I must admit that my favorite movie slogan ever belongs to (yet another) bad sci-fi movie that I’ve featured in this blog — Robocop: “Half Man, Half Machine, All Cop.”
I like it for a few reasons:
- It’s a clear, powerful piece of positioning
- It’s an effective positioning for a hybrid product
- It’s a situation that arises often in the marketing of hi-tech products
For example, at Business Objects I had the challenge of positioning BusinessObjects 4.0 which was the first product that combined classical Query & Reporting (a la BusinessObjects 3.0 or Cognos Impromptu) with basic OLAP (online analytic processing a la Cognos PowerPlay). We were providing a single product that combined query, reporting, and OLAP functionality; we integrated into one product what were two different products in our competitor’s product line.
Some people rightly questioned the use of the word “integration” in our launch because it sounded like we were fusing two separate products into one; integration connoted weaving or welding, so people might look for the seams. In reality, there were none because we had built one product that provided all three functions. We were not bolting two different things together.
As a critical thinker, my concern with “integration” was different. I wondered if you could reasonably argue that “integration” was always a good thing. In my opinion, you couldn’t. I used two canonical examples to make my point.
- The clock/radio, which has virtually eliminated the market for bedside alarm clocks.
- The car/boat, an invention that has been tried again and again, but has never found commercial success.
The former just works. The latter presumably suffers from a worst of each problem: it floats like a car and drives like a boat.
Siebel comes to mind. They wiped out the independent markets for salesforce automation (SFA), enterprise marketing automation (EMA), and call center applications by uniting them into a single category called CRM. Unfortunately, Oracle and SAP have done unto Siebel as Siebel did unto others — by then uniting CRM into ERP / enterprise applications.
Sagent comes to mind as well. Sagent was a startup launched in the mid/late 1990s on the quite reasonable premise that integrating ETL tools with BI tools was a good idea. As a simple example, an ETL tool knows a lot about the data it manipulates — where it came from, when it was refreshed, what transformations were applied, etc. If one could make it easy to access this metadata in a BI tool, then you’d have a better BI tool.
I always found Sagent paradoxical because commercially I knew they would have trouble declaring a war on two fronts — against both BI and ETL leaders. Technically, it struck me as a good idea, and one that has since widely implemented.
So often in hi-tech the value comes from integrating things in new ways, so it’s worth spending a lot of time trying to characterize and think about the different forms integration takes.
At Mark Logic, our product has elements of DBMSs and enterprise search engines. As I’ve said before, from the outside our product looks like a DBMS — in goes a query language and out goes data. It supports transactions, concurrency, recovery — all the things you’d find in any DBMS. When you open up the hood, however, it looks much more like a search engine.
In fact, our VP of products, Ian Small says, “if you were to pop your head up into random places in the code, sometimes you’d think you were in a search engine, and sometimes you’d think you were in a DBMS.”
It is that hybrid nature that makes MarkLogic so unique and so powerful. It’s not “integrated” in the same sense that BusinessObjects 4.0 wasn’t integrated — it isn’t a search engine and DBMS welded together. It’s a special-purpose DBMS, built from scratch, that makes use of search engine scaling and indexing techniques. And the design point is to optimize the system for content, and lots of it.
So, when I have to boil it down to a soundbite, I come up with this.
MarkLogic Server: Half DBMS, Half Search Engine, All Content.
What do you think?