Wednesday, March 28, 2012

Fear, uncertainty, and cloud

I recently read an article entitled CIOs fear business leaders see cloud as way to circumvent IT (h/t to Jon Reed who h/t'd Vijay Vijayasankar) and I was floored.

What Sooraj wrote was:
The study says CIOs are concerned that cloud provides business teams with a way around IT teams by acquiring cloud services on their own, which undermines the strategic partnership they are trying to build with business leaders.
But what I read was:
With lines of business increased control over IT spending and IT's complete inability to meet the needs of the modern business in a timely and pleasant fashion, here's just one more reason why I fear not only for the office of the CIO generally but for my job specifically.
I find that more than a little disheartening -- I've long said that working like you fear for your job is the best way to do a terrible job. That said, everyone's got kids to feed, so how does IT survive this "whole cloud thing" and leave the business better off to boot?

  1. Manage these cloud connections internally. Dominic Wellington recommended this at the end of the article, and rightly so. Lines of business don't want to manage their cloud services or anything else technology-related, but often they must in order to get things done. Managing cloud services needs to have some policy and procedures around it certainly, but taking 3 months to get around to filling out a web form that takes 5 minutes isn't going to cut it.
  2. Make everything simpler. Whether it's Steve Jobs talking about saying "no" to lots of things, or Vishal Sikka talking about removing layers, it isn't hard to find smart people willing to talk about taking complexity out of your business as a good thing (when done thoughtfully, anyway). Doing extra things can be dumb. Tactically employing "the cloud" can help you do less dumb things. 
  3. Take this opportunity to restructure IT intelligently. An investment in the cloud doesn't have to just transform the way your company does business, it can also transform the way you deliver services to your business. As a company moves to the cloud, it probably needs less pure-developers for day-to-day operations and should invest in skills such as security and configuration. This will also allow your best developers to move onto more interesting and fulfilling opportunities, like mobile development, and your less good ones to do something... less developer-y.
The cloud has put CIOs on notice, as well it should have, but being put on notice and being doomed aren't necessarily the same thing. Like with most things, proactivity is the key. Embracing the cloud now may prevent it from putting you in a choke-hold later.

Monday, March 19, 2012

Is SAP Still Missing the Mark on Mobile?

We are at a point in enterprise mobility where everyone has assumed it is a given but no one has exactly figured it out. Is SAP shooting itself in the foot by trying to have a big fixed cost (SUP) and high variable costs (uber-expensive apps), meaning that they'll have to talk people into surmounting expensive barriers to entry on two separate fronts. 

Consumerism and Platforms
Most people think that the "consumer experience" is about things working seamlessly and being super intuitive. That's half the story (and a big half, to be sure), but the other half (and the easier half, to be sure) is choice. We've all, I'm sure, read the Google Platforms rant. The reason Platforms are great is because if you can own the data and the infrastructure (which are the most valuable bits -- sort of like the bacon of IT) and you let others extend it and propogate it, you win without doing a whole lot of the work. You want to own the platform.

Let's use mobile Twitter as an example. Sure, you want to sell directly (the mobile Twitter website) and you'd like to have a brick and mortar store of your own (the official Twitter iOS app) and you might even buy a distributor (Tweetdeck), but in the end you really mostly just want people to be writing and reading tweets. That's why you build out an API and let all sorts of other people sell your product for you. And other people selling your product is a good thing. 

That gives people choice, and people like choice (I hate Tweetdeck, but I love Hootsuite, which enables me to continue to use Twitter's core product without using it's distribution). It also generates competition, which makes your product even more valuable in the long run. Enterprise software vendors need to remember that what they really want to own is the tweets data within their systems of record.

Developers Wanted
SAP is getting closer to having a fleshed-out mobile developer ecosystem, but at present it isn't open enough to generate a lot of innovation or competition -- the table stakes are just too high. SAP needs to make sure the price is low enough to engage enough developers to give users choice. Need a mobile solution to approve work orders? It sure would be nice to have a couple of different apps that are inexpensive enough to try out. 

In order to get those apps you need a robust developer ecosystem. In order to get that robust developer ecosystem, you need to leave enough money on the table for those developers, and you need to make it really easy for them to develop software (check out the 4:05 mark of JD-OD's video).

The Back of the Napkin
Per SAP CIO Oliver Bussmann, SAP manages some 12,500 iPads and those iPads have downloaded around 120,000 apps from their Afaria appstore (that's around 10 apps/user). For SAP. Customers, each of those apps will cost somewhere between 25 and 100 euros, so we'll say that these each average $50/app (yes, I did just switch currencies - it's cool, though, since SAP can handle that). That means that for each user, which might download 10 enteprise apps, we've effectively doubled the price of a $500 iPad (presumably people important enough to warrant more memory or 3G connections would warrant more/better apps as well) and that is before the price of SUP, Netweaver Gateway, and Afaria are taken into account. 

To justify that sort of expense for the average mobile-enabled user, you have to basically disable their desktop experience to save on that cost. I for one don't know many people that are willing or able to totally leave their desktop or laptop behind yet. The easy solution? Charge less for apps. The problem is, that's the only place third-party developers can currently make money (and they have to make enough to cover their own infrastructure first).

So Where Are We?
The solution -- and one, I might add, that'll be hard to swallow -- is that SAP needs to give up on making money at every possible point in the "enterprise data to mobile device" supply chain.
  1. Give developers access to SUP, Netweaver Gateway, and Afaria for free to learn the technology. This will allow certain shops to participate in the market that never would have otherwise. You'll end up with some crap applications, but you'll also unearth some gems. The free market works in an app store.
  2. Charge customers for the mobile infrastructure one-time. Let them connect to it via whatever means they want. This should not be a tremendously high price. How much extra do Workday or Salesforce customers pay for their mobile access?
  3. Do what you can to keep App prices down. This will largely be up to the developer, but work with them to encourage "creative" pricing. 
SAP-broadly should be trying to learn from what BusinessObjects is doing in the mobile space. It has had some struggles, but it is poised to move enterprise business intelligence to the mobile device because it has figured out the cost (an enterprise mobile service which isn't cheap but also doesn't seem to be a tremendously huge obstacle) combined with a couple of solid if not perfect free applications. They also have a developer ecosystem that is willing to play because they have low barriers to entry. This means you only have to pay once for the platform and then can distribute it however you see fit. This is the sort of model consumers have come to expect and that enterprises should flock to. 

I understand SAP wants to make money everywhere in the process, but they have to remember that their back-end systems still being relevant in 10 years will be largely dependent on enterprises being able to access and manipulate that data from anywhere. Making it really expensive and unattractive to do that right now isn't going to help on that front.

Monday, March 12, 2012

Is it OK to get excited about Explorer yet?

I can't remember the first time I saw Explorer, but it seems like a really long time ago. It's been a part of every major SAP or BusinessObjects keynote that I can remember, shown as the delivery method for high-value information to people who don't otherwise know how to get at information, and it was really BusinessObjects first foray into Mobile BI (if, that is, you're able to ignore the Blackberry version of Mobi. Based on what I've heard, most of you were able to ignore  the Blackberry version of Mobi).

The thing is, besides coming out with an iOS interface, Explorer hadn't really changed much, at least in terms of appearance, since it first came out of the labs and was called "Polestar" (a relatively exhaustive history is available here). Why is this? The commonly held assumption is that no one bought the damned thing*, making more investment into it seem dubious. Based on some of the things happening around Explorer these days, the investment faucet has opened back up.

Which is good, as I'm taking a look at it for my employer and have run into several limitations, most of which have been resolved in the most current release (which of course I'm not on) or in the upcoming FP3 (which I am also obviously not on). Per SAP (ASUG login required), the following have all (or will be directly) added to the tool.
  • Second Dimension - Now you can chart by product by region. That's a big deal. 
  • Auditing - If I can't prove people are using it, it's tough to justify spending more on it. Based on some feedback I've received this last week, I'm not entirely sure that everyone who has the latest version is aware of this feature.
  • OpenDoc - Or more accurately "OpenDoc-like functionality." I desperately want to be able to link from a dashboard spot to an Information Space with several facets already selected. I clearly do not want to bookmark and record the URL for every one of the possible combinations of selected facets. 
  • Calculated Measures - These had historically only existed at the session level and are now (or will be soon) available as part of the Information Space properties. This really broadens the value of the Explorer Tool since it allows for (simple) measures to be calculated during runtime at higher levels (such as revenue/store) where storing them at the lowest grain possible (individual transactions) would lead to inaccurate aggregations.
  • Offline Capability - In the current App Store version you can bookmark a particular view and keep it around (and even manually resync it when you are connected), but everyone knows that isn't really the ideal end game here because you lose the ability to analyze it. Based on some conversations I've had, I'm not sure that SAP intends to extend the functionality far beyond that.
  • Time Trending - This allows data stored as "Wednesday" and "Mittwoch" (Wednesday in German) to be grouped automatically in facets and sorted properly in charts. 
  • Geospatial/regionalizing/maps - For years I've rolled my eyes when salesmen would talk about this sort of functionality in BI tools, but after seeing some of the things Centigon has done I'm fairly certain geolocation has a use case (albeit a limited one) in the field. I think this is more a play for the future than the present, but it will be interesting to see it shake out.

Obviously the big thing about Explorer in FP3 is the release of Exploration Views (which is NOT supposed to be an additional license beyond Explorer). This essentially turns Explorer into a data source for sexier and more customizable visualizations, which is a good thing. An even better thing is that this application has been shown around for a long time, so it should be pretty close to fully baked when FP3 first comes out.

Why should we care about Explorer so much? Since some customers showed interest in it (and voted with their wallets) the tool has and is continuing to become a lot more enterprise class. I think in the not too distant future we'll see companies drive their "self-service" analytics needs through a tool like Explorer and their hot and heavy ad hoc users through Webi (although these folks could probably just as easily use Crystal and skip Webi altogether).

The important thing is that Explorer is finally getting beefed up, so if you haven't invested in it before, it's probably about time.

* Number of downloads is often a metric used to prove adoption of a certain application (or more often an app) but I find it flawed. Everyone and their brother/sister have downloaded the SAP Explorer app from the app store, but until they connect it to their own environment's repository, I couldn't care less. Hopefully most of SAP's BI customers aren't basing their decisions on 20 year old World Cup results or Formula One metrics.

Monday, March 5, 2012

BusinessObjects, the Shelfware Killer

Last week I attended the BI2012 conference (recap podcast here) and was fortunate enough to present two sessions, each of which were centered around the different data connectivity options that a user has when connecting to data that doesn't originate in SAP applications. The conference organizers, cognizant that "SAP BI" now includes a whole lot of customers who don't run anything else from SAP (and may, eventually, include a whole lot of people who don't run BusinessObjects either), sought out speakers who could talk about the BOBJ toolset from a more database-agnostic perspective.

Not wanting to waste anyone's time, I warned people upfront that if they were there to find out how to connect to BEx, BICS, BW, BWA or any other SAP-centric data source/connection they may want to move on. I lost quite a few people every time I mentioned this, which was OK, but I can't help think that a lot of SAP customers are missing out on some big opportunities to turn off some other BI solutions that they have in house.

In my first session, I asked how many folks were already running BOBJ. Almost all hands went up. I then asked how many also had, somewhere in their company, another BI solution to support another applications. About half of the hands went up (and I'd wager those that didn't just don't know about what shadow IT has put out there).

In case you haven't heard, the economy is pretty tight these days, so why are companies still paying maintenance on two BI tools when BusinessObjects can almost certainly handle all of their reporting needs? A few possibilities exist.
  • Because they don't know they can. A lot of SAP customers probably just bought BOBJ because they were told to, and once in, the account rep had no incentive to encourage them off of their existing tools even though the BusinessObjects family can report off of practically anything
  • Because it would cost more. Obviously this will depend on what your licensing looks like, but it is probably worth a look. I haven't seen a formal study, but I'd wager that if you took into account the key cost drivers in a switch from BI tool X to BusinessObjects (retraining people, porting content in, eliminated maintenance and support) you'd come out at least even if not better off, with tons of future savings in only having one toolset to support. 
  • Because change isn't easy. Even if an organization knows what BOBJ can do, and that it will save them money, sometimes it really isn't organizationally worth the hassle. It's "because we've always done it that way" at it's worst.

So what (legitimate) reasons do people have for having an enterprise reporting solution that doesn't cover their enterprise? Am I crazy?