1Sean Cull (@seancull)  8/26/2014 1:21:43 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Bill,

thanks for putting the effort into writing this - I am sure that it took a lot of time and care and I hope that IBM appreciate it.

The complexity of the Sametime and Connections install has definitely put a lot of Business Partners off getting involved with those products - which was detrimental to everyone.

I know that OpenNTF has promoted some IBM pre-assembled on-premises connections environments that are available to developers. I guess this approach works OK for development but when you come to upgrading a live system with data and customisations it is more difficult. There again I guess there is not a lot of data held in Sametime ?

2Mick Moignard  8/26/2014 1:49:41 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Hear hear. This insane complexity that seems to be striking at everything is a large part of what is making me think its time to move on. Not just connections, but even "simple" things like modern programming are so much more complicated than they used to be, and I think far more complex than they need to be, either.

3Paul Withers  8/26/2014 2:01:37 AM  IBM’s Lost Art of Installation is Costing Them Revenue

I think this complexity is why many Domino customers (most?) have not taken up the entitlement to Connections Files and Profiles that has been available since 8.5. And the same for embedded experiences in 9.0.

I think it's also behind a lot of the concern about the backend of Mail Next - not that there may be non-Domino aspects to it, but that there may be the kind of complexity with Connections that means no one ends up installing it.

A great product is no good if people cannot install it easily.

4Henning Heinz  8/26/2014 3:19:54 AM  IBM’s Lost Art of Installation is Costing Them Revenue

I am always surprised that people care so much about IBM revenues.

For the full year, IBM's 2013 revenue was reported at $99.8 billion. The software division is extremely profitable (and has been for many years).Many of their problems currently are with hardware.

Sometimes I wished IBM would struggle a bit so that they would have to adjust their strategy but as IBM in most cases only reacts with cost cuttings am not even sure this would help.

As long as customers pay more for less this is not going to change and IBM is not the only vendor doing this.

Personally this is a problem because my customers want smarter software products but there must be some companies worldwide that still buy IBM and if it does not work they are buying their services too.

Business wise this seems to be a great concept and the revenue IBM loses with their cost cutting strategy is being adressed with a myriad of acquisitions.

Star Analytics, UrbanCode, SoftLayer, Trusteer, Daeja Image Systems, The Now Factory, Xtify, Fiberlink Communications, Aspera, Cloudant, Silverpop Systems, CrossIdeas, Lighthouse Security Group and those are acquisitions since 2013.

5Alexey Zimarev  8/26/2014 3:22:32 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Agree, absolutely. Whilst Connections is still manageable but Sametime footprint went out of control. New components like MCU and VideoManager demand insane resources from servers, cannot be virtualized and can only be installed on RHEL.

6Keith Brooks  8/26/2014 6:11:44 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Bill, Solid post. While I see your points, I see some of them in reverse, hear me out, I am not pro or con you or IBM.

Complexity=$$$ and IBM, and ourselves as BPs, are in this to make money. Yes, it scares off some people that installation costs more than the whole licensing, but such is life. I do not see this changing anytime soon.

An appliance would be nice and has been discussed by the many groups inside IBM ever since the infamous Lotus Foundation Server. It would help speed everything up and reduce the cost, perhaps. But it would be a massive server(s) and in truth small companies just don't want that expense or size to get what they can get for near free from a Cloud. If they want in house systems, then they have to pay up now. The days of cheap Domino shops, which I have advocated for years was the wrong way to sell/pitch your Domino infrastructures, is over.

You and I can install Domino in our sleep. Some of us can do the same with ST and a few even with Connections. If you recall what it was like pre-DOmino era, when R4 was really tinker toyed together, that is how I see Connections now. The difference was Notes had a huge footing that was growing. Connection is growing, but at a slower pace (my opinion, not IBMs) but has certainly picked up more steam since R5 of Connections came out.

R5 Domino is when Domino really took off, maybe R5 Connections is that catalyst as well.

Should it be easier or harder to install something today, over 10-15 years ago? Much harder I say. Look at all we now deal with at security levels, directory levels, international languages, various hardware platforms. I think it is a challenge to us all.

Patching is a nightmare and we do need a better "surveying app" to let us know which pieces need to be updated.

See you soon if not before Orlando.

7Carl Tyler  8/26/2014 6:26:12 AM  IBM’s Lost Art of Installation is Costing Them Revenue

I've been telling IBM this for years.

If you're a partner that does customer deployments this is initially good news, you'll be busy doing installs. But, the number of customers you'll have in the future will go down, as they will have moved to other easier to manage and maintain platforms.

I saw the following, customers spent so much getting the installation done they had no budget left to do integration projects with the installed software. So for example, Sametime is getting used purely for IM, not integrated into workflow processes or other applications, know what the means? It means it's easy to switch out with a different vendors solution when they want to, no tie, no value add, no increased ROI.

All this stuff is poor planning on IBMs part, it's all short term gain, that will impact them in the future.

8David Schaffer  8/26/2014 6:57:37 AM  IBM’s Lost Art of Installation is Costing Them Revenue

As a small user (30 seats) there was never a chance we would implement Connections and Sametime 9 in house. The resources required are too much even if we had a partner do the setup. The options are stick with Domino-based systems, go to the cloud, or go to a competitive product. We're still running Sametime 8 and Quickr while planning our next move.

9Bill Malchisky  8/26/2014 7:01:54 AM  IBM’s Lost Art of Installation is Costing Them Revenue

@1, 2, 3: Thanks, Sean, Mick, and Paul.

Yes a lot of time on this Sean; I appreciate your noticing the effort. Intersting take on other partners not getting involved too.

@3 - good point on slow adoption.

@4 - Not sure about others Henning, but I know revenue is what IBM understands and in the spirit of offering a professional perspective to affect a positive change, I utilized familiar terms. No successful company likes losing revenue

Thanks for commenting too.

10Bill Malchisky  8/26/2014 7:06:10 AM  IBM’s Lost Art of Installation is Costing Them Revenue

@6 - Keith...

Clients have fixed budgets for projects. If a significant percentage is consumed by just installing and patching the product before you can do anything, then the funds usually set aside for performance tuning, configuration, and training disappear. Installation should be a trivial event, comparatively speaking. The business partner seeking to extend, customize, or tune a quality enterprise product should spend the bulk of their time there, and let installation be simple and efficient. Calling technical support during installation increases costs for everyone involved.

There is a difference between short-term gains and the longer-term perspective with customers. Complexity can create an immediate cash influx, but also less work overall. I've seen this many times when on projects and managers gripe about other vendors.

Microsoft understands the easy installation model and well, they have more marketshare now.

11Bill Malchisky  8/26/2014 7:12:23 AM  IBM’s Lost Art of Installation is Costing Them Revenue

@7 - Thanks, Carl. So few people understand the constrast of tactical cash flow versus strategic gains. The customers for which I've been doing busiess for several years provide the most work.

It is very difficult to remain solvent on one-and-done opportunities, year-over-year.

@8 - and not everyone wants data in the cloud, particularl with the NSA issue on non-local data. I think an appliance akin to the Protector model could be a fourth permutation.

12Warren Elsmore  8/26/2014 7:13:34 AM  IBM’s Lost Art of Installation is Costing Them Revenue

I'm 'on the outside' now, so my comments amount to approximately nothing. but..... ;-)

I had a similarly bad experience with IBM's archiving solution - which was SO bad I've even forgotten the name. The install documentation for that one talked you through how to create the database schema. It was that terrible...

I had the thought then, and believe it even more now, that if I were the person responsible for procurement there is NO WAY that I would accept a cost for the software, then a separate cost for the installation. My business wants the end result, not a series of separate things happening to do something. If I were in charge, I'll pay you for something or nothing at all.

As an aside - In my micro business, it's why I've moved my mail to Office365. It just works and I just pay for the end product. No need to worry about the install time, hosting costs etc etc - EVEN THOUGH I can do that for free myself. The ROI doesn't stack up.

If customers don't want a cloud solution and are happy (dumb?) enough to pay IBM consulting the money to install it, then good for them and IBM I guess. But I've seen seven digit (GBP) pieces of work go ahead to install a single piece of software FOR PILOT and I know for a fact that the project subsequently tanked and the sale was never made. Too expensive to recover the install costs apparently...

13Mick Moignard  8/26/2014 8:08:50 AM  IBM’s Lost Art of Installation is Costing Them Revenue

@12: Warren, CommonStore? Could the installation at Unipart be the one you remember?

14Richard Moy  8/26/2014 8:25:58 AM  IBM’s Lost Art of Installation is Costing Them Revenue

As a business partner that focuses on the SMB market, this is a primary reason why IBM Connections and all its version will never penetrate this market and why they are losing market share in this area. You can say that IBM is only focused on the enterprise which is true. However, the SMB market is the core that support the enterprise. Sometime I even complain about installing Domino being too long. You have other platforms that are coming up that are simpler and easier to install and implement. These are the platforms that IBM will be competing with. They may not be fully enterprise ready but they are coming up fast and are being used by enterprise already. One example is Nodejs. To me as a Connections newbie, if it requires that much of a mess to install, than it is a reflection of how these components are integrated together. We tried recently to install Sametime 9 and gave up and we internally are no longer using Sametime. We recently were completing our Notes 9.01 training course and wanted to install Sametime 9 entry are part of our course. After trying many times we determined that we did not have the hardware nor patience to install it. We ask all our customer what version they were planning to use with Notes 9 and the answer was Sametime 8.52. They found that Sametime 9 was just too much trouble for them to invest into for now. And these are really big enterprise companies. So just like having a bad UI having a bad installation process will prevent companies from accepting and using your product.

15Darren Duke  8/26/2014 10:50:57 AM  IBM’s Lost Art of Installation is Costing Them Revenue

If Connections were a waterbed, there would be a whole lot of water, but not much area to sleep on.

At it's heart it seem relatively modest product suite offering over complicated by the appearance of a skunk-works like development and production process. As a really smart person once said to me "IBM makes software for IBM". Lest we all not forget that.

16Luis Benitez  8/26/2014 10:54:18 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Thanks for sharing. I've shared this internally with our team.

17Jeremiah Benjamin  8/26/2014 11:52:24 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Bill, great post and as a smaller customer we definitely have been hit with this issue increasingly over the past few years. Cloud would be nice, but I don't see it happening for us anytime soon (management decision).

For the benefit of anyone interested, here is a brief history:

We first deployed Sametime at version 8.5.1. The initial cost (consulting, licensing, hardware) to deploy was surprising, but acceptable, given that it was entirely brand new for us and we wanted to "do it right". It has worked very well for us, especially the webmeetings, so no regrets. But, as everyone knows, A/V didn't work well in 8.5.x. Thus when ST9 was launched with "new and improved A/V" I pitched for it. "We already own the licensing," I told approvers, "and we already have the hardware. All we need to do is pay someone to do the upgrade." Sounded like a no-brainer! But, alas. It's really a full rip-and-replace -- hardly the "I'll run the installer this weekend" process I use with Domino. Furthermore, the extra cost in consulting time for the "new and improved" A/V components was not approved. Recently we needed to give a series of video conferences, so we went with a competitor's solution which was very simple to setup. It lacks the level of integration, yes, but it works well enough.

As Carl's post above points out, we still haven't had anything left over (in time or $) for integrating Sametime with phone systems, custom applications, our web site, etc. So the overall feeling is we've bought a full-size farm tractor but all we do is mow the lawn with it. It COULD do more, yes, but we've already spent enough on it just getting it running.

Now we're looking at Connections and the sheer complexity of it really begs for an appliance option. Keith's argument that smaller companies would not want an appliance is, I think, true -- if IBM only shipped a "one size fits all" physical server. But I was thinking more in terms of a VM-based appliance. (We also use Protector, and one of the great things of that system is that it's pre-packaged as an appliance that runs on a VM. It's very simple to manage, upgrade and extend.)

We already happen to own licensing for Connections as well as all the hardware necessary for all the VMs, but the approval for consulting has stalled somewhat because of the sticker shock. This is not how it should be. It will become increasingly difficult to pitch these projects because the cost of implementation seems to increase with every upgrade because the complexity increases.

If IBM released these as the beta, I would be completely happy with that. Then IBM could put a couple months into making a nice installation process and/or coming up with the VM-based appliances before launching the product officially. If they do that, we'd still be paying the same money for consulting but, going back to Carl again, we would be integrating with other stuff in our environment and deriving more business value from it. That's when it becomes easy to approve new purchases again, and that's what I would love to see.

I think IBM's overall strategy is solid, the products are solid, but the "hidden" costs of implementation are not in line with other systems we use.

18Nathan T. Freeman  8/26/2014 2:16:30 PM  IBM’s Lost Art of Installation is Costing Them Revenue

Had to go back in the archives to find this one, from Feb 2007, I'll quote myself...

"The zombie horse takes another beating over at Ed's -- where a bunch of people joining in again to complain about IBM's marketing to the SMB sector. I have mixed feelings about how IBM handles their marketing in this regard, but I think it's much more important to think about how IBM handles their product strategy.

I've been pretty vocal before about how disappointed I am in IBM's new tendency to put relatively simple products on these uber-platforms, where the technical expertise required to deploy jumps by an order of magnitude. It's a personal pet-peeve because I've been so passionate about usability lately. But I've also come to believe that this is an important part of marketing to SMBs, where the single most important marketing tool available is AUDITIONING. Every business manager that I've ever known in organizations of < 1000 people has the same question about every new piece of technology: When can I try it out?

Here's the simple question to ask yourself: what does the CEO of a 200 person company have to do to try out, say, blogging? He can start something at Typepad or Blogger or MSN in 10 minutes, right? What if he wants to host it inside his own domain? What does he have to do to try it out on Wordpress? On Sharepoint? On Domino?

What about trying out email? With gmail, all you need is a credit card and 15 minutes to fill out a web page. Don't like the risks involved with that approach? What does it take to get an Exchange server up and running? What about a Domino server? What can you get results on sooner?

So what if he wants to try out instant messaging? What does it take to get the different options running? AIM vs. Trillian vs. MSN Messenger vs. Sametime? If even EVALUATING competing choices requires 100 times the effort on one versus the other, do you think that's going to be a decision criteria? It certainly is for me. And that's why I get bent out of shape when I see what the setup requirements are for IBM's new Sametime Gateway, or Activities server, or, I fear, Lotus Connections.

One of the saving graces of the Domino platform is that, although it can be somewhat intimidating to actually buy it, it's at least fairly easy to try it. It requires some determination, but you can figure out where to download it for free, and you can run the install on just about any box in your environment. And you can get visible results inside of an hour. That has allowed it to be an exploratory technology. It doesn't require heavy commitment.

And while I appreciate the engineering brilliance of something like the Activities server, the truth is that in an organization of < 1000 users, no one really needs the scaling capacity of a DB2 backend, do they? So when a product gets brought to market that requires a terabyte RDBMS behind it, then it's automatically off the radar screen in smaller organizations.

The transaction cost of working with IBM software is simply too high.

I want to say that again, really loudly: THE TRANSACTION COST OF WORKING WITH IBM SOFTWARE IS TOO HIGH.

And this is what I think a lot of people secretly mean when they complain about IBM's marketing to SMBs. There's awareness -- nobody in business doesn't know who IBM is, or fails to realize they offer an incredibly wide array of software to businesses of all types. But turning to Microsoft or Google frequently happens because it's just so easy to do. Even if there isn't already a relationship, it's going to be very easy to figure out how to get one. There is very little learning curve involved in setting up a gmail account or installing the latest version of Exchange.

That's one of the reasons the business segments of those companies profit from their consumer ventures -- the consumer relationships reduce the transaction costs of working with them on a business level. No, it's not the ONLY reason. Yes, there's a lot of emotion involved. Yes,there's some technology involved. But don't underestimate the importance of these transactions costs. When a company has to make an investment in IBM simply in order to find out what's in their product catalog, that's a big barrier. It's like driving 300 miles to shop for a car -- you'll do it if you need a fleet of 10. You aren't going to swing by there to take a test drive.

How can this get fixed? A big part of it is product strategy. IBM should decide that it hasn't delivered a solution to the SMB market until the solution is auditionable. And that means it has to be easy to find and easy to try. Imagine this: what if you took all the people that write install and setup routines and made them report to SMB marketing instead of product dev? How would that change priorities? How much more trial would your best products get? "

19Bill Malchisky  8/26/2014 2:37:58 PM  IBM’s Lost Art of Installation is Costing Them Revenue

@12 - Warren, as an outsider now, your comments are quite valuable to me and sage in that you chose what was best for our needs and simplicity won

@14 - Great insight and examples, Richard

@15 - Nice analogy, Darren

@16 - Thank you for reading, Luis. I appreciate your sharing this post.

@17 - Jeremiah -- excellent comment; awesome case study to which I would like IBM to take to heart.

@18 - Spot-on.

20Soumitra Limaye  8/27/2014 3:11:05 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Thanks Bill for the feedback. Has been shared internally.

21Lars Berntrop-Bos  8/27/2014 6:20:51 AM  IBM’s Lost Art of Installation is Costing Them Revenue

Hear hear.

22Bill Malchisky  8/27/2014 6:24:45 AM  Follow-up

@20 - Appreciate the interest, Soumitra

@21 - Thanks for the support, Lars.