even if selling commodity products, might want to find out what uses the customer has and help them do better with those uses.

    There is a danger that nearly all of current advertising will go away. So, instead of the enterprise interrupting people — even well targeted customers — at unpredictable times and places and poking them in various ways to get them to respond, it may be that mostly customers, when they want to buy something, will go to Froogle, get a list of candidate vendors, shop at the vendors’ Web sites, and then make a selection.

    Then we are left with a situation: Customers might actually LIKE to receive some notifications, even at random times, for new products in certain categories. Okay, set up a company to ‘broker’ such ‘flows’ between the enterprises and the customers. The ‘broker’ would get all the customers because they had all the enterprises and would get all the enterprises because they had all the customers, would please the customers by giving them only well filtered high quality material and would please the enterprises by giving them a lot of information to help targeting. Would want some applied mathematics in there, mathematics able to match customers and ‘ads’ where the list of ‘characteristics’ of both are quite general. Hmm ….

    Interesting enough?

  10. One way web 2.0 can quickly enter the enterprise is as a tool for individual or small-team productivity, especially via mobile devices. My new blog, “Web 2.5”, covers the fusion of web 2.0 & mobile tech. See link below…

  11. Ed. I agree with your fundamental premise. The business world is beginning to adopt some of the innovations that have penetrated the consumer web in the last 2-3 years. You are also totally correct in identifying that users in enterprises are adopting innovative technologies under the radar of the IT department. This is being driven by the fact that these users are experiencing great innovation in their personal use of the web and want to bring those benefits to their workplace.

    I have spent the last three years growing a hosted software application for marketing. Nearly every client started small and grew their use of the application year-on-year. In nearly every case the marketing department adopted the [hosted] application without major involvement from IT. In every case funding was from operating budgets, not capital expenditure, so IT budgets were not used.

    There were a few comments about how expensive, time consuming and risky it can be to sell into the enterprise. This is true, particularly with old enterprise license software models. However, the advantage of the hosted/open-source models that genuinely offer greater usability and higher adoption are that new sales and pricing models can be used. Trial-to-buy, annual licensing, open user dialogue, shorter release cycles, etc all reduce purchasing risk for the enterprise (and consequently for the software company’s investors).

    In our case hiring the big sales guns was a step backwards and we saw a big fall in revenue for a period. Those old-style elephant hunters just didn’t understand that a smaller, faster sale with scope to grow was far better that a high-value, 12-month sales process with a high risk of failure.

    I also have a question. Why, when these conversations turn to the enterprise, does everyone talk about the Fortune 500? There is so much money to be made in the long tail of small and mid-sized businesses.

  12. Ed, I’m curious to how you would define ‘the enterprise.’ (its not a trick question, I truely would like to understand your definition of ‘enterprise.’)

    Is Enterprise only Fortune 500 / Global 2000? Or is Enterprise “everything that isn’t SMB (business with more than 250 employees)? It seems like we all have different definitions of what ‘enterprise’ means.

    Seems like most all of the Web 2.0 apps are geared towards consumers. Their ability to bridge their business models and products to the enterprise will probably never happen (with a few exceptions like Newsgator, maybe Zimbra, and perhaps Socialtext (but only as a professional services company)).

  13. Yes, everyone has different definitions. As my post was meant to stir the pot, I wanted to just juxtapose the biggest enterprises against the consumer. The reality is that alot of money from these web platforms will be made in small and medium enterprises as the sales and marketing to reach them are more efficient and cost effective, especially today versus yesterday. Yes, there will be some larger Fortune 1000 department-level groups using some of these services to solve problems but this will be the exception more than the rule. I am not advocating elephant hunting as you know from all of my prior posts that Frictionless Sales is a big deal to me.

    http://www.beyondvc.com/2005/12/frictionless_sa.html
    http://www.beyondvc.com/2005/09/skype_and_siebe.html

    While everyone is thinking consumer it is not a bad idea to think about the business world

  14. Ed

    You wrote:

    > it costs a lot of money to sell to the big boys and a whole different kind of support infrastructure.

    Did anybody commenting here read this?

    It is such a fundamentally different proposition that most would-be entrepreneurs wouldn’t know where to start.

    Of course, that doesn’t mean that they shouldn’t figure out where to start and start up the learning-curve …

  15. Pingback: Moonwatcher
  16. You shouldn’t attack all consulting firms. In reality, many of the larger firms actually do provide any real value other than helping an enterprise realize their own potential…

Comments are closed.