John Del Pizzo from the IBM Sametime team responded to my post - but, I'm not persuaded by the counter-arguments.
He dinged us on:
"...the total cost of ownership is not clearly outlined, a reasonable TCO model would establish a timeline (say 3 years) and include the total costs, including various weighting factors for planning, operational support, integration costs, help desk, change management, and so on."
As I helped work up these estimates, I'll take the heat for this one. While our goal for the Lotusphere keynote was to provide concrete examples of how Sametime could pay for itself within a year, we did not set out to provide an exhaustive TCO model. We have to leave something for folks like Mike to do. :-) But... if you were at the keynote, you may have noticed that our composite Renovations, Inc example was followed by one from a German manufacturing firm. They provided data on their 3 year costs including things like education and support. Their return was 9 times costs. So, while Renovations, Inc was not a full TCO, I'd argue the composite return of 3.5 times costs was actually fairly conservative. Of course, your milage may vary and you should examine the TCO in light of your specific organization.
Round2: A fair point. The more specific example is appreciated but that doesn't really excuse IBM from portraying an incomplete / inaccurate argument. I would have preferred to eliminate the simplistic example then and stick to a testimonial that reflected a real-life situation. I think real scenarios always trump fabricated illustrations. Saving the time on the over-simplistic demo would have allotted more time to dive more deeply into an actual use-case.
On replacing the licensing fees of a hosted web meeting service with Sametime:
I dislike this example because IBM fails to accurately portray the total cost of ownership for an on-premise implementation of Sametime vs. a SaaS solution.
To an extent, this is the same issue as above. Yes, there are costs associated with running your own infrastructure versus outsourcing it. However, the point stands that you can pay for the software licenses, hardware and a full-time administrator with what you would save by eliminating 1000 Webex Meet Me licenses - in year one. And since you don't have to "rebuy" Sametime every year - maintenance fees are only 20% of the purchase price - the savings go up dramatically in the out years. AND you get a full unified communications platform with an extensive range of capabilities above and beyond only online meetings. I'm not saying that the hosted model isn't attractive for some - just that there are scenarios where it makes more sense to own the infrastructure.
Round 2: The argument is still incomplete though - are all situations handled with only one administrator - how are complex ST implementations that require placing components of different servers handled - when is more than one admin needed and what about the other support people (help desk, system engineering, networking, server support, change management, etc? Clinging to the notion that all you have is license cost and one administrator is very simplistic. You also have costs incurred when major releases need to be evaluated and an infrastructure upgrade scheduled. I realize the sensitivity concerning WebEx since a good number of ST shops use ST for IM but other vendors for web conferencing - but we need to be a bit more accurate here.
On using Sametime to IM enable a support site to reduce the number of calls.
This example is a stretch for me - a big stretch - most customer-facing applications on web sites that are handled by call centers do not use off-the-shelf generic IM products. They typically use real-time collaboration tools that are included within their call center suites and/or CRM applications. The reason is that call centers do a lot of multi-channel management so you do not want someone involved in an IM chat to appear available to handle a phone call or email response. You need to integrate on-hook/off-hook signaling, integrate with other customer response management tools (telephone, web, e-mail, IM, etc). IBM is correct that organizations should be looking at these tools - but incorrect that they should look at IBM and Sametime
Sorry, Mike but I have to completely disagree with you here. We specifically included this scenario because we have customers who use Sametime for exactly this purpose. How long have we been talking about the benefits seen by companies like Celina Insurance? Or a large financial services firm that is using Sametime on their loan website? In the latter case, they integrated Sametime with a partner application to route the IM to the next available agent and are evolving the implementation to connect to their telephony system to show onhook/offhook status. This way IM's wont be routed to someone already on the phone. I probably get this question once a week from our sales teams... so clearly companies are thinking about extending their web sites and call center applications with Sametime.
Round2: Sorry - you made my point. You need to add in the cost of the partner application or the customized integration into the e-Service/CRM suite that the customer is using if you want to be accurate. Implementing ST as a silo, without the integration into the multi-channel interaction management environment of large enterprise call centers is a huge oversight in the IBM argument. You need to include those costs before you can address any potential savings. And if you include those costs - how does that compare with the native real-time capabilities found in leading call center applications that address e-Service. IF those costs had been included - I'd be fine. Presenting them in a way that dodges those issues is disappointing in many ways.
The "time saved" argument is one that has been used over and over again to justify technologies ... if you save someone 10 minutes it does not mean they will take that 10 minutes and apply it to real work. The more structured and directed the activity is (e.g., call center) the better that argument resonates. But the more discretionary the work, there a greater chance that the person will just chat with a co-worker, go grab a coffee, etc.
This is the one Mike brought up twice, so I'll address it last. (And I really got a kick out of his example of what people would do with all their new-found free time.) But... we didn't go near the productivity gain potential in the Renovations, Inc composite example. While nearly every customer I've ever spoken to has told me that the productivity gains are the reason Sametime is such an invaluable tool, the gain is usually difficult to quantify ahead of time. So for Renovations, Inc, we stuck to what could be quantified with a reasonable degree of certainty before an implementation. I'm not sure why we got two thumbs down for an argument we didn't use.
Beyond that minor detail, I'm afraid this line of reasoning wouldn't hold water for me anyway. Perhaps we should go back to using typewriters because word processors have led to too much golf? Seriously, though, IBM has documented that it closes its books 3 days faster every quarter because of Sametime. While they maybe hard to quantify beforehand, productivity gains should not be dismissed when it comes to Unified Communcations & Collaboration software.
Round2: I will match your argument with 10+ years of being in the analyst business and talking to multiple hundreds of clients that consistently point out that "time saved" is over-rated because it is often very difficult to prove real savings. Yes, you can identify time freed up - vendors are right to make that claim - they fall down on assumptions regarding how that time might be re-applied. Too many variables in some situations. For instance, in those situations where work/workers are transactional or semi-transactional (structured task workers), then you can indeed get pretty close to converting "time saved" into real efficiencies because those activities will have metrics to substantiate it. If that is the case - then you can make that point and be well-received. But once you get to what people refer to as "knowledge work" - (e.g., people re salaried, are paid regardless of time saved; or where the work/tasks have a lot of flexibility in how they get done and over what time period; or where work is not really output-driven - then the time saved argument is pretty weak. It can get a project approved - that happens a lot - but when you go back to look for the savings - it's like chasing vapor. You can be somewhat flippant in your response, but it just shows a very weak understanding on the part of IBM regarding the challenge of measuring productivity. You can go back to many industry efforts that have tried - economists and folks in the educational arena. It's a tough intangible to measure. Fortunately, IBM is not alone - Microsoft made the same grandiose claim around its OCS R2 launch. It's a very common approach - and vendors should be called on it.