Was thinking about this all day. Where does IBM go from here?

Clearly they have made the decision not to pursue hardware (or, 
realistically, hardware-based services). And their research division, 
while sporadically impressive, is mostly there to fuel patents-per-year 
quotas with associated bragging rights. I guess the future vision is a 
purely software company, a la Oracle, with adjoining software-install-
deployment-integration consultant revenues.

But IBM sells dozens of software products. Where best to focus?

The current thrust is CAMSS (Cloud, Analytics, Mobile, Social, Security). 
IBM themselves coined this term, and, indeed, numerous employees have 
been marked as 'deficient' in these skill-areas, with pay docked 10% 
while they attend mandatory sharpen-up-on-CAMSS training. If there is a
'bet for the future,' CAMSS is it.

=-=-=-=-=-=-=-=-=-=

Cloud -- the first offering (IBM SmartCloud Enterprise) was a failure, 
         and, most prominently, cost the corporation its Amazon-CIA bid. 
         IBM has since purchased SoftLayer as a second swing. In Jan 2014, 
         this cloud had no customers, but (Q3 earnings suggest) after 
         being re-branded 'BlueMix,' a modest sales-trickle is forming.

Mobile -- the company has provably floundered and procrastinated for 5+ 
          yrs in this area. Mobile computing even within IBM was (and 
          remains) a directionless mess; tech acquisitions like BigFix 
          and MaaS have gone nowhere. The IBM-Apple alliance might turn 
          this around (as IBM doesn't want to be a hardware company 
          anymore, they can now back whatever platform(s) they want), but 
          will this late-breaking news really matter much, relative to 
          x86 and Android?

Social -- no evidence of any progress in this area. IBM's approach to 
          social media has historically been Lotus SameTime (shudder) and, 
          more recently, a series of homebrewed "Jam" sessions harvesting
          social-media chats so as to extract common themes, buzzwords, 
          trends, etc. Unless Big Blue creates (or acquires) their own 
          social-media platform, no prospects here.

Security -- on paper, this seems like it should be good. IBM owns (has 
            acquired) a diverse array of info-security products, and 
            maintains a stable of pretty sharp consultants down Atlanta 
            way. Their pricing (way too expensive) and intra-brand 
            competition (one product team bashing/undercutting another) 
            is tangled and Byzantine. I would guess that this will reduce 
            to install + deployment help and "we'll-throw-in-2-staff-
            with-your-software-buy."

Analytics -- okay, this is truthfully an area where IBM does shine. 
             They boast numerous big-data collection tools (structured,
             unstructured, or otherwise) and quant-crunching methods. 
             At sufficient scale (big enough company, big enough data-set, 
             big enough time-span), every problem eventually becomes a 
            "big data analytics" problem.  The Watson AI is one celebrated 
             example. This (Analytics) area is doubly promising, because 
             IBM can sell both big-data software and big-data consulting 
             services.  In the "BlueMix" cloud, even.

=-=-=-=-=-=-=-=-=-=

So. What is that, two out of five? One-and-a-half out of five? Does this 
mean the corporation should shrink down to 30% or 40% of its current size? 
Or that it should somehow split its operations into two distinct sub-
companies? To me, it suggests a lingering mass of misspent investment (too 
much time/cash/headcount without results) not easily fixed by thousand-
staff reductions or delivery-outsourcing.

I do not envy Big Blue the road ahead.