In the middle of May, just about 6 weeks ago, the SQL Server BI team launched PowerPivot. This weekend, we have we passed an important milestone: 100,000 downloads of the Excel add-in. For a BI application, this really is a huge number. I guess it truly underscores something we have said all along: that PowerPivot is aimed at a much wider market than traditional BI. (As Amir Netz says, those who can spell Excel, but can't spell BI.) I blogged about this very topic recently when I asked Who is PowerPivot For?
For me, it simply reminds me of one of the reasons I love working at Microsoft; we are the only company who can bring analytics of this power to this size of audience. It's not that other companies don't do great work; I love what many of our partners and rivals are able to accomplish. Yet only at Microsoft can we impact so many people. It's a huge privilege, and I know from my own experience that none of us take it lightly.Of course, it is early days yet. There is a lot more to come. Some of you may be wondering if these 100,000 users already represent the bulk of those who may be interested? I can tell you that the download rate is increasing, and increasing rapidly.
For now, having passed this landmark, I think the team deserves to feel quite satisfied with the early reaction. It promises to be an exciting year!
You've probably seen some press lately about how pervasive Microsoft is in lots of areas (here, here and here). We've also seen SQL Server show up in more and more enterprises - you can read lots of case studies here. There are lots of reasons that you're seeing SQL Server numbers grow - we're less expensive than other offerings, have a pretty good track record and a great toolset to monitor and manage the system. We've also got good support, a well-established repuation in large enterprises and all the way down to small businesses, but there's one factor that I'm seeing make a difference. It's our community.
I've used lots of software over the years (I'm really old) and I've seen lots of community efforts. Open-Source does a great job of keeping people in contact with each other, as do a few other vendors. But I see a LOT of community for SQL Server - people of all levels of skill, all around the world, in multiple industries helping each other out, freely and willingly. Thomas LaRock (SQL Server MVP) recently blogged about this phenomena as well.
I think community makes a difference - if we (Microsoft as a company) listen. And I think we do. I know the folks I deal with constantly monitor social networking sites, blogs, press reports and other sources for a mention of their product area. When they see someone struggling with a feature, they really take that to heart. No, we don't base every product decision on a sampling of people talking, but we sure do listen to what they are talking about.
If you're not connecting with the SQL Server community through social networks, blogs, user groups, SQL Saturdays, PASS or any number of other sources, you're missing out. There's lots of folks out there willing to help - and happy to do it.
I monitor a number of different database discussion forums and you can always find a healthy debate on the best database platform. The problem with this question is the answer is always, it depends. It depends on the business and technical requirements today and over the next few years. We just published a case study on a company called Arabah. Arabah is is the official U.S. distributor of the Jericho by Paloma line of beauty products made from Dead Sea minerals. Arabah started down the path with MySQL but as they began to compare the business and technical requirements to what MySQL offered they realized it wasn’t the correct choice. I’m sure you know how the story ends. The morale of this story isn’t that SQL Server 2008 is a better database platform than MySQL (even though it is), it’s that you must first define your requirements before you select your solution. Now does anyone know what’s the best car?
Ann All recently posted an interesting article at IT Business Edge: Microsoft’s Schizophrenic Approach to Business Intelligence. I’m not too insulted: I tend to agree with Janet Long that, “Part of being sane, is being a little bit crazy.”
Ann’s argument is straightforward enough. We have said that we hope to win fans in IT with PowerPivot, but Howard Dresner (and others) see PowerPivot as putting power in the hands of end users. Howard, and Ann’s, question is: “Who is your primary customer?”
Now I should say here, that I rarely see a dichotomy without wanting to prove it false, and this one is no exception. There, are to my mind, at least three fatal flaws in this argument:
· I think the term “end user” is too inexact for the scenario in question;
· I doubt the distinction between business and IT is as clear as the question implies;
· I don’t agree that a “primary” customer is necessary or desirable.
Let’s take them in turn.
Who is the end user of PowerPivot? The Excel jockey? That is hardly a complete picture. In fact, the majority of users will see the end result of PowerPivot in the form of interactive apps, dashboards and reports served up with SharePoint, and with SQL Server on the back end. These end users are consumers of analyses and they typically outnumber producers – the Excel users – by perhaps ten to one. These end users do not serve themselves: they may explore, and interact with the analysis, but others have built the solution, and others manage the infrastructure that serves them. So, “end users” form a complex ecosystem, which PowerPivot serves in carefully differentiated ways, with a powerful Excel add-in for producers, and a manageable, easy-to-use server-side story for the consumers.
So, if end-users are a complex bunch, what of this distinction between IT and business? Well that’s not so clear either. Many a department in many an enterprise – small or large – has a business user who, in effect, provides IT services for her peers. The SQL Server WorldWide User Group runs a splendid course for the Accidental DBA. Many small SharePoint installs, or Reporting servers, even departmental OLAP boxes are cared for by these untitled, and unsung, practitioners. They provision infrastructure for their peers, but their day job is in the business side. These are not the sole IT audience of PowerPivot, but they will be an important segment of the market.
Even with more “traditional” IT roles – with job titles to match – I find the artificial distinction between business and IT rankles somewhat. The implication is that IT has a set of concerns that are of little or no interest to business users, and possibly vice versa. Well, I must say that I rarely see the vice versa in action: even barely functional IT departments take great interest in the success of the business that pays their wages. Increasingly, business users, especially at higher levels, are aware of the business drivers of IT decisions. After all, companies can be fined, and company officers can even be imprisoned, for breaches of data and security. The integrity of a server may have as much bottom-line impact as the business transacted upon it.
Alors, revenons à nos moutons. The third flaw is the most serious. I simply doubt that a “primary” customer is needed – and indeed I reject the idea of either IT or business being “primary” in a decision to adopt software which affects both of them. To think that one must be primary is to assume that there cannot be an common interest between business and IT. It should not be so.
I expect at least two patterns of adoption for PowerPivot, no doubt there will be a greater variety than I can cover in this blog.
On the one hand, I see Excel power users discovering PowerPivot as a tool they can use to build more powerful analyses. They will adopt on the desktop but will soon need to share their work with consumers. As Ann said in another blog post: Happy Employees Are Collaborative Employees. At that time, they will push for server adoption from IT: either “real” IT or “accidental” departmental IT.
On the other hand, I can see IT taking a proactive role. Recognizing that PowerPivot provides a solution to the continuing demands of business users for ad-hoc analyses, IT can provision the infrastructure and advise business users to go forth and serve themselves.
Both scenarios are equally valid and I see no harm in us pursuing both of them. And of course, there is a third one. Business and IT together (they do talk, you know) decide PowerPivot is an appropriate solution and agree (it does happen, you know) to implement managed self-service BI.
Maybe I’m crazy after all, but I think it might just work.
I like it when *other* people review Microsoft software. I still volunteer as a DBA, so I still use multiple versions of our own products (and comptetitors, by the way) in the "real world". That's how I normally present, that's how I teach at the University of Washington, and how I work with my Microsoft clients. When I talk about the product, it's something I use - but of course I work here, so folks always like finding other people that have tried the new versions of our software and read what they say. Sometimes that's good, sometimes the reviewer has issues that they talk about. In any case, I like when they show detail about what they have tried and how it worked. And here is just such a review - it's by Jason Brooks over at e-Week.com - enjoy: http://www.eweek.com/c/a/IT-Infrastructure/SQL-Server-2008-R2-Offers-Enhancements-New-Management-Capabilities-518969/
I first met Sean McCown when I worked on the team developing SQL Server Integration Services. Sean was simultaneously one of our most active supporters and one of our most trenchant critics. He can very – what’s the word? – very forthright. You know, blunt, outspoken and generally on the money. So, I keep a look-out for Sean’s articles and certainly his reviews of our latest efforts.
So, frankly, I was at least relieved to see the headling of Sean’s latest article - Seven reasons to care about SQL Server 2008 R2 – and glad to read it. Sean calls out what I would agree are the important reasons to upgrade: Managed Self Service Business Intelligence, Report Components, Master Data Services, StreamInsight, Multi-Server management, DACPACs and Sysprep. But don’t just make do with the list: Sean’s comments are insightful and critical where he needs to be, and he calls out great use cases for the features he likes.
For an example of his criticism, Sean takes us to task for the ways in which we have factored Enterprise (now supporting 8 CPUs) and DataCenter. Time and customers will tell – especially new customers. Mark Beyer, of Gartner, had a different view of our changes: “It’s about time. The amount is reasonable, the per socket is still better than any percentage per core that other vendors use, and SQL 2008 R2 remains a solid value for the price.” SKU factoring is always a difficult balance, particularly where existing customers find themselves right on the edge of two editions. We’ll have customers in that situation and I trust our field teams will help them make a choice based on our need to build a long-term relationship, rather than overselling an edition for the box price. Traditionally, this is exactly how we have worked – thus the very high numbers of SQL Server Standard Edition that we see.
Enjoy Sean’s post here and please do add your comments. Would love to see what you think is compelling in SQL Server 2008 R2 too.
Donald Farmer
Well, I've almost recovered from TechEd this year. I had a fantastic time - learning, teaching and most of all, interacting. I'm really glad that Microsoft stuck with New Orleans - they've had a tough go of it since Katrina, but the city didn't show a trace. It was clean, positive and vibrant. Lil' Buck insisted on coming as well, so I brought him along.
I got out most every morning for a walk and to meet with the "SQL Cool Kids" at Cafe Dumonde, for coffee and French donuts. The first morning was with Donal Farmer and Brent Ozar.
I also visited with lots of the SQL Server community. While I did go to a few of the sessions, I have always thought that the most valuable part of an in-person conference was the in-person part - meeting with folks in and around my profession.
(Left to right: Denny Cherry, Andy Rowland, Sean and Jen McCown, Mark Stempski, Thomas LaRock, Steve Jones, Brad McGehee)
I also had an opportunity to visit several sessions this year that were far outside of the SQL Server arena. Knowing more about Windows, the cloud, development and SharePoint was my focus this year. (I saw the cloud - it's a big blue box that lives at TechEd and you can walk in it)
And of course I took liberal use of my evenings there in New Orleans to - well, be in New Orleans. My mother's side of the family comes from this area, so I was right at home.
So all in all, was it worth it? Oh yeah. I'll definitely be back.
I hope you’re able to join me next week at TechEd 2010 U.S. In New Orleans, Louisiana. I love taking this time each year to learn in an environment where I’m “out of my element”. What I mean by that is I have not only SQL Server knowledge all around me, but also knowledge for the operating system, Office Automation products, SharePoint, and one of my new/old favorites, High-Performance Computing. It's a time where I can focus for an entire week on "recharging" my knowledge batteries.
If you are coming (and I really hope you can), make sure you get the most out of the visit. Check out this link: http://blogs.msdn.com/b/cdndevs/archive/2010/05/31/a-few-tips-amp-tricks-on-attending-microsoft-teched-2010.aspx to learn a lot more about how you can prepare for a TechEd Conference.
We’re all facing budget pressures, so you have to figure out whether you should go to a conference right now. To help you make an informed decision, check out this link: http://northamerica.msteched.com/default.aspx?fbid=TJgYwH0BwHo
And if you are coming down south with me, stop by and say hello – I’m doing a presentation or two and also working at the booths, including our Surface Computing System demo area.