I'm also planning to forward my domain www.davidprovost.com and consolidate my online presence. We'll see how this goes...
Wednesday, December 3, 2008
If you're a regular reader of Semantic Business you'll recognize that things have been changing - I've added a jobs feed, some of the formatting is slightly off (to my chagrin), and I'm planning to start serving ads in a relatively unobtrusive way. All of this is a precursor to my move over to TypePad (preview here), which looks much better suited to my needs going forward. Please bear with me during this time - this isn't a random decision and I'm aware that links are likely to be broken, things may be lost (hopefully just temporarily), and I may spill my coffee. Or not.
Posted by David Provost at 12/03/2008 04:34:00 PM
The excitement of Android's launch has died down a bit so this seems like a good time to cast Google's entry against Nokia. Sounds like an unfair contest right? After all, one is a giant in its industry, a leader in R&D and technical innovation, with a long history of support for the developer/open source community, not to mention a globally recognized brand. I'm referring to Nokia, in case you were wondering. I'm not ignoring Apple and the iPhone, but I don't get the impression they're pressing as hard to shape a play that stretches from fundamental infrastructure up to the end user experience - which I do believe has crossed the minds of people at Google & Nokia.
Both companies have money, talent, recognition, and deep commercial relationships. Nokia probably has the lead in governmental relationships due to the highly regulated nature of the telecom industry. Google's got search so solidly nailed it's scary. With Android, Google is moving onto Nokia's turf. In the meantime, Nokia's plans appear to call for an increasing emphasis on Web services and certainly mobile content, as demonstrated by its launch of Ovi and its own music store (maybe Nokia's taking a shot at Apple, after all).
I won't dwell on the increasing sophistication of mobile devices or that they'll become a prominent means of Internet access (or even primary for some people), instead, I'll focus on the evolution we're seeing on the WWW toward the SW. Since the SW is simply an extension of the WWW, it's safe to assume that wherever the Web can be reached, the Semantic Web can be reached as well - it's all due to HTTP after all.
Here's where the respective paths of these companies begin to diverge. Google's demonstrated its schizophrenic approach to the SW pretty clearly, as I explored in this post. It seems they're quietly exploring the fringes of the SW, but they're probably expending just as much effort in denying the existence of these activities. I haven't looked through any of Android's documentation for SW references, but since I haven't found any elsewhere, it's possible that there simply aren't any.
On the other hand, Nokia's involvement in the Semantic Web goes back to at least to 1997, when Ora Lassila wrote a brief note titled "Introduction to RDF Metadata," likely when he was a visiting fellow at W3C. Ora's gone on to play an influential role within Nokia, promoting the SW all along the way. Others within the company have followed suit so that now there's a raft of internal SW projects here and here (they're not all SW projects, but a good number are). So, given the complexity of mobile environments, the differences in platforms, operating systems, regulatory environments, etc., it seems like SW technology might be well suited to service composition and delivery, provisioning issues generally, and offloading processing requirements from the handset.
Oh, did I forget to mention that Nokia acquired Navteq over the summer? Frankly, I don't think there's enough of a difference between Google Maps, Navteq, Yahoo! Maps, or others to really make a difference. But the point is that by purchasing Navteq, Nokia no longer needs to rely on Google and it no longer needs to divulge any information regarding how location based services might be developed and deployed. That's knowledge Google will have to acquire on its own. Since I've already indicated in my earlier post about Google that there's no evidence they're recruiting people with expertise in RDF, OWL, ontologies, SPARQL, graphs, triples, etc., it doesn't look to me like they're going to catch up any time soon. As a matter of fact, now that Google's cutting its workforce , I'm guessing the company's appetite for innovation may slow down, meaning that if they want to catch up in any contest with Nokia, they're going to be sucking serious wind.
Diversifying away from a reliance on search makes all the sense in the world to me. But can a mobile operating system succeed on its own without its very own fleet of handsets, infrastructure, content beyond maps, software development in what I believe will be a key technology, and deep experience in global telecom regulation? I don't know, but it's not a bet I would take.
Posted by David Provost at 12/03/2008 03:10:00 PM
Thursday, November 20, 2008
Since publishing On The Cusp at the end of September, a number of companies have contacted me to schedule briefings and quite frankly, I'm very happy to do so. Of course, now I feel self-conscious because I have a bit of a backlog to attend to...
With that said, I'm planning to provide updates on the companies I've profiled while also profiling new companies (or at least, new to me) in the Semantic Web industry. Additionally, as I become aware of new information or discover something interesting, I'll provide updates accordingly. Along these lines, anyone reading this is welcome to contact me about companies they'd like to see profiled, even if it's their own(!)
In the meantime, you'll be seeing new profiles and updates starting in the very near future.
Posted by David Provost at 11/20/2008 06:18:00 PM
Inform Technologies, Inc.
HQ: New York NY, USA
Products (Primary): Inform
Contact: Josh Kirschner
Vendor Category: NLP
Installed base: --
It’s safe to say that Inform has built a successful business by using NLP technology to process its customer’s Web sites and then enrich these pages by linking to relevant content published elsewhere on the Web. One look at Inform’s customer list, which includes names like Ziff Davis, The Economist, Wired, and many more serves as solid evidence that the company has a winning product. Each of these customers has decided that it’s in their best interest to seek out relevant external content and then publish it to supplement the original article on the page.
Increasingly, publishers seem motivated to engage in this practice in the belief that it reinforces their authoritative standing in the eyes of their audience. As would be expected, customers can control what external content is deemed suitable for publication on their site by creating white lists and black lists, as well as whether or not to keep visitors within the publisher’s “family” of media properties.
In practice, using Inform’s solution is easy enough – Inform hosts the back end where the processing is performed, and when a publisher creates an article its submitted through Inform’s API. The article is processed, the (desired) relevant content is identified and linked, and the results are returned to the customer for final publication. Aside from expanding the content presented to visitors, these results also play an important role in Search Engine Optimization (SEO), with customers reporting page views increasing by 10% to 20% and in some cases as high as 25%.
In the course of its existence, Inform created a well developed, professionally maintained taxonomy. This complements the work of the company’s library scientists, linguists, and ontologists and has had the effect of positioning the company well to pursue specific vertical markets such as health. As a result, Inform is prepared to move into select industries and may well do so from a position of strength, unburdened by the need to play “catch up”.
A very interesting (and fully operational) example of just how far Inform’s solution can be extended is found at NewsDaily.com. Reportedly, this site is operated by a single individual who uses a Reuters feed subscription to form the content kernel for processing by Inform. Clicking through the top-level Reuters content leads to pages the clearly include related articles from a wide range of publishers. Readership or audience numbers weren’t supplied, but the low fixed costs of this business suggest that modest advertising success could yield solid revenues for a one man operation.
Six/Twelve Month Plans:
Aside from its potential entry into specific verticals, Inform is considering opportunities in advertising. While it’s easy to imagine the extraction of primary concepts from an article and then associating relevant advertisements, the actual implementation has challenges which Inform has yet to fully define. Aside from these two possibilities, Inform is deep into execution mode and at this point, a primary goal for the company is to continue building on its track record of success.
Inform has a roster of believers who have contracted for their services. In many ways, the business case is fairly easy to express – in a traditional publishing environment there might be a number of editors who spend part of their time tagging stories for a variety of reasons. Time spent tagging means time taken away from other, higher value activities. Inform’s solution reduces this burden on content editors which translates into time savings, cost savings, and higher productivity (arguably, these are all ways to describe the same business result). These savings, combined with an improved audience experience, create a compelling argument for publishers to take a close look at this technology and how it fits with their overall goals.
Posted by David Provost at 11/20/2008 06:15:00 PM
Thursday, November 6, 2008
I don't know anyone that can conclusively say what Google's up to with respect to the Semantic Web, so I'm embarking on a small mission to figure this out - or at least shed a little more light. There are more reasons than I can think of for taking a look at this issue, but for starters:
- Android vs. Nokia's raft of SW efforts related to mobile environments, and Nokia's stated strategy of relying on three revenue streams deriving from handsets, Web services, and mobile content.
- Google has a lot of smart people, many of whom are hired out of MIT (probably from the same building where the W3C is headquartered).
- Unconfirmed reports that upon visiting W3C in Cambridge MA one or two years ago, Eric Schmidt commented that there was lots of "good stuff" going on there (but no research contracts were forthcoming).
- I just searched on Google's US hiring page and got zero (0) results after using the following nine terms (one at a time with no operators): rdf, owl, sparql, ontology, semantic, uri (although url only got one result), linked, triple, graph.
- Google's a member of the W3C and since May 1, 2008 nineteen (19) individuals with "@google.com" in their email address have posted on the W3C's public mailing lists (the lists that the working groups use). Some are quite prolific, particularly if they're chairing a working group. Lots of focus on geolocation.
My take is that the search terms have either been scrubbed from the posting results, scrubbed from the job descriptions, or Google just isn't hiring anyone with competencies in the nine areas searched on above. I find this last point unlikely for any forward looking company whose reason for being is the Web itself.
I'll persist in my research and report back - I'll also fill in some links to the points above as well.
Posted by David Provost at 11/06/2008 01:12:00 PM
Friday, October 24, 2008
I sat in on the VC panel at Web 3.0 and took note when they discussed what they look for in the deals they fund. Not a lot of surprises, but I think their points are worth repeating:
- What problem do you solve? I look at this as a 25-words-or-less plain english statement. I've spoken to a tremendous number of entrepreneurs who struggle to do this and it's not easy. I've created these statements in companies where I've worked and it takes time, thought, and a lot of testing to figure exactly what message gets your point across in a way that people can understand. Obviously, the other benefit is that the more precisely you define the problem, the easier it'll be to solve - hopefully!
- What's it cost to get a customer? In the late 1990's, I spent a lot of time working with companies like Expedia, E*TRADE, Fidelity, and other household names in e-commerce. Regardless of the industry, each successful e-commerce site came to the realization that their customer acquisition cost was a huge issue and one that needed to be monitored constantly. This point applies equally well to consumer facing sites as well as corporate oriented solution vendors.
- What do you need to believe in? In other words, if you're starting a social networking site, you need to believe that 1) people are essentially social animals that will usually look for ways to connect with each other; 2) you've identified a niche and created an experience that will actually attract and retain enough of an audience to make the venture pay off; 3) the sun will rise tomorrow morning, or whatever else is essential to success. Being able to state these beliefs helps set context and provide perspective, both of which lend themselves to understanding by a third party, like a VC.
- What are your dependencies? This is distinct from your beliefs, above - these are the things that have to happen to make your idea work. If you're pursuing a solar power idea, some examples might be 1) silicon based solar panels will continue to fall in price and rise in efficiency; 2) there's a finite supply of oil and over the long term its price will continue to rise; 3) because of construction lead times and political issues, nuclear power won't be a competitor any time soon. And so on - in any case, the point is to identify and clearly understand the things that have to exist for your business to succeed.
- What's the competitive environment? Well, this one's pretty obvious, assuming you know what business you're really in - see the first point above. Be sure to include substitutes and not just other competitors that happen to look just like you.
- What's the quality of the team? Putting together a business with your best friends from high school may not be such a great idea, unless they happen to be uncommonly well suited to the tasks they've been assigned. Otherwise, you're far better off reaching out to the best people you can find, which means talking to a lot of people you don't know, getting a glimpse of lot of things you don't know (through their eyes), and hopefully, finding people who are so smart and so talented they leave you thinking you'll have to work hard just to keep up with them. Inevitably, investors need to be convinced that the team in front of them can take their idea and execute on it better than anyone else around.
Whether you're experienced and have dealt with all of these issues or you're new and facing them for the first time, these points are worth bearing in mind as reminders or guides. Now go get 'em!
Posted by David Provost at 10/24/2008 08:50:00 AM
Monday, October 20, 2008
This was a small (~200 attendees) but successful conference. Dan Grigorovici invited me and I'm glad he did - I'd actually geared the release date of my report to be in advance of this conference and I was very nicely surprised at how many people said they found it helpful. Some quick thoughts:
- I was surprised at how many speakers focused on online advertising or Natural Language Processing (NLP). For awhile I began to think I'd missed something in my registration materials, but I got over any misgivings in time for my panel discussion. Fortunately I had the chance to point out the in the US we have a $13 trillion economy and an audience member volunteered that online advertising accounts for roughly $85 billion. In the remaining $12.9 trillion I'm certain we're leaving many opportunities for Semantic Technology untouched.
- When I gave a brief talk in Cambridge, MA (MA) two nights earlier, I'd been asked if any corporations were exchanging URIs in the course of conducting business. I'm not aware of any large (Global 1,000) companies engaging in this practice although I can see plenty of reasons to do so. During my panel I ran through a very simple example of Kimberly-Clark (KMB) exchanging URIs with Weyerhaeuser (WY) for production, supply chain, and demand management.
- In a later session, a collection of West Coast VCs provided a very interesting panel discussion on what they look for in the companies they select for investment - more on this in a separate post.
All in all, Web 3.0 was time and money well spent.
Posted by David Provost at 10/20/2008 06:08:00 PM
Thursday, October 9, 2008
Response to my report has been surprisingly strong and my thanks to everyone that's written in. Several companies have contacted me to provide a briefing and I plan to write up those conversations and post them here on my blog. I'm not sure of my own next steps but I plan to offer consulting services in the interim - stay tuned and I'll post details shortly.
And special thanks to Paul Miller for having covered my report - he also kindly invited me to do a podcast which you'll find here.
Next week I'm off to the Web 3.0 conference in Santa Clara where I hope to keep learning more about how companies are using Semantic technologies to achieve their strategic goals. If you plan to attend, I hope you'll introduce yourself and say hello.
Posted by David Provost at 10/09/2008 02:37:00 PM
Tuesday, September 30, 2008
I'm delighted to announce the availability of my report titled "On The Cusp: A Global Review of the Semantic Web Industry". In it are profiles of 17 industry participants and analysis that highlights the following:
- The Semantic Web is a commercially competitive technology.
- Linked data will be extremely valuable - when it's better understood.
- Natural Language Processing is an important step in deriving value from "World Wild Content".
To download a copy of this report, click Download Semantic_Web_Industry_Review.pdf (1416.1K)">here.
I'll be part of a panel discussion at Jupitermedia's Web 3.0 conference on Thu Oct 16 and I'll hope to see you there!
In the meantime, my thanks to everyone I spoke with at the companies covered in this report - their participation and involvement made this report possible. Lastly, and far from the least, my thanks to Jim Hendler. For the past several years Jim's been a close advisor and a staunch supporter who's never let me down in a pinch.
Posted by David Provost at 9/30/2008 10:02:00 AM
Sunday, September 14, 2008
I'll be on a panel titled Product Marketing, Key Biz Strategies and the findings in my report will be a key part of my comments. If you're planning to be there I hope you'll say hello!
Posted by David Provost at 9/14/2008 10:32:00 AM
Tuesday, September 2, 2008
Fri Sep 5 marked the end of the period I budgeted for interviews and data collection and frankly, I'm delighted with the number of companies that have participated and the thoroughness of their responses. The companies that will be covered in my Sep 30 report are:
In the final report, each of these ventures will be profiled individually to describe their primary offering, what it does, near term plans, and commentary. Several patterns have emerged and I'll discuss these in an industry analysis section that will also be in the report. In the meantime, I'm still on schedule to publish on Sep 30. I'll announce the availability of this [free] report here on my blog and I'll also be sure to provide a link to the .pdf.
My sincere thanks go out to the people and companies that took the time to participate in the interviews and then later, for reviewing the profiles for accuracy - their help and support has been central to this process.
Posted by David Provost at 9/02/2008 09:29:00 PM
Tuesday, August 19, 2008
It's time for another progress report - I haven't posted any recent entries to this blog because I've been busy with this project. Even so, since my review of the Semantic Web industry will include a description of my method, I figured I'd get a head start by outlining my approach below. A version of this will appear in the report:
- Picking the companies is never a perfect process, especially for an initial review. Factors that contributed to being covered by this report include being primarily product based as opposed to consulting services based, some kind of sponsorship at a conference (it's an indicator of success & maturity), and being in "general release" and not beta (a fuzzily enforced criterion). At this point I estimate that 18 companies will be covered.
- Assemble a list of relevant questions designed to uncover basic business issues, product capabilities, trends, and future plans. These questions were reviewed by a small subset of vendors to ensure fairness, relevance and adequate coverage of key issues.
- Interviews with key personnel. This is really the heart of the process - the discussion, Q&A, and general exchange involved in covering my questions is what provides the illumination that I'm seeking.
- Writing company profiles based on the discussions and publicly available information. These are all going to be between one to two pages long and they'll contain a description of the company's primary product, what that product does, plans for the next six to twelve months and some light analysis. There'll be other basic information as well.
- Submit the profiles to the companies for review. In my view, this is a key step to ensure accuracy. Note that the analysis contained in the body of the report will not be made available for review in this fashion.
- Summary and analysis, which is where I review the company profiles, develop my findings, and state my observations and comments.
I'm very much in the thick of things right now, but there's always time for trivia:
- 11 interviews complete, 7 profiles written.
- The companies span time zones ranging from GMT -7 to GMT +9.
- 11 are headquartered in North America, 7 are headquartered in Europe, 1 is in Korea.
- August is global vacation month - no kidding. Why do we even bother with Q3?
Posted by David Provost at 8/19/2008 08:11:00 AM
Wednesday, August 6, 2008
If you're reading this, you may know that I'm engaged in a Semantic Web industry review based on interviews with the leading industry entrants. Once this phase is complete, I'll reflect upon and analyze my interactions and write up the results. My target for publishing this (free) report is the end of September. Here's where I stand:
Aduna - interview being scheduled
Cambridge Semantics - interview complete
Franz - interview being scheduled
Garlik - interview being scheduled
Intellidimension - awaiting initial response
Mondeca - interview complete
Ontoprise - interview being scheduled
Ontos - interview complete
Ontotext Lab - interview being scheduled
OpenLink Software - interview complete
Primal Fusion - special case: pre-launch
Saltlux - interview being scheduled
Sandpiper Software - interview being scheduled
Siderean Software - interview being scheduled
Sindice - special case: pre-launch
Talis - interview being scheduled
Thetus - interview scheduled
TopQuadrant - interview being scheduled
Dow Jones Client Solutions - interview scheduled
ITA Software - special case: pre-launch
The Calais Initiative - interview complete
Twine/Radar Networks - interview scheduled
Yahoo!/SearchMonkey - awaiting response
Posted by David Provost at 8/06/2008 05:54:00 PM
Saturday, August 2, 2008
The World Wide Web is the virtualization layer. All digital assets (databases, files, executables, down to the record level) can connect in a consistent, universal, two-way fashion.
The Semantic Web is the conceptual layer. The virtualized assets can be combined and linked. Those links can be used by people, machines, other (imagine emailing someone a link to a Web page). It's an important concept, because you as a person are more than what's contained in a credit card database, blog, school transcript, etc. With the SW you can link all those assets together and create a more complete picture of yourself.
In an enterprise, this could be a more complete picture of a customer, chemical compound, financial market, etc.
See the ugly picture with non-standard symbols & terminology (I'll replace it with something nicer - full size here):
Posted by David Provost at 8/02/2008 11:06:00 AM
Friday, August 1, 2008
I've been in radio silence temporarily as I ramp up the organizational process and start speaking with companies. Response to my invitations has been swift, strong, and enthusiastic. Nearly all the companies I've contact have agreed to participate in my survey and one or two more have written expressing interest in being included. To be fair, the companies that have chosen not to participate are still in pre-release/stealth mode, but I wanted to make sure they at least had the choice to join in. At this point, I've either scheduled conference calls with each company or doing so is in process, so there's been no hesitation on anyone's part to get started.
The two vendors I've already interviewed have been around for several years and their comments reflect their experience. What struck me the most was their evolution has been based on real-world experience that occurred before the Semantic Web was formalized by the W3C. Each founding team came with hands-on exposure to significant IT infrastructure problems and the recognition that Semantic Web technology is very well suited to solving these problems.
What's also come out of these conversations confirms that the differences will be in the details. Bear in mind I've only interviewed two companies so far and they don't compete with each other. But each one has taken subtle but substantial steps to differentiate the results their customers receive and likewise, the value of these results. Sorry, you'll have to wait for the report for me to clarify that rather vague statement.
More updates to come.
Posted by David Provost at 8/01/2008 09:40:00 AM
Wednesday, July 23, 2008
I've sent the invitations and I've been surprised at the speed and extent of the acceptances. Cutting straight to the list of vendors I've invited to participate in my industry survey:
And the "deployers" (trying to think of a better term...suggestions are welcome):
Dow Jones Client Solutions
The Calais Initiative
I've created lists for industry reviews before and I've always found that having a few basic guidelines or criteria for inclusion can be very helpful. Here are the ones I've used:
- The entrant has to be a company or part of a company.
- Professional services firms (consultants) are not included. I want to focus on vendors or makers of products, Web presences, or Web services.
- Defense contractors like Northrup Grumman, Boeing, etc. are not included. Those markets are too specialized and restrictions related to classified information arise far too early and often to allow real exposition of the issues.
I've tried to keep this simple and I don't guarantee perfection. As a matter of fact, I've always found that it takes several iterations to get the criteria, categories, and key issues right. This is a starting point where I hope we can all learn together and build on successfully.
Gotta go - I've got to get back to a bunch of people to schedule conference calls. I still owe a couple of entries on this blog related to the process of building shareholder value and separately, using a simple investment portfolio model as an analogy for corporate IT portfolios.
Posted by David Provost at 7/23/2008 12:19:00 PM
Saturday, July 19, 2008
I was chatting with a friend the other day whose business is really starting to take off. It's a great problem to have. As would be expected, juggling priorities and responding to the host of demands on his team's time, attention, and energy is becoming a pretty significant challenge. It's one thing for us to set priorities for our own activities, but as you grow, how do you create a consistent framework that the entire team can use as a guide without having to re-visit each and every new situation? Here are some thoughts for a lightweight, easy-to-use guide:
Prioritizing and Qualifying Business Relationships
There are two paramount factors when setting business priorities:
- Money, in the form of profit or shareholder value.
- Time, because a dollar today is worth more than a dollar tomorrow.
Every business faces demands on its time, personnel, infrastructure, and effort, which I’ll refer to as resources. These resources can be quantified in terms of money. If two businesses face similar demands and provide comparable levels of service, the one that targets the highest value activities and executes on them with the greatest efficiency will win.
In rank order, business relationships follow this pattern (and each category, or set, can have its members prioritized within the set’s boundaries):
Customers – parties that are prepared to pay money (or in rare cases, barter) for goods and services. The money provided by customers is the lifeblood of a business, and the more customers and money, the better.
Fundamental qualifying questions:
- Can the customer convince the vendor that it has the money required to make a purchase? Customers seeking to make the largest purchases take priority, assuming equal decision timeframes. Vendors must ask: Do you have a budget? What size is your budget? Has it been approved? Once a decision has been made, what’s required to release the funds?
- Is the customer presently in the market as a buyer and how long will it take for them to make a buying decision? If two customers will spend the same amount of money, the one that will make a decision first takes priority. Vendors must ask: What’s your decision timeframe? Will there be any additional approvals required and if so, how long will that take?
Partners – parties that provide access to customers, or by providing quantifiable value, allow a vendor to pursue customers it wouldn’t otherwise be able to attract.
Fundamental qualifying questions:
- If a partnership is formed, what do the parties get that they wouldn’t otherwise?
- What will ensure that each party gets whatever it is that’s been specified?
- How much time should be allowed for the specified results to be achieved?
- What happens if expectations aren’t met? Will one party be liable to the other on a cash basis?
Employees – people who are paid to enhance the value of a business through their labor.
Fundamental qualifying questions:
- What clearly identifiable skills does the person bring to the business?
- Can the value of these skills be calculated? Are there specific customers or markets that will make buying decisions as a result of what this hire brings to the team? Similarly, will this hire free other personnel to catalyze buying decisions?
- What specific knowledge, relationships, or credibility will this person bring to the business? Can these be quantified in monetary terms?
- Can a durable, highly motivated core or team be built around this person?
Analysts/Media/Consultants – people and organizations in a position to directly influence the behavior of customers.
Fundamental qualifying questions:
- What relevant customers do these people/organizations serve?
- Is their influence needed?
- Aside from consuming the company’s resources, is there a cost for establishing and maintaining a relationship?
- Is the analyst/consultant or media property credible? How credible is it?
Knowledge seekers – people contacting the business to increase their knowledge without any exchange of value.
Fundamental qualifying questions:
- Can this person’s questions be answered through existing and readily available material?
- How much time would be required (or should be allotted) to speak with this person and answer their questions?
- Is it conceivable that this person might add value or influence customer behavior at a later date?
This framework can serve as a consistent guide, but the exercise of business judgment will always be a necessary ingredient. With experience, different qualifying questions may emerge to suit a specific industry, relationship, or opportunity. Adapting to these changing circumstances will be key.
Posted by David Provost at 7/19/2008 11:01:00 AM
Wednesday, July 16, 2008
I’ve been involved with the business of the Semantic Web (SW) since 2004 and I think this is an excellent time to launch a review of the “industry players.” The SW industry now has a number of participants and there’s an opportunity to clarify who they are, what they do, and why anyone, particular buyers, should care.
Over the next few weeks I’ll be contacting companies like (and no, this is not a complete list):
- Cambridge Semantics
- ITA Software
- Open Calais/Reuters
- Twine/Radar Networks
My conversations will follow a consistent format and I’ll have at least two classes of players – the first will be conventional vendors like Siderean, Thetus, and TopQuadrant and the second will be “deployers” of the technology like Open Calais/Reuters and Yahoo!. As I go along I’ll post observations on my blog, Semantic Business, and when I’m done I’ll write up the findings in a final report which I’ll make available for free.
It’ll be free because:
- There’s a growing trend among analyst firms (at least the startups) to give their reports away for free. Where they make their money is in retained relationships, corporate sponsored presentations, and custom client projects.
- I want to maximize industry (corporate, vendor, research & scientific) engagement, conversations, and involvement, so I want to minimize the barriers to interaction. Sounds a lot like open source, doesn’t it?
- I’m willing to take the risk that this approach will work. It might not.
The level of activity in the industry is higher now than it’s ever been. That means it’ll be hard for me to find all the vendors, deployers, projects, etc. If you think I’ve missed something, please write to me and let me know.
I’m looking forward to this and I hope you are too!
Posted by David Provost at 7/16/2008 11:24:00 AM
Tuesday, July 15, 2008
Things were moving along pretty well - I knew the problem I wanted to solve, I'd confirmed customer interest in my approach, the open source technology already existed and I had the the support of the community leaders, and I had some pretty clear ideas about money. Now to start building a team, starting with the right partner...
If you've read the previous entries, you can imagine that I was starting to get pretty jazzed about my prospects. But I'm not a technical guy and in 2004, there were literally only a handful of people available to fill the role of CTO. I felt I needed someone with unquestionable technical skills, a significant profile within the SW community, and the ability to articulate the value Human Element was trying to deliver in clear, understandable terms that potential customers and investors would find appealing. I'd been in the mix long enough to know exactly who I needed to reach out to. (All names changed to protect the innocent, the guilty, the shrubbery in my back yard, and whatever else.)
- Candidate 1: I'd interviewed Bill in the course of researching my thesis. He'd been really helpful and at the time, his own venture seemed poised to do well. In the interim, it actually folded, which made me think he might be interested, even though he was on the West coast. The reality: I've never been able to track the guy down. I don't know anyone that's heard from him since, and I've never seen any record (at all) of his activities, even to this date. It was easy to move on from this one, but it still seems strange. Maybe he works for the government in a secret bunker a half-mile underground...
- Candidate 2: James is one of those other-worldly smart Ph.D. types, but I knew who he was and his qualifications were unquestionable. I pitched him on the idea and he was lukewarm at best. I persisted, and he asked, "...what's the end game?" Recognizing that cutting to the essence of the matter was crucial, I answered simply "Money." I'm talking about a business after all, and that's what businesses are supposed to do - make money. James responded "figures, coming from a business guy..." I tried pointing out that in a business, people count on people like me who think about money because they need to get paid, pay their rent, kids' tuition, eat, etc. I've been in touch with him since then and I'm convinced that one day he's going to turn the world on its head, but I guess not with me. Moving on...
- Candidate 3: Harry had been in an SW venture before and actually had some business experience. He'd left his company and had some time on his hands, so we talked about the idea and he seemed interested. I continued to ask and he continued to not say no, round and round we went until one day he basically said "Listen, I'm going to finish my Ph.D. and I really want to focus on that." OK, I understand that, particularly since I just graduated from a mid-career masters program myself. Time and possibilities were running out at this point...
- Candidate 4: There were a few other candidates that I haven't mentioned, partly because I ruled them out quickly or because they never responded. Carl was different. He'd been advising me for some time and he knew what I was trying to do. I also think he was genuinely hoping I'd succeed. Naturally, I pitched him. Carl had never been in a business venture before. He'd always been in academia, but he made it clear that he was losing enthusiasm for that lifestyle. As my sales conversations were moving along, he assured me he'd deliver a demo. I persisted in asking him to join me and he reported back that what I was saying synced up well with what his lawyer was telling him. This conversation took place over six to eight (valuable) weeks...radio silence ensued.
During this period when the reality of not finding a partner was sinking in, I was on the phone with a close advisor one day. I was telling him that I was uneasy about my search for a technical partner and we were talking about how this was a make-or-break issue. Our conversation moved along until I mentioned that I had to leave - I was on my way to a meeting with Polaris. He knew the person I was meeting with and advised me based on his experience. Then he asked "How are you going to deal with the technical questions and what are you going to say about your search for a partner?"
Instantly, my stomach sank and my heart was in my throat. I don't remember what I said but I ended the call shortly thereafter and left for my appointment. By the time I arrived at Polaris I knew what I was going to say and frankly, the meeting went well. I'd set expectations for an informal "this is what I'm working on what do you think" kind of visit and it ended up being very productive. I've been back since then, but only to discuss a different venture.
By this time I was out of candidates and I'd exceeded my timeframe for attracting a technical partner. There weren't any other candidates and financially, I couldn't afford to pursue this venture any longer. I'd met my roadblock.
Milestone #5, attracting a qualified technical partner within the required time period, failed. This was the most important task I had to complete and it was a deal killer.
The following week pretty much sucked. I'd extended my deadlines because I wanted to believe I'd landed a technical partner. Now it was time to kill my idea and move on. Fortunately, I bounced back quickly and I have to admit, my experience as an athlete helped - being conditioned to move on from a single defeat, a loss of a point, a rally, a heat, whatever, made a big difference.
I learned a lot from each one of the tasks I set for myself and obviously, I lived to tell my tale and fight again another day. I'm still looking for the right venture and the right team, and if you've read this far who knows, maybe we'll find ourselves on the same team some day.
Posted by David Provost at 7/15/2008 07:26:00 PM
Sunday, July 13, 2008
In Part One, I talked about identifying the problem I wanted to solve, expressing it in plain English (or plain language, for all non-English speakers), using open source as a head start, and getting the support of the open source community's leaders. Things were percolating nicely, but how was I paying the bills and who was going to join me? Read on...
Money is a many-faceted topic, so I'll take some care in how I cover this one.
- Paying the bills. There were two things in my favor - I had just finished up a well-paying contract with Digitas and I had some money in the bank. My wife also has a thriving recruiting practice, so combined, I had some reasonable (but not unlimited) runway to pursue this venture.
- Funding the serious development that would eventually be required. I didn't want investors, I wanted customers. Essentially, I wanted my customers to fund my idea through their willingness to purchase the product. When bringing early stage technologies to market, there are customers that recognize there's a certain give and take in the sales and product development process and that's the kind of customer(s) I was looking for. My attitude was "If customers don't think my idea is good enough to fund, then my idea simply isn't good enough." Some people might think this is a harsh standard to meet, but the reality is that if customers don't believe in your product then investors won't (or shouldn't) either. I'm certain I could've persuaded investors to fund my idea. But imagine developing and marketing a product for six to twelve months and then having no customers, running out of money, and facing the prospect of either closing up shop or asking the investors for more money - neither of these is an attractive prospect. It was really unclear if my idea would succeed and if I took investment capital and never got a single customer, or if Human Element became an endless money pit, my credibility in the entrepreneurial and investment community would be badly damaged. Consequently, my ability to return to the community with a future venture idea would also be damaged. That's bad karma.
- Even so, since I was planning to succeed, I was still actively speaking with angels and VCs. Come on, this is a business we're talking about and if investors come along and propose to add value in the form of funding, that's a conversation you need to have. In my opinion, anything less would be remiss and perhaps even management malpractice (unless you know with absolute certainty you won't need the money or you have a maniacal disregard for reality.) So it's a fact that I was working my networks and reaching out to people known for making investments, angel groups in the Boston area, and I was having informal talks with VCs in an effort establish awareness in that community, if I ever needed to go there.
- I had (and have) a very structured view on how to build shareholder value. Every business person is an advocate for what their company offers (or at least they're supposed to be.) That means you're an advocate to customers, employees, partners, the media, and of course, investors. Investors are a tricky audience because dollars (sorry - Euros, Yuan, cigarettes...) define the value of a transaction (investment), and since we're trying to predict the future when starting a new venture, it can be difficult to identify reference points for this value. My approach is to break the process down into steps, or phases that might exist in a project management approach. Each discrete phase (which presumably uses up some cash) should add value to the venture. By using other ventures as analogs, it's possible to start building a framework that can anchor a discussion related to value. I'm planning a future post on this topic, so look for more on this later.
- What's the exit plan? To this day, it kills me when people say they plan to go public (and this has happened within the past twelve months.) Even during the dot com boom, most successful companies were acquired - they didn't IPO. I think the ratio was around 4:1 (acquired:IPO). For those of us in the real world, acquisition vs. IPO isn't the question - the question is, from the outset, what potential acquirers can be identified, what specific gap in their portfolio will your venture fill, what markets will your company open for the acquirer, and what's that likely to be worth? Obviously "I don't know" won't cut it, but at least knowing the questions will help your credibility.
Coming soon: Recruiting a CTO, and making some painful realizations.
Posted by David Provost at 7/13/2008 12:14:00 PM