Archive for the ‘Uncategorized’ Category
Business intelligence is too focused on internal data and doesn’t help decision-makers understand their external business environment
When you’re driving a car, you spend most of the time looking out the windshield: What’s ahead? When to turn? Any unexpected obstacles in your way? Every now and then you glance at the dashboard. But that’s just to check how fast you’re going, whether you’ve got enough gas and that the engine isn’t running too hot.
In BI, it seems to be the other way around. BI traditionally focuses on data from internal systems, providing a view on internal operations and past performance. Data on the external business environment and the future belongs to the fragmented world of corporate, strategic or market intelligence with their proprietary databases, pivot tables, PDF reports and custom research delivered in static slide decks.
In other words, decision-makers in the enterprise spend a lot of their time looking at the dashboard — sometimes literally — but only have a blurry and fragmented view of their surroundings: the markets they operate in, the economies they belong to and the demographics they target.
There is a lot of good data out there, from public and proprietary sources alike. Government databases are opening up and contain more valuable information than most people realize. Syndicated research — trackers, forecasts and surveys — is plentiful but hard to find and quickly gain insight from. And data from custom research, whether internal or from research vendors, is usually delivered in static formats. As a result, too much of it ends up sitting on hard drives somewhere with no good way to search, compare or access later – let alone to keep an eye on updates to the underlying data.
This is fundamentally inefficient. It means that decisions aren’t made with reference to the best available data; time is lost digging through piles of static documents; and companies are unable to make the most of the sizable investments they’ve already made in market intelligence.
I believe this is the next frontier for intelligence systems. The task is to provide a window on companies’ external business landscape, expanding BI as it has traditionally been used to encompass the rest of the universe of corporate intelligence.
- – -
Note: This was originally published as a guest article on DailyTekk in May 2013.
Below are the slides from my presentation at European Data Forum in Dublin, April 9 2013.
These are the slides from my lightning talk at the Hacks/Hackers meetup in Boston, March 28, 2013
About a year ago DataMarket set up a sales and marketing office in the Boston area. At the time it was just me, but we’ve since grown to a team of three and plan to have 6-8 people working out of the region before the end of the year.
Yesterday we moved into new offices on 184 High Street, right at the edge of Boston’s financial district. The house is an old lead works building, originally built in 1886 and still sports the original ornaments and signage. You can read more about the building’s history here.
These pictures were taking during our move-in yesterday. The space is – for obvious reasons – still quite empty, but it does have a great spirit. We seriously look forward to continue building something great here!
Interesting side note: Our first office here was in the Kendall Square area in Cambridge, known as a hot-bed for startups and innovation. We wanted to stay in the area, but found that were not a lot of vacancies and the prices had soared since a year ago. The reason is apparently that all the big guys – Microsoft, VMware, Google, IBM and most recently Facebook, to name a few – have all been expanding or setting up presence there, driving up demand and thereby price. At the same time Boston’s Financial District is still recovering from 2008 and has a lot of vacancies. Personally I’ve heard of several start-ups that have recently moved or are about to move from Kendall Square into the Financial District, for the very same reasons. While backed up by a couple of other people I’ve talked to, this is still only anecdotal evidence, but it is interesting if the “cool kids” are moving out because of the big boys that want to hang out with them
There are currently 13 different chart types to choose from when exploring data on DataMarket. Several off them are variations on the same basic chart types (line charts, bar charts, tables, etc.), but each variation has its strengths and weaknesses given the data at hand and the point you’re trying to illustrate.
While written with our own users in mind, this overview page might be helpful for anyone working with charts – whether on DataMarket or not:
To all our users, friends and followers:
Thanks for a wonderful year of data and joy,
The DataMarket Nerds
Slides from a presentation I did at the Data Scientist Seminar Series in Boston, December 3 2012.
Note that most of the images are links to relevant data, technologies or additional information, so click around!
Below are the slides from my presentation “Best Practices for Publishing Data” given at the Strata Conference in New York, October 2012.
The slide deck includes a lot of links to additional resources, so go ahead and click around.
Note that this is an enlarged and improved slide deck from a presentation with the same title from Strata London.
Last week DataMarket introduced a new product, an energy specific data service called simply: DataMarket – Energy
The venue at which we introduced the service was quite unusual. We were lucky enough to be invited – along with a selected group of other startups and innovators working with energy data – to present our work at the White House at an event called Energy Datapalooza. We have since jokingly said that in order to top this venue for our next product announcement, we will have to book the International Space Station. I’m working on that.
Here’s a short video of my presentation there and the unveiling of our new service:
For those of you that have been following DataMarket for a while, you will notice that the business model for this new product is significantly different from what we have previously been running with.
When we originally kicked DataMarket.com off with international data early 2011, there was only one thing users could pay us for: A low-priced premium subscription that gave access to additional features, such as more advanced data export formats, automated reports and a few other things. A couple of months later we added the first premium data to the site; data from premium data providers such as the Economist Intelligence Unit (links to EIU data on DataMarket), resold through our site.
However, using the site’s core functionality – the ability to search, visualize, compare, and download data from the vast collections of Open Data that we aggregate – has always been free. As such, DataMarket.com has become quite popular in certain circles. But quite frankly, the two revenue sources have not taken off in a big way.
What has however taken off is our technology licensing business. We’ve seen high demand for our data delivery technology from other information companies. The ability to normalize data from a wide variety of data sources, and enable users to access that data through powerful search and online visualization tools is something many information companies, such as market research and financial data companies, have identified a strong need for. So last February we formally introduced our data publishing tools, most prominently what we now call the Data Delivery Engine, a white-label solution that is already up and running for a few well know information companies, (including Yankee Group and Lux Research) with several other in the implementation stages. This licensing business is where most of our revenues comes from today, so one could really say that we’re now more of a software company than a data company.
The upcoming launch of DataMarket – Energy is another stab at the data side of the equation, but the approach is different in several ways:
- Focus and scope: By focusing on a single industry or vertical we can make the service much more relevant to its users. Instead of solving 10-15% of everybody’s data needs with the kind of macro-economic and demographic data that can be accessed on DataMarket.com, we aim to address 90-100% of the data needs of a much more targeted audience.
- Premium access: We’re selling access to this service at a substantial premium (final pricing is still being decided). Those that see value in the discovery and aggregation services that we add on top of the data will be charged for the “job they hire our product to do”. This indeed means that some data that has been made publicly available for free (Open Data) will only be available to DataMarket users behind a paywall. As explained in the presentation above, that doesn’t take the least bit away from the value of the Open Data. On the contrary: The data is still available in its original form from the publishing organizations, but we add a choice on top of that: A nicer and more user friendly way to access the data for those that are willing to pay for that value-add.
- Targeted sales: Instead of relying as much on PR and viral distribution as we have with DataMarket.com, we’ll use more direct, traditional sales approaches for this new service.
One of the interesting things about running a technology startup is that the same technology can be turned into so many different products without a single line of additional code. Often the only difference is how you promote it, price it and sell it. This can be both a curse and a blessing, and usually a few things need to be thrown at the wall before you find what sticks. Luck is involved too, but as the famous
Norwegian Swedish alpine skier Ingemar Stenmark is quoted saying: “The more I practice, the luckier I get“.
It will be interesting to see if we’ve practiced our data marketing skills enough for the DataMarket – Energy approach to work out.