Friday, July 9, 2010

All fans of Paul the Octopus need probability lessons

This one is for sooooo many updates on my facebook and twitter page

For the unaware, Paul the Octopus is a common octopus, used as an oracle to predict the results of German international football matches. Paul is presented with two boxes, each containing food, and each marked with the flag of a national team. He is reported to have correctly chosen the box with the flag of the winning team for all the German football matches.

Paul the Octopus rather the ‘Oracle’ who has an uncanny ability to get the results of all the fifa world cup matches right. Blessed with foresights, he unravels the mystery of future for football fans. But wait, should we be impressed. I don’t think so.

So there are just 3 results possible i.e win, loss or a draw. And 3rd match onwards it comes down to just 2 possible results. My basic knowledge of probability tell me:

Possible permutations and combinations for 6 matches: 3*3*2*2*2*2=144

And to have an impeccable record in world cup (7 matches): 3*3*2*2*2*2*2=288

So to interpret it 1 in 144 octopus/coin/bird etc. would get the first 6 matches right while it improves to 1 in 288 chance for getting it right for all 7 matches.
1 in 144, yes that’s pretty much it. So basically one of 144 octopus can replicate his record (Gosh but we don’t see so many more celebrity octopus)

Still believe in the ‘oracle’, well let me remind you that our ‘oracle’ has got 2 matches wrong in Euro (source: his wiki page). And common sense tells me, getting it wrong even once proves that he has no powers (if I could see the future, I will never ever be wrong unless I am betting which I don’t think Paul ever did). Well being an erstwhile Argentina and now a Spain supporter, I am happy with his predictions but I am not celebrating yet. Neither should you.

Who is the winner: Obviously the media who got some TRP and news (we live in a word of 24 hrs news channel so something needs to happen) + the keeper who made a mountain out of a mole hill and will be laughing his way to bank.

Monday, February 15, 2010

Commoditizaton of IT

Technology and Commodity used to be miles apart a while back. But the changing landscape might alter the dynamics of the industry. IT landscape is changing. But then we have simultaneous trend of Cloud, CI amongst others. What does this lead to.

Hardware: Moore’s law has hit hardware the hardest. Almost all hardware is now commoditized. With compute capacity increasing manifold we are seeing declining hardware prices already. Who will not buy ISS servers with BCS traits or some good and dependable cheap storage. The free fall in terms of $/GB is huge (yes I know we are living in age of information boom, all we need is to develop good software to dedupe and save the relevant information. A lot is just noise). Further developments in Virtualization and related Cloud technologies will improve IT utilization rates and Moore’s law will ensure more cheaper compute capacity and storage. What about Level 3-7 of network infrastructure (do you know ProCurve’s ambitions) or Mainframes (IBM is already talking about moving away from hardware, is that a sign of acceptance of end of mainframe dominance) etc. But hardware is just one part of the stack

Software: Open Source, SaaS and Cloud. Can they lead to more commoditization? The answer is yes and no. With growing complexity and reducing hardware margins, the incentive to support open source might not be there (it’s not been long since one of the Open Source poster child Sun was set) but then we already and have a slew of companies with great products in open source. The areas which are fairly developed like CRM for SMB’s (maybe the wave for Social Media CRM might delay it) will and are surely being commoditized while other where complexity is increasing will be the new margin leaders ex. Large Enterprise ERP or Analytics. I don’t see Cloud as making it more commoditized. On the contrary developing great technology say in sixth sense technology or Cloud it might lead to high margin and high growth for a company.

Services: All three segments need separate mention:

Support Services: I anticipate a trend toward commoditization is non Critical and SMB datacenters for support services. But for enterprises with Global scale and Mission Critical requirements, the day is still very far. Multivendor and MCS capabilities for large complex IT environments is not an area where every Tom, Dellick and Harry can play. I expect to only see HP and IBM leading in it.

Outsourcing: Anticipate it to be cloudy. We can anticipate commoditization in Outsourcing as we can expect outsourcing to compete directly with Cloud delivery model. And any premium will have to be justified. (Did I hear public cloud concerns, oh!! wait for some time and concerns like security etc will passé)

Consulting and Integration: Can consulting be commoditized. I anticipate largely yes but a fairly high growth. With growing complexity of IT (Yes Cloud esp. Private means more complexity) and to exploit the true potential of the developments in IT industry. Companies could use more and more consulting services. But will these be non-commoditized. For a fairly high number of companies the answer is no (ok so you are bank with 200 branches, we can sell you the same advice we sold to so many of your peers with some customization) and some smaller customers might not even pay for it if they are committing to other purchases from the vendor (So you are a Software development Startup, this PaaS is what you need). Almost every deal will involve some Consulting or as the management jargon goes “Consultative Selling’. While large run rate opportunities will still be Non Commoditized (So you are a fortune 100 company want some breakthrough impact, HP and select vendors can do it but not everyone will be able to help you)

Overall, a fairly large portfolio is moving toward commoditization. But how can we forget we operate in technology sector, some new garage ideas might make it not at all commoditized again or Completely Commoditized.

PS: Graphic Representation to come soon

Tuesday, February 2, 2010

The problem with Cloud user survey



Cloud is at the zenith of hype cycle or the in vogue word. It has also given a lot of fodder to Syndicated research firms, news wires etc etc of the world and we end up seeing a lot of reports, articles, news on the same.. Almost everyone has an opinion and these esteemed organizations want to capture the same. So this all sounds great but for…

The problem begins and ends with the word. Each and everyone has a different definition for Cloud. So here I am seeing a user survey report saying X% of SME have moved into cloud (what cloud??) and then here we have Z% (much higher) many users moved into SaaS. Now when did SaaS not become the part of Cloud (maybe rest of them have this lower multitenant etc etc version of SaaS which as per some is not cloud). How many times we come across X% of users are interested /evaluating the Cloud. Y% have deployed cloud and then completely different numbers when we look at IaaS, PaaS and SaaS results plus if you do some simple maths. And then someone says Google is Cloud (How many don’t use it). Almost all personal email is cloud (as per few thought leaders :) ) and online storage. Gosh!!! its such a large list. Did any one care to explain it to the respondents? What do you mean when you ask someone questions on Cloud. Do you or your respondent really care (Isn’t everyone on Planet Earth smart enough to figure outthese trivial things!!)

End result, we see a lot of conflicting views, more confusion and a lot of Gibberish. I complain because it makes my job tough!!!

Free advice: When doing surveys on Cloud Computing please avoid using the term and focus on terms that are clear(categories may be) and if you still fancy the term, please conduct a training for your respondents and maybe a test to check if they got your version of definition right

Or is it 99% of all statistics are just made up

Saturday, August 1, 2009

Predicting Cisco’s Strategy

A lot of discussion has been happening around Cisco should acquire Dell also on why Cisco could have acquired Sun…… . I am wondering wouldn’t Cisco have already done its research on possible permutations and combinations (esp. when it has 40 billion in cash). More so many companies are looking for suitors for a while now (Sun searched for a long while as well) and Cisco’s decision of entering into blades is also well planned. So is their Cloud Vision.

This makes me wonder if Cisco has something else in mind. With UCS, they are targeting a market which is at present niche and they must be expecting it to grow by leap and bounds (Almost everyone is expecting Cloud to grow). Think they also believe that installed base doesn’t give them a transformation edge. Additionally very few big companies have got the right set of mature technologies to add transformational value to their Cloud vision. Almost everyone is new to Cloud and no one yet knows what will eventually lead to nirvana for cloud.

Further the way IT market is coming off age, its turning that behemoths like HP and IBM have added quite a lot of breadth and depth to their portfolio. Everything said and done they are the strongest contenders to own the datacenter of the future.

At present the behemoths are exerting influence and potentially weakening the niche players (Cisco including) with their might (promoting proprietary products + endeavors of making them best in class+ OEM partnerships). The road ahead appears blocked till niche players come together combining their strengths and creating a formidable competitor (they will HP and IBM to join but will make efforts to avoid them). Cisco with its leading networking and strong balance sheet might think itself as being perfect for exploiting this partnership. This option of having multiple permutations and combinations gives customers (with Cisco in center) ample opportunity to choose the best fit and having it under one umbrella with can be a potential inflection point. The biggest challenge will be coordination and running the show which Cisco hopes it will be able to do.

So a major big acquisition will affect Cisco’s plans as firstly say for ex. Dell (or even Sun if we go back to last quarter big acquisition) is losing market share/ or are weakened that’s why they want to sell. A major turnaround story will definitely be a drain on cash on one hand and what it will make Cisco is more like HP and IBM but comparatively weaker. Cisco thinking that it is smarter will like to avoid the scenario. Not acquiring a big player gives it perfect opportunity to partner companies which can add value and Cisco can just pick up perfect fit.

So what will it do with all this cash? Well there are definitely some companies that are integral to Cisco’s vision at present. Top of mind being Vmware/EMC, Accenture and to an extent BMC etc. What if IBM or HP enters into an agreement to acquire them? That will be an ultimate offensive move from the behemoths against Cisco but wait Cisco is ready. Cisco will battle them out. Pay crazy premiums that IBM or HP can’t. Shareholders hate leverage and everyone knows that. To take an example say HP enters into an agreement to acquire EMC and its crown jewel Vmware. Not that impossible to do but what if it has to face Cisco in the battle. HP’s strong balance sheets doesn’t have the degree of leverage Cisco has giving Cisco an edge. So eventually Cisco can choose how important is EMC/ Vmware than risk losing its important partners

To add to this, the feeling of strong partnership with a primarily Very Strong non-competitor against the biggest competitors (HP and IBM) will make these players more sure and less likely to sell (read will ask for higher valuations)

Most importantly, this will dissuade IBM / HP to make the ultimate offensive moves. Thus again keeping the market open for Cisco to partner and choose the best available options.

So what will Cisco do. I think they will acquire smaller companies with niche technologies important to their vision. I don’t see them making first moves for an Accenture or an EMC in near future. Additionally I can see them build proprietary capabilities for Cloud. Most important, form very very Strong relationships and partnerships.

More importantly prepare for the battle which has just begun

Monday, April 20, 2009

Oracle+Sun!! Exactly what doctor ordered

Finally Sun finds a buyer, and must say its what the doctor ordered. Industry will definitely get interesting post the buy as Oracle has the money and buying Sun will make it an integrated player. Applications to Java to Storage to to servers and yes the Cloud as well. Not sure how much Larry loves the cloud though(Pun intended)
Sun was/is definitely a gem of company but its balance sheet couldn’t stand the impact of crisis and they were forces to downsize. Rock got delayed and there were serious questions on its survival making customers jittery in investing in their products. Add to it the IBM acquisition news and they were left with no option. On the other hand Oracle with its stable annuity business is a powerhouse in a crisis. And their ability to integrate acquisition is also incredible.
The challenge for Oracle will definitely to turn Sun around. Also Oracle doesn’t have the reputation in Sun’s markets which many customers might question. Also Oracle doesn’t have the services arms which Sun definitely requires for a easy turnaround. But what Oracle does have is money to turn things around. It can and should invest heavily in Sun.
The success will largely depend on how quickly the acquisition is integrated. Post the integration, there are opportunities galore but till then IBM and HP will be happy acquiring Sun’s customer and building even stronger relationship with SAP and others. Not to forget preparing for the battle.
The nirvana for Oracle will be how quickly they can churn out more ready machines like exadata but this won’t slow down other players. It does provide Oracle a lead but Cloud will have its own implications on the market.
Overall the datacenter customers will like the acquisition but so will Accenture. Can see it becoming a hottest acquisition target (Will blog on that as well)

Wednesday, March 25, 2009

Cisco not yet in race to buy Sun, does that imply something

A lot of discussion has been happening around why Cisco not IBM should be acquiring Sun...... . I am wondering wouldn't Cisco have already done its research on possible permutations and combinations(esp. when it has 30 billion in cash). More so Sun has been looking for suitors for a while now and Cisco's decision of entering into blades is also well planned. But all the arguments of accessing installed base, technology, storage, java etc all make sense.

This makes me wonder if Cisco has something else in mind. With UCS, they are targeting a market which is at present niche and they must be expecting it to grow by leap and bounds and they believe that Sun's installed base and most of its technology will not add transformationak value in the space. Almost everyone is new to Cloud and no one yet knows what will eventually lead to nirvana for cloud.

Further the way IT market is coming off age, its turning that behemoths like HP and IBM have added quite a lot of breadth and depth to their portfolio. Everything said and done they are the strongest contenders to own the datacenter of the future.

At present the behemoths are exerting influence and potentially weakening the niche players(Cisco including) with their might(promoting proprietry products + endevours of making them best in class). The road ahead appears blocked till niche players come together combining their strengths and creating a formidable competitor (they will HP and IBM to join but will make effors to avoid them). Cisco with its leading networking and strong balance sheet might think itself as being perfect for exploiting this partnership. This option of having multiple permutations and combinations gives customers opportunity to choose the best fit and having it under one umbrella with can be a potential inflection point. The biggest challenge will be coordination and running the show which Cisco hopes it will be able to do.

So Sun acquisition and its portfolio will somehow affect Cisco's plans as firstly Sun is losing market share and it might not be able to turn it around as it just has cash but no experience in Sun's primary markets or presence in emerging countries etc. Sun will definately be a drain on cash on one hand and what it will make Cisco is more like HP and IBM but comparitively weaker. Cisco thinking that it smarter will like to avoid the scenario. Not acquiring Sun gives it perfect opportunity to partner companies which can add value and Cisco can just pick up perfect fit.

MAYBE thats why Cisco is not acquiring SUN