Marco on CEP

In 1989 GM bought Saab Automobile (The car company part of Saab, not the fighter jet part). One would assume that the people set to lead GM are among the best leaders and smartest corporate leaders available. Now in January 2009 GM is shutting Saab down and from what I understand they never made any profits from Saab at all.

Does this sound like something smart people would do? Given the power of GM and 10 years, how come they totally failed with Saab?

 

Who’s Getting Who?

In 2010 there will be a number of smart people buying CEP companies. A post on predictions would not be complete with some guesses on who will acquire the most interesting companies.

In 2010 I predict that CEP will find its way to people outside the core event processing community. As a consequence people with no background in CEP will find out what we CEP companies have been doing the past years. Those that will find these companies or their technology the most interesting are probably those that are drowning in real-time data. I would not be surprised if a major stock exchange, something like Reuters, SCADA company or an integration company buys one of the CEP vendors in 2010. Actually my best guess is that a CEP company is bought to be used as a base for SCADA solutions. SCADA is almost never mentioned when talking about CEP, but that’s where there will be some action in 2010.

If would need to name a company that I think will be acquired first, it would be Aleri. Maybe really soon? Who knows. The open source project Esper is also a good candidate to be bought or assimilated by another open source company. Maybe RedHat or Sprinsource are in the mood for some shopping?

The New Generation

Considering the small amount of pure play CEP vendors available, will there be anyone left after the wave of acquisitions that I think will happen in 2010?

This brings me to my next prediction. New CEP vendors will emerge in 2010! I think we will see several vendors with interesting ideas on how to do CEP. These vendors are in a good position as they can learn from the current vendors and take the ideas and develop them further without any concerns for existing users.

I would guess that at least a couple of them will fail, but maybe one or two will do OK and it will be among those that the new cool ideas will come from. I don’t think open source will lead the way here. From what I see, the open source community is good at taking existing mature ideas and creating good implementations of them. Not inventing new concepts and ways of doing things. Something that is required for the next wave of CEP. I have the distinct feeling that the current CEP tools are just the first wave. There will be at least a couple of waves before things settle and we get some kind of collective agreement on how CEP should be done. Compare to SQL in the database world.

The Great Escape from Wall Street

To many people CEP equals algo trading and other kinds (legal) tricks for gaming the markets. Several event processing companies talk a lot about capital markets and less about other application areas. It’s not that there’s a total lack of efforts in trying to capture other markets. But still, capital markets are where many vendors have done OK.

In 2010 I think that many of the leading CEP vendors are trying to figure out how to break out of the world of finance. I think they would love to find new markets and have started to probe into various fields in 2009. In 2009 many companies at Wall St. took some beating and I bet this made the life of the sales reps at the CEP vendors a bit more, well, interesting. My prediction is that they won’t come that far with the great escape in 2010. I think the vendors will find that the capital markets will suck up all their resources and moving beyond selling very expensive solutions doing all kinds of financial magic and trickery is not at easy as is sounds.

Driving in Snow

This year I have been driving in snow for longer than most winters. And the winter is not over yet. We have had temperatures around -10 C for a long time here in south Sweden where I live. Just a couple of hours to the north, they had -40.4 C (about the same in F) this night.

Here my idea was to have some clever transition from snow driving to CEP, some authors are really great at doing that. But it seems that I’m not one of them. So I’ll continue just a little bit more about driving in snow before getting back to CEP.

The above tire is actually from my very own Saab. I read in the news that today is the last day for Saab. So now I’m driving a car which is not produced anymore. It will be interesting to see what that does to prices and availability of spare parts.

I think my transition had to do something with Saab being discontinued or that Saab have always had a performance profile. The Turbo Saab is somewhat of an icon here in Sweden.

The Cult of Performance

Whenever you talk to with other vendors, system integrators, clients, prospects that have heard about CEP they constantly focus on performance. It seems that the main perceived benefit of CEP is performance, and performance alone.

It’s simply seen as a way to do things faster than competing offerings. This is both good and bad. It’s good for CEP to be viewed as something that is needed when high performance is required. But it’s bad because Moore’s law will outrun us. If performance is the only selling point, then you can be sure that a normal desktop PC will replace you within a couple of years. Also, compared custom built code in, say C, any generic engine will most likely be slower.

But in 2010 I predict this will change. In 2010 CEP will find uses in other areas where speed in the not main selling point. I think that agility is a word that will be used more and more when talking about CEP in 2010. CEP solves many problems in a natural way, even if they could be solved using other tools. This is something that will be discovered by users outside the CEP community. We, the CEP community, know this already so it’s not a very bold prediction.

Show Me The Money

2010 will be the year when all kinds of psychedelic business ideas with no chance of ever generating profits will fail. Now it’s time to start making profits. Now more playing around. This is good news for CEP and bad news for Twitter and others. CEP is what I call a old school business idea. That is, there is a plan how to make money by providing something of value to a business. Our clients make money, we make money; everyone is happy.

Still, 2010 will not be an easy ride. Considering the amount of investments made in many of the CEP vendors, there’s bound to be some pressure to increase revenue and start making considerable profits. For example, a wild guess would be that the StreamBase’s investors would like to get back at least $600M. So how much profit do you need to make each year in order to justify that valuation? There’s lots of CEP licenses that need to be sold. The dependence on the capital markets can be a problem here. A guess would be that at least one of the current vendors will see some revenue related problems in 2010 in the CEP area. For the smaller vendors mainly doing CEP this can be disastrous. For bigger vendors where CEP products is just one part of the portfolio you might expect re-branding and other kinds of creative moves when they discover that selling CEP takes more time than they thought.

It sounds like I think that there’s no market for CEP. There is. But viewed from a customer point of view the CEP market is in a tricky phase. No standards, no agreed upon way to do CEP, we don’t even agree on what CEP is. This is confusing for the customers and I think many of them will build instead of buying during 2010. But the demand will increase, CEP solves real problems providing real benefits.

A Major Event

In 2010 I think something critical will happen that puts CEP in the spotlight. A customer success story or something that the press will write about. This will be the starting point of spreading CEP.  Unfortunately I think that the news about CEP that everyone will talk about can be rather negative. Considering the greed culture of Wall St. and the focus of CEP on capital markets, I would not be surprised that some clever uses of CEP will play a major rule in another PR disaster at Wall Street.

Another, not very bold, prediction is that there will be at least two CEP books published in 2010.

 

 

 

Complex Event Processing