Wednesday, December 12, 2007

Postgres 8.3 is smokin' fast!

Stefan Kaltenbrunner has been testing and comparing the latest beta release of PostgreSQL 8.3 against PostgreSQL 8.2 and has some significant results to show.  Though the data is anecdotal, the results are fantastic.

POstgres 8.3 performance

Stefan's configuration information and simulation details are in his post, but here is a quote:

The benchmark database is initialized with a scaling factor of 100 (equals to 10M rows) which seems to be a reasonable size for a table in an OLTP style database.  All testing was done with 100 clients and 100,000 transactions/client which comes out to 10M transactions and an average runtime of about 1.5 hours.

Nice results.  You can see all of the details here.

And be sure to read the comments.  Here is a comment from a MySQL person who finally "gets it":

...when I write to the pgsql mailing lists, I get a response--and usually a VERY informed one--within an hour. Often within 10 minutes. Try getting that response time from the MySQL community on their forums online. You're lucky if your thread gets answered in that huge labyrinth of forums at all. As for the documentation, the website has a great deal of stuff, and whatever isn't there, just search the email archives (many such archives exist), and if that's a pain too, just drop a note to the pgsql list. You'll have an answer or excellent pointers more quickly than you imagine. I'm a recent convert to PgSQL (8.2.3) and while it took me a month to really figure things out and tune my database, I am simply amazed at the way things are structured, at the performance, and at the resilience/functionality offered by this stupendous db. (Having been a devout MySQL fan for over six years now). There's a reason a product creates passionate zealots.

Monday, December 10, 2007

Improving Google Adwords Results

In the world of B2B marketing, just about everything happens "on the Net" these days.  And Google is at the top of the list.  Adwords are being used to advertise just about anything and everything on the web today and our company spends a significant amount of money on Google Adwords.  The number of click-thrus is impressive, but some lower level analysis is needed in order to make sure that the clicks are valid.

Here are some insights that I have learned over the years from both my prior and current jobs.  Nothing is rocket science here, but it is often overlooked.  The best marketing is simple and well thought through. 

First things First.  If you are not breaking up your Google Adwords spend into campaigns, then you have no visibility.  Use campaigns to create a laser focus on WHO you want to attract and WHAT will attract them.  Assume you sell sporting goods and want to advertise footballs.  Instead of just targeting footballs. create different campaigns with similar ads and do some A/B testing to see which ads perform better.  Focus your efforts on a certain type of buyer - adults with male children ages 6-10.  Sounds difficult, right?  But, figure out where they go and then you will get better results.  Make a buyer persona.  Spend some time identifying what the hot buttons are for your customer.

For us, there are multiple hot buttons and they are clear:

  • They have been raked over the coals by Oracle and need an alternative
  • They want to use open source, but MySQL just can't get it done
  • They are using Postgres, and they need some help

Next, make sure that your ads don't just drop them on the home page.  This is probably the single biggest mistake that people make.  You have a targeted ad, you need to send the prospect to a targeting page.  That page needs to reflect the message found in the ad and need to provide some compelling content.  Back to the football.  Instead of dropping the prospect on your home page that may advertise lots of different things, send them directly to the page with the football.

And finally, give them something to do.  There needs to be a call-to-action and it needs to be "above the line".  When I say "above the line" I mean that the user should not have to scroll down to see the call-to-action.  Maybe its an offer to download your product or read a white paper.  In the case of the football, it should simply be a way to order the product right there.  Depending on the cost and complexity of the item, the call-to-action may or may not be a "Buy Now" button.

Check out your current Adwords campaigns and see if you have done these three things.  If so, great.  Now run some A/B testing and improve upon it.  Keep up with Adwords, change the ads frequently, experiment with different offers and call-to-action. 

If you are not, then spend the time to do it now.  You are throwing good money away.  You may get clicks, but I guarantee they leave quickly and never get your point!

Friday, December 7, 2007

In-flight Internet Access coming...finally

JetBlue will start a pilot program (no pun intended) to enable limited internet access on one of their planes next week.  Though the offering is on one plane and limited to 2 Blackberry devices, it is a move in the right direction. 

My life would be so much easier if I could do email on the plane and I would be so much more productive too.

I just hope that they don't allow cell phone voice communications.  While the idea that your cellular signal could interfere with the plane's operation is ridiculous (truly), I like the fact that I don't have to listen to 200 people trying to talk over the hum of the engines at 30,000 feet.  I think I would shoot myself!!!

You can read the article here.

Tuesday, December 4, 2007

Responding to the Register Article

I have not been blogging a while, simply because I have been so busy with work and family commitments that this site has suffered.  But, given the news publish today by the Register in the UK, I thought it was time I used this blog to talk about EnterpriseDB in a little more detail.

By now, I am sure that you all have seen the article.  If not here it is.

So, let me respond.  As Andy stated we are in our best quarter ever.  That is true.  It is also true that we have already far surpassed last year's sales.  In fact, we are calculating our growth in multiples as opposed to percentages. In 2006 we did millions of dollars of sales.  2007 will be x times that number.  By whatever standard you measure sales growth, that is fantastic.

Next, we now have almost 200 customers in less than 2 years of sales!  Some of the names we have talked about. Sony Online Entertainment, Sun, Vonage and FTD are the most notable.  Others are less known but just as significant such as:  Tomax, FortiusOne, Ticketline, Agri Stats, Atomogy, Proxicom, and xPrima.

And there are some we have yet to announce including:

  1. One of the Top airlines in the US has built their entire loyalty program on EnterpriseDB Advanced Server
  2. A $20 billion dollar public sector solution provider contracted EnterpriseDB for a multitude of large projects
  3. A leading provider of IPTV that boasts over 32,000 subscribers and 300 channels
  4. EnterpriseDB replaced Oracle as the database for a leading provider of Pharmacy Benefits Management (PBM) software
  5. The world's leading provider of mammography software for the most of the world's mammography machines, swapped out SQL Server for EnterpriseDB.
  6. Leading GPS monitoring company replaced MySQL with Postgres and EnterpriseDB and now processes 200,000 transactions/day and volumes of reads
  7. The world's number one online broker uses EnterpriseDB to support their Postgres implementation
  8. A publicly-traded, billion dollar insurance and financial services company uses EnterpriseDB for their Postgres support
  9. The world's leading high performance networking provider uses EnterpriseDB Advanced Server
  10. A social networking provider with over 50 million worldwide members uses EnterpriseDB for their Postgres support

And that is just the customer news.  There are so many other things going on including:

  • EnterpriseDB beat Oracle 11g for Best Database Award at LinuxWorld.  We have won this award EVERY year since our inception.
  • EnterpriseDB Advanced Server now works with both Cognos and Business Objects.
  • Our latest beta release includes more Oracle compatibility including over 2 dozen additional system views, bulk array and bulk binding capabilities and support for Oracle hints
  • EnterpriseDB contributed the single most-complex, and significant change to PostgreSQL in almost a decade.
  • EnterpriseDB Advanced Server is now qualified on AIX and z/Linux from IBM
  • EnterpriseDB is now qualified on HP-UX from HP

And, since I am the marketing guy, I will give you some insight there as well.  As Curt Monash can tell you, we are seeing huge demand for EnterpriseDB.  Our lead generation activities are resulting in a pipeline that is better than we have ever had.  And the types of companies that are in our pipeline include major ISVs and Fortune 50 companies from just about every vertical.

In 2007, we also started a regular set of training classes for both EnterpriseDB Advanced Server and Postgres.  They are very popular and almost always filled to capacity.

So, I have yet to answer the question that everyone is asking after reading the Register article.  "Why, then, did you get rid of some sales people?"

Here is my take (and if I haven't stated it before, I DO NOT SPEAK FOR ANDY OR ENTERPRISEDB.  THESE ARE MY OWN THOUGHTS.)

As with every start-up, you start out with a business model that is based not on real customers, but on the promise of real customers.  After a year of business, you evaluate what you learned and how you attracted your customers and you adjust.  You do that every year and every company does it.  The difference is that it is not as apparent with big companies like IBM who have been in business for years. 

So, what did we learn?  We learned that we are starting to turn a corner.  In the beginning we had no track record and just about every single deal required an on-site visit to a prospect to make them comfortable with this "start-up".  Now, we have the customers and references to back our claims.  And, we have learned a few things, mainly:

  • We do not need to visit every customer
  • Most customers are buying a smaller license up-front and then coming back in 1-3 quarters to buy a much larger license.

This is significant and why we are focusing more on inside sales and less on the outside sales.  Companies are able to download and test EnterpriseDB without a sales person.  They get comfortable with the software and make a small purchase to "try it out" in a non mission critical environment.  They make these purchases via the web or over the phone.  Shortly after the realize how good Advanced Server really is, they come back and buy a lot more.  At that point, we may send a sales person to visit them.  Its not that we don't care about them until this point.  Its that they don't need to be bothered with a dog and pony show when they can just go get the software themselves and test it and our support organization for FREE.

Make sense?  It does to me.

So, from a marketing standpoint, I want to get more companies buying the smaller licenses and coming back to purchase a larger license faster and we need a bigger inside sales force to do that.  Then, when a sales person does make a visit, they are going to a very well qualified customer, not just a lead.  Their job is streamlined, as is the business model for EnterpriseDB.

And the customer is happy because they can call us when they need us and we aren't hounding them with direct sales guys always trying to schedule a meeting.


I know this was a long post, but I hope you read the whole thing because this is important to me and should be to you as well if you care about EnterpriseDB for your business.  Besides, I haven't posted for a while, so I am making up for all of it at once...

Laugh of the Day

I have always contended that Gartner's style of analysis is old-school and caters to the laggards, but recently it seems to me that when it comes to open source, most within Gartner just don't get it.

Kudos to Matt Asay who calls this out in his blog.

It is no secret that Gartner mostly caters to the large, multi-national companies. And, in many ways, these companies are so risk-adverse that they will not implement new technology until it has been a standard for some time. I also feel that, when it comes to open source, Gartner is afraid of it mainly because many of the analysts never tried to understand it.

Their view of open source is conflicted even within the organization. Folks like Mark Driver "get it". Others do not.

I was even told by one analyst that:

"Open source is like communism and you see how well that did in Russia."

And that was less than a year ago! (Of course, he said he would deny saying it.)

In fact, here is a recommendation from a Gartner analyst at last year's Open Source Summit:

Begin developing skill sets for Linux as a production platform, especially for DBMSs.

They have really gone out on a limb there. To recommend that a company BEGIN to develop skill sets for Linux in 2006 is preposterous.

Below is another gem from Gartner. And you wonder why organizations are gravitating towards analysts that are more in touch like RedMonk, The 451 Group and Forrester.

"Unless you’re hiring the best electronic engineers and computer science graduates every year, you’re not ready to use open source in mission-critical areas"

CIO Magazine - April 3, 2007