(Another delayed report from a talk last Friday at the Web 2.0 Expo.)

Joe Keller is the marketing officer with Kapow, so I was expecting a marketing talk, but there was a nice amount of technical content to keep most happy. Joe was talking about “getting business value from enterprise mashups”. Kapow started off life as a real-estate marketplace in Europe ten years ago, but moved towards its current focus of mashups after 2002. Referencing Rod Smith, whom I saw last year at BlogTalk 2006, mashups allow content to be generated from a combination of rich interactive applications, do-it-yourself applications plus the current scripting renaissance.

According to McKinsey, productivity gains through task automation have peaked, and the next productivity wave will be data-oriented as opposed to task-oriented. Joe says that Web 2.0 technologies are a key to unlocking this productivity. He also talked about two project types: systematic projects are for conservative reliability, whereas opportunistic projects (or “situational applications” to use the IBM terminology) are for competitive agility. Mashups fit into the latter area.

The term mashup can apply to composite applications, gadgets, management dashboards, ad hoc reporting, spreadsheets, data migration, social software and content aggregation. The components of a mashup are the presentation layer, logic layer, and the data layer (access to fundamental or value-added data). In this space, companies are either operating as mashup builders or mashup infrastructure players like Kapow.

The main value of mashups is in combining data. For example, HousingMaps, the mashup of Google Maps and data from Craig’s List, was one of the first interesting mashups. The challenge is that mashups are normally applied to everyone’s data, but if you’re looking for a house, you may want to filter by things like school district ratings, fault lines, places of worship, or even by proximity to members of your LinkedIn / MySpace network, etc.

He then listed some classes of mashup data sources. In fundamental data, there’s structured data, standard feeds, data that can be subscribed to, basically stuff that’s open to everyone. The value-added data is more niche: unstructured data, individualised data, vertical data, etc. The appetite for data collection is growing, especially around the area of automation to help organisations with this task. The amount of user-generated content (UGC) available is a goldmine of information for companies, enabling them to create more meaningful time series that can be mashed up quickly into applications. According to ProgrammableWeb, there are now something like 400 to 500 mashup APIs available, but there are 140 million websites according to NetCraft, so there is a mismatch in terms of the number of services available to sites.

Kapow aims to turn data into business value, “the right data to the right people at the right time.” Their reputation management application allows companies to find out what is being said about a particular company through blogs, for sentiment analysis. They also provide services for competitive intelligence, i.e., how do you understand the pricing of your competitors in an automated fashion. Asymmetric intelligence is another service they provide for when people are looking for a single piece of information that one person has and no-one else possesses. Business automation is where mashups are being used to automate internal processes, e.g., to counteract the time wasted by “swivel-chair integration” where someone is moving from one browser on one computer to another and back again to do something manually. Finally, opportunistic applications include efforts whereby companies are aiming to make users part of their IT “team”, i.e., by allowing users to have access to data and bringing this into business processes: Web 2.0 infrastructure allows companies to use collective wisdom using Kapow technologies.

About RSS, Joe said that almost every executive in every corporation is starting to mandate what feeds he wants his company to provide (and RSS feeds are growing as quickly as user-generated content in blogs, wikis, etc.). Kapow’s applications allows you to create custom RSS feeds, but he gave a short demo of using Kapow to build an on-the-fly REST service. His service produced the quote for a company’s stock price by extracting identified content from an area of a web page, which could then be incorporated into other applications like an Excel spreadsheet. I asked Joe if it is difficult to educate end users about REST and RSS. He pointed to the ease with which most people can add feeds to iGoogle and said that its becoming easier to explain this stuff to people.

Kapow’s server family allow portal creation, data collection (internal and external), and content migration via mashups which Joe reckons are often more useful than static migration scripts since they can be customised and controlled. Kapow also provide a free “openkapow” API and site for developers to share how they build mashups and feeds.

In summary, Joe gave these take aways:

• The next business productivity wave will be via data and know-how automation, not routine task automation.
• Knowledge workers need self-service mashup technology to take advantage of this.
• Access to critical (value-added) data can create a competitive edge.
• Web 2.0 technologies complement existing IT systems to maintain the competitive edge.