Bhaskar, I can totally understand how these conversations would have gone even a few years back. While these are the challenges, they also point to the significant untapped strengths these companies have (knowledge of customer's processes and levers for operational efficiencies etc.). In these difficult times I am sure the savvy ones will figure out how to leverage them better.
A tangential note on processes: most of IT work is related to business processes, and that is one of the reasons I feel automation is not an existential threat for the industry as a whole. It is just a bump. On Wed, Oct 19, 2016 at 11:06 AM, Bhaskar Dasgupta <bdasgu...@gmail.com> wrote: > one of the examples I had asked to be funded was to leverage their data. > this company manages banking processes. what i wanted was to tie up with > ISI (not that one) and hire a small skunk work of data scientists and a > data design / visualisation centre. And then wanted to do what rolls royce > have done with their trent engines - they make a stupendous amount of money > by monitoring their engines on a real time basis in flight and saving > airlines shed loads of dosh. So i would’ve provided a set of tools, > constantly evolving, to the heads of operations on their process flows, > heads of sales on sales analytics, heads of product design on competitive > features, and so on and so forth. And once I have sufficient coverage, I > can setup a banking product market place. World Domination! result? god no, > we cant pay these phd’s that much! no? then you will lose them to american > firms who can and will. but that will cause the pay scales to be fucked up > internally. ok, lets spin off this firm. we don’t do spinoffs. why? 100% > owned subsidiaries are good and actually you can IPO it as their multiples > will be better. Oh! that decision is above my pay grade (this is the > president of the division!) you can fuck off. /facepalm. > > forget about creating new products, buggers don’t even leverage what they > have! they are sitting on a fucking gold mine of rivers of data (if you > don’t mind me mangling metaphors) and are happy to sit there and fish for > minnows or get paid lowly for tending the sodding river bank. > > > > > On 19 Oct 2016, at 06:06, Suresh Ramasubramanian <sur...@hserus.net> > wrote: > > > > IT companies buying product companies in a desperate bid to innovate .. > let us just say that I’ve seen a lot of that happen at a previous workplace. > > > > The usual end result is that the founders and key employees quit in > disgust after a while and those that are left are gradually absorbed into > the company doing something totally different than what they set out to do. > > > > And meanwhile the product itself is killed off immediately, or maybe > dies a slow and lingering death with a few legacy customers left behind and > practically zero further development. > > > > Big companies that don’t have DNA beyond being pushers of software that > most if not all users have a visceral hatred for, and/or bloated services > contracts, are absolutely not going to infuse any magical fresh DNA into > them by acquiring successful product companies > > > > The prospect of such foreign DNA taking root in the company is far less > than in the case of an organ transplant – the sort you get in mad scientist > movies where a scientist transplants human dna / tissue / whatever into an > ape and suddenly ends up with a super intelligent planet of the apes or > Gorilla Grodd variety animal. > > > > Mohandas Pai is a smug and opinionated twit but he got one thing right > though. The software industry didn’t die – it will survive and it will > probably hang on, but the traditional indian (or even foreign) services > model is long dead in favour of automation. The only things that won’t be > automated to a large extent are higher up the value chain than such > companies generally play around at. And the hanging on will be the way a > really old and sick man keeps hanging on – perpetually in the chasm between > Allopathy and Tirupathi. > > > > In other words, the days of 15% raises are dead and gone, companies > outsourcing basic bargain basement sysadmin and datacentre work will > outsource far less after automating the hell out of everything they can, > testing will be automated. > > > > Of all the cash cows out there, telemarketing and support still needs > humans to a larger extent and will hang on but higher up the value chain – > because a lot of it has moved to social media, marketing rules in this > space have tightened etc. > > > > And even that is going way down after all the tech support scams in > India that flourish tarring even the legit players with the same brush, > with at least some of the legit players looking wistfully at the “upsell” > angle that, if pushed a few hairs farther down the line, becomes those > scams where someone claims to be “tech support” for your OS or device > manufacturer and cons you into paying for a $50 a month perpetual contract. > > > > --srs > > > > On 19/10/16, 10:15 AM, "silklist on behalf of Deepak Shenoy" > <silklist-bounces+suresh=hserus....@lists.hserus.net on behalf of > deepakshe...@capitalmind.in> wrote: > > > > Apologies for the plug but I wrote a piece a year back: > > http://capitalmind.in/2015/01/the-inflexion-point- > > for-the-it-service-industry-long/ > > > > So my point is that the problem isn't with IT companies - they will > survive > > as IBM and HP etc have, and perhaps grow in single digit percentages > and > > generally get a lot lower price to earnings multiples. They do stuff > no one > > currently wants to do and overhauling a system to use stuff that other > > smart people want to use is too expensive. (Like COBOL - it may be > outdated > > and all that but it still forms the basis for an irrationally large > part of > > banking) This work will continue until you have removed the very need > for > > the basis that the older software has been required, so there will > always > > be business for the Infy/TCS/Cognizant types from this kind of > maintenance > > work. > > > > But things can change very fast. 10 years ago I couldn't imagine that > BigCo > > would be able to manage something like 5000 servers using less than 5 > > people including bringing more up when required, at run time, with > complete > > reporting/control available even on a mobile phone app. It's possible > now. > > It's routine now. And as more companies are discovering it is, the > on-site > > and off-site infrastructure work that was handled by the IT cos has > seen > > dwindling business. SAP now offers a cloud based pre-setup solution, > where > > you need none of the server infrastructure - only the initialization > pieces > > which the IT cos still do; but at some point SAP will create modules > that > > are one-click installs for the kind of industry you are in. > > > > Automation isn't robotics - you don't need machine learning or AI for > most > > of the work required. For instance much of the testing work that is > done > > tends to be checklist driven, and some of that has already been > automated > > at multiple levels using tools. What these IT cos should have been > doing is > > buying the product companies that build these tools, but they simply > don't > > have the DNA. (I would even say buy minority stakes in them with a > board > > seat) Their own products are absolutely crap; see the quality of the > stuff > > TCS and Infy have built for say the MCA, versus teh quality of design > that > > seems to be in teh new Modi camp (vidyutpravah.in for instance or > such). > > > > I heard of a bigco - a friend works there - where an insurance > company was > > being pitched by various vendors for a certain solution. The big > Indian > > names went in with the powerpoint smoke and mirrors thing and offered > > things like 6 months to a year to finish with X headcount etc. A > russian > > company with the founders as programmers pitched a working prototype > that > > they said would take a couple more weeks to finalize and they'd get > it live > > in amonth at a cost that was not even in the same area code forget the > > ballpark. They actually won the project and BigCo guys had multiple > > meetings to "create risk mitigation techniques" and such things. It's > kinda > > funny. > > > > I think if IT cos react, they will get lean. If they get lean, many > many > > people - and I'm speaking thousands in bangalore alone - will find > out that > > their skills are drastically short of the real world requirement of > jobs. > > This is the problem - not that the IT cos won't survive. IMHO. > > > > Deepak Shenoy > > Capital Mind: Financial Macro and Market Analytics > > http://capitalmind.in > > Twitter: @deepakshenoy > > > > On 16 October 2016 at 15:34, Bhaskar Dasgupta <bdasgu...@gmail.com> > wrote: > > > >> i was interviewing for one of the IT corporates some time back for their > >> COO position and once i managed to dig a bit into their financials, i > >> backed out. the majority of their revenue streams are from processing in > >> advanced stuff, processing code, processing transactions, processing > >> quality control. They do this very well. Very very well. standardise the > >> process, six sigma the shit out of it, hire the great unwashed herd of > >> graduates pouring out of the universities - retrain them to be great > >> processors and great business model. But this kind of model is very > >> susceptible to dis-intermediation from further advances in technology. > When > >> I asked if can have some serious seed funding to develop products rather > >> than just provide services, there was a bit of a hoo ha. I think a > product > >> plus service model is the best option, create great products and then > have > >> a long tail in services and maintenance contracts. we have some of these > >> products but not enough. not easy to develop products - the eco-system > >> isn’t there yet. > >> > >> so whilst i don’t think its the end of the road, but for example, every > 2 > >> months I am in a conference where vendors pitch up talking about robotic > >> process improvement or AI and how they are showing 20–50% reduction in > warm > >> bodies in agency/outsourced/offshored units. Where will these 20-50% of > >> highly trained processors go when the infosys or TCS lets them go? > >> Thankfully the economy is ginormous and we are well accustomed to > poverty > >> and pain and still have the joint / extended family to fall back upon. > But > >> for the IT industry? pain... > >> > >> i agree with Srini, changing careers is not easy for us desi’s….(says > the > >> man who has made a career of changing careers, heh). > >> > > > > > > > > > > >
