Dan Keldsen Interviewing Dennis Cahill of Factiva, discussing "Agile Information Intelligence Solutions" – a concept combining search technology and strategy, with metadata and governance, which leads to discovery and true “intelligence” – reacting to workers demands but also proactively “pushing” or alerting on a continuous basis.
What can "agile development" teach the enterprise about enterprise search, among other things? Listen and find out!
Full transcription is available by following the permalink to the full entry.
Complete Transcription follows:
Dan Keldsen: This is Dan Keldsen, Senior Analyst and consultant with the Delphi group in Boston, Massachusetts. We can be found at www.DelphiGroup.com or give us a call at 617-247-1511 or 1-800-DELPHI-0 which is 1-800-335-7440. This interview is with Dennis Cahill, who is V.P. of Products for Factiva.
We are going to briefly discuss "Agile Information Intelligence Solutions", a concept combining search technology and strategy with metadata which leads to discovery at true quotes intelligence reacting to workers demands but also proactively pushing or alerting on a continuous basis. Dennis , can you provide some quick background on Factiva and your own background to back up for our conversation?
Dennis Cahill: Sure. Let me tell you a little bit about Factiva. At the heart of Factiva is a content archive of over 10,000 authoritative news sources from around the world. We currently receive content from 122 countries in 22 different languages and our content archive is over 400 million documents that go back over 30 years.
What Factiva does is builds business solutions around that for Global 400 type customers. We currently have over 80% of the global Fortune 500 and our goal is to combine this rich content set with innovative technologies which achieve truly drive better decision making by those customers.
Dan: Now, Factiva has been around for a fair amount of time and is quite the consumer of both leading edge technologies which is competitive to the more established technology. At least for you folks, if not for the rest of the enterprises out there in the world and search in particular, than the ever present human knowledge worker because not everything is automatic as hopefully everybody realizes.
Now in the context of information intelligence which is our roll up and Factiva you are talking about the concept of agile information intelligence solutions. So, let's start at what does that even mean? What does agile mean and in your context - just a little background.
On my part, this week apparently it's agile week for me. Just yesterday I was at a usability conference locally here in Boston, where agile development as a concept was mentioned and most people were quite unfamiliar with it. It is only just being discussed in those circles which is kind of ironic since the usability crews of the world tend to work pretty hand-in-hand with development teams.
It seems like that concept hasn't quite permeated into the development world and on the enterprise side either. What does agile mean within the context of Factiva?
Dennis: Let me take this in two bits. Let's start with the agile development and then we'll talk a little bit about how it will affect enterprise search.
Dan: Yep.
Dennis: We jumped into the agile development space probably about three and a half years ago when prior to that we had been more traditional. We would go out and interview tons of customers and spend eight or twelve months collecting all of this great information about what a customer wanted. Then we would spend another six months building it and getting it to customers.
Dan: Right.
Dennis: What we found is that by the time we got back to the customers their wants and needs had changed. We also found that they couldn't really express their wants and needs very well.
Dan: Right.
Dennis: So, we switched to an agile model. It's funny when we first got into it we focused on it from the perspective of allowing us to develop faster. What we found is we started breaking everything up and saying let's get to market quicker data type solutions. Let's get them in front of customers and let's react to customers' immediate input on what the solution is.
What we found by doing that is it's not so much about getting the development more efficient. It's more about getting the right product to market much more quickly that we are allowing our customers and the users of our tools to drive the real value ads that they see in the product line. It's less about the guess work of research and it's more about direct customer involvement and feedback.
What we found is that when we get to done a whole lot quicker and that customers can really focus in on what they want. I would say the easiest way to get a customer to tell you what they want is to show them something that isn't it.
Dan: Right.
(laughter)
Not quite it anyhow.
Dennis: Yes, right. Just show them something that's close enough and they will tell you exactly what you need to do to it.
Dan: Right.
Dennis: We found it to be a tremendous success that customers are really ready for this type of involvement, that they are ready to see products that aren't fully ready for prime time. They are ready to see the early views of them and they are ready to work with you on getting them to where they need to get to.
I think that's a great model for anybody who is launching any sort of new software based capability today. It really maps the strength of software development which is the very quick turn around type capabilities that you can do when you are working with software with the needs of a customer.
Dan: Right.
Dennis: The reason I think that it maps very well to enterprise search is that when I look at enterprise search and I look three or four years down the road I see something that doesn't resemble enterprise search today at all. I think when we go three and four years down the road enterprise search is not going to be ‘enter some key words and get back a list of articles or a list of records in a CRM system'.
When we see enterprise search, we see holistic dashboards that are stellar to specific job functions that combine all of the information assets in an organization that truly drive efficiency for the job function that is using the tool. It's going to be less about searching and more about browsing and understanding the content that is at your disposal and all of it.
It will be the rich, competitive intelligence information that exists out on the Web today. Combine that with the information on customers that sits inside your CRM systems. Your purchasing systems combined with the breaking news from premium sources like Factiva also combining the internal documents like the latest Microsoft Word document written about that customer, or Power Point that exists for a presentation that has been given to a customer.
Dan: Okay.
Dennis: I see an environment where all of these are richly integrated into a dashboard specifically targeted at a job function. It's all about driving that efficiency that allows those job functions to competitive advantage to drive more efficiency within their work processes.
Dan: Okay. Now I think dashboard is the best term because that seems to come from a sort of business intelligence sort of world versus the world that I am typically more comfortable with personally, which is enterprise portals. Is there a difference?
Dennis: There really isn't. It's nomenclature. The reason I like dashboards is that it applies. When I think of enterprise search, I think the mantras behind business intelligence map very well to the problems that unfortunately, business intelligence in the past has largely been focused on structured data and so much of the great information that is out there.
When you look at the information that is sitting on message boards and blogs today around competitive intelligence, it's just staggering. You are really looking at enterprise search as a way to truly expose that.
If I am a competitive intelligence professional, I am not going to need to run to five different products and run five different searches. I'm going to have the latest blog hits around one of my competitors in a specific issue that I'm tracking directly in my environment. It's not going to be so much about reading unstructured data. It's going to be about technologies that extract facts and provide analytics around that data that truly drive the efficiencies.
Dan: Yeah, okay. I know the reason why we have coined the term ‘information intelligence' is because we did see that business intelligence had a whole lot to offer except that it is for the most part still, if not in fact, it's still perceived as being very much about structured data. So I was just curious what you saw about that.
Dennis: It's a great way to explain it to people because when people see business intelligence, when people hear business intelligence, they don't think of ‘let me look at thousands of little records'. They expect some report, something that is telling them the high level story around a decision they need to make.
When we look at unstructured data, we need to look at it the same way. It's not about a list of headlines. It's about the facts that those headlines are telling me. It's really about driving the decisions around that content that is the whole.
Dan: Right. So we talked about enterprise search a bit. What other technologies are merging that can help with this and that you are either using right now or that is coming down the pipe?
Dennis: This is obviously a very rapidly evolving technical space. There are very strong taxonomy and ontology tools out there today both the generation of taxonomy, ontology, the application and the maintenance of those. We find those to be very strong. We use them today. We find that a very good way to drive value around information management within the enterprise.
There is also a whole suite of technologies that are evolving very rapidly in the entity and fact extraction and link analysis space where defined dictionaries and thesauri are being provided with these tools and make them much easier to drive into the enterprise and utilize.
Then probably the area that we are most excited about is the statistical clustering and categorization techniques that are out there that truly allow us to check themes within piles of unstructured content and then map those themes directly to structured information that exists within the enterprise.
Dan: Right. There has been a whole lot of interest in taxonomies and we've been doing our proving ground for four or five years now, both publicly and privately, which is largely about taxonomy but ontologys and thesauri and all sorts of other controlled vocabularies that are adding value to search and content management portals and all of that kind of stuff.
Now, a sort of obvious question here is what is driving any of this? Where is the money coming from as far as why are your customers looking into this? What do they see as the advantage? How do they justify their investment?
Not everybody is a visionary or reports for a visionary anyhow and the hard and soft savings are frequently what people are looking for, which drives me insane, because people aren't in business to drive costs down. They are in business to make more money. So why don't they see both sides of the coin? What are you seeing from your customers and what you are using?
Dennis: Yeah, I think you just hit the nail on the head regarding the way enterprise search is treated today by the masses. I think the masses treat enterprise search as a tool that is needed within the enterprise, but one that they look at from the point of reducing costs and keeping under cost control. They don't view it as a driver of their business.
Factiva has one luxury and that is we tend to work with very large global companies, global 4000 companies that tend to already respect the value of great information. They already value the competitive advantage that great information can give you.
What we have seen over the last five years is that they have actually pushed us. They have seen their vision in this space and to them it's about doing it better. It's about competitive advantage and they see all of these rich information assets that exist for them. They know that they can drive their own competitive advantage by understanding how to use that content better.
An example which several large companies have come to us to work on is their lead generation processes. A good example is where historically they use structured data bases to do lead generation where they would go in and say, "Find me all of the companies in this industry that have more than 500 employees on the East Coast."
Dan: Right.
Dennis: Then they would do lead generation off of that. Now those companies come to us and say, "I want to mine your news feed. I want to extract companies from your news feed. I want to map those companies to structured data the taxonomy, the symbology that Factiva has and I want to make this an alerting service around lead generation, not something that occurs once a quarter or once every half year when the destructured data is updated."
Dan: Right.
Dennis: So, that's a good example to show the real way that effective use of all of your information assets drive competitive advantage. It's not so much about having an enterprise search engine.
Dan: That is probably the point of all of our role in information intelligence. It's not like search is going away but most people's conception of search is pretty narrow. They are not fully aware of searching just key words, to not even just synonym rings with thesaurus and other things, make your results smarter but frustrating and ontologies and all of that so as you drive it into a larger solution or view like information intelligence or competitive intelligence or some sort of dashboard or something like that the value is much magnified. It's not a linear addition of capabilities. It just drives you to a whole other set.
Dennis: We absolutely agree. We tend to have the luxury of working with some pretty big organizations and you can see where they are taking this. They view information the same way.
I often hear people describe this problem as information overload. I think it's the worst name that ever existed. When you look at the amount of information that is available to organizations today it is so compelling. You basically have customer service logs of your biggest competitors sitting on blogs and message boards out on the Internet.
Dan: Right.
Dennis: Organizations that can truly execute solutions that will allow them to drive decision making around that are going to have such a big competitive advantage. With the big global companies around the world you are seeing that type of interest. They understand where this can take them.
Dan: Great. Now, why Factiva? Why would somebody come to Factiva to use any of your solutions? What is in it for you? What drives you as a company?
Dennis: I wish that I could tell you this is some big strategy of Factiva's over the last five years and it's coming to a head. The fact of the matter is that it's our customers who have pulled us into this space. They recognize what Factiva has done with their products and they have pushed us to do it for them with their content around Web content, their internal content.
They want it behind their firewall so they have actually come to us and really pulled us into it. That caused us to form our advisory and technical consulting groups, probably about three or four years ago now, to really go in and help organizations do what Factiva has always done but do it to a broader set of context than within their firewalls.
Dan: Great! Well, I wanted to keep this podcast a little shorter than the more recent ones that I've done so I think we will stop now although we could go on for quite a while about where people are at right now and where people are going. The future is looking pretty interesting. There has been just a lot of raw research and actual commercially deployed solutions in this whole general space of search and information intelligence as a larger role up.
I think as much as I hate the ‘information overload' term on some days it seems worse than others, but people can stop drowning for a second and take a look around and see what's out there. There actually are quite a lot of capabilities out there that can be tapped that are not free but they are also not hideously expensive. Especially when you consider the advantages that they can bring to you as long as you are trying to really solve a business problem.
Thanks for your time today Dennis: . For everybody listening or reading the transcript afterwards for more information on Factiva go to www.Factiva.com or 800-369-0166 and once again this is Dan Keldsen, Senior Analyst and Consultant with the Delphi Group in Boston. We can be found at www.DelphiGroup.com or give us a call at 617-247-1511 or 1-800-DELPHI-0 which is 1-800-335-7440.
Your feedback on this podcast or the transcription or the concept of information intelligence and "Agile Information Intelligence" would be greatly appreciated. Please join the conversation on my blog which you can find at http://DelphiGroup.blogs.com. Thanks again Dennis.
Following up on Dennis's comment that "information overload" is the wrong focus. I agree. I think people are too quick to say "there's too much information out there." What they're really saying is that "search" isn't the right methaphor for harnessing billions of pages. Search can surface the 10 best pages, but what about all the other pages? Text mining helps you leverage the rest by learning what's being said in the aggregate.
Posted by: Glenn Fannick | April 03, 2006 at 04:16 PM
Glenn - Thanks for the commentary - and yes, without appropriate tools, information overload is a definite possibility, but with smarter, or more "intelligent" tools, summarizing, text mining, fact extraction, etc. - information becomes an enabler rather than a disabler.
Posted by: Dan Keldsen | April 03, 2006 at 04:45 PM