deal architect

Barbarians at the Cloud Gate

In the last year, as I built the model for SAP Nation, I have had conversations with several cloud vendors and shared my view that outsourcers/SIs from the on-premise world are looking for new parking places for their massive workforces and infrastructures. Cloud software vendors should leverage this bounty but be careful of the baggage many of those providers may bring along.

Some of those warnings need to be amplified as Phil Wainewright writes about the Salesforce partner ecosystem

“In annual revenue terms, they generate at least three times Salesforce’s own revenues, or close to $20 billion, according to new research from IDC presented on stage at the morning’s partner keynote.”

Time for me to write “Salesforce Nation”?

No.

Let’s put things in perspective. In SAP Nation 2.0, I actually increased the estimate for the economy from $ 200 to over $ 300 billion a year. And the SAP economy has a long track record of massive overruns and failures. The number above does not include amortization/writeoff amounts.

So, the Salesforce ecosystem estimate is not that worrying. Also, unless the cloud model deteriorates dramatically, it will never approach the size of the hosting, upgrade and application management economics which plague the on-premise model.

What is more worrying is the focus on outsourcers as a brand validation and sales channel versus them as a customer delivery support extension. Too many software vendors focus on the former and not enough on the latter.

If cloud vendors do not start tracking and benchmarking success/service levels/economics the partners are delivering, in a few years I may have ammunition to write Salesforce or Workday Nation. I hope not.  

Indian outsourcing: The lost decade

The WSJ profiles Dave Smoley, CIO of AstraZeneca in a story about how cloud computing is pressuring Indian outsourcing. Dave is one of my favorite CIOs and he had allowed me to showcase in SAP Nation his remarkably brave initative to invert the 70% outsourced model he inherited and about the cloud investments he has been making.

But the WSJ could have gone further back. In The New Technology Elite, Dave had described an even braver decision to go with Workday and its implementation consultants for a global HCM project when he was CIO at Flextronics. In 2008, the much safer decision would have been to stick with SAP and a global SI. But Dave did not.

Actually the WSJ could have gone even further back. In 2006, I had blogged about a contract format I had started negotiating for other similarly brave CIOs which blended application management and hosting, on a per user basis. I called it Software as a Customized Service (SaCS) and for years I have implored Indian vendors to move to this model.

Three challenges have kept most of them from getting there, even a decade later. First, they have shied from CapEx needed to offer infrastructure services, so invariably have to partner with a hosting or cloud provider. Investments in Indian world are measured in recruiting and training days, not in data centers and servers. They have also shied away from automation investments to cannibalize some of their labor. Second,  (as a corollary to their labor intensity), they are still stuck with headcount/FTE pricing models and reluctant to offer as-a-service pricing.  Finally, and probably, much more alarmingly, a continuous improvement culture (which was proudly developed when their commitment to CMM and Six Sigma was intense in the 90s) has vanished. That had helped offset some of India’s wage inflation.

It is gradually changing. Vishal Sikka, formerly of SAP, now CEO of Infosys is pushing for more automation. "The company has set a long-term internal target of up to 70% automation in its infrastructure management business, and up to 55% automation in its business process outsourcing business. Currently, 35-40% of its BPO processes are automated.”

But that automation should have started years prior at Infosys and other Indian vendors. As should a change in business models. Now brave CIOs like Dave are much more mainstream, so the urgency has only increased significantly.

NASSCOM: More Moore, less ABAP

My friend Phil Fersht was at the annual Nasscom event in Mumbai last week and wrote “More MBA, less ABAP is the recipe for India’s digital success”.

There is much to like in what he says, but I would go further

a) Much Less ABAP

My book SAP Nation shows a $ 200 billion a year ecosystem ready for disrupting, even collapsing. Part of it can come from Indian firms cannibalizing their own services in that market. Much more could come from Indian firms diversifying to SaaS vendors that SAP customers are migrating to and providing  third party maintenance to SAP customers. Unfortunately, many will just tell their customers to wait and see how SAP’s next-gen product, S4HANA matures.

b) Digital with Big D

As I wrote in this post, Digital Transformations require a wide portfolio of skills. “There are so many opportunities to digitize products, channels, shop floors, logistics, business models etc.  So many legacy systems need to upgraded and integrated to. The portfolio of skills needed to build a vibrant Digital Enterprise is very wide.”  Many Indian firms have some of these components in place, but many need to acquire digital agencies, product centric/R&D focused practices, and as Phil says maybe even some MBAs for digital strategy and business model transformation.

c) Refocus on Moore’s Law

This is probably the biggest change Indian firms need to make. They have to quit leading with labor intense models. A blend of symbiotic man-machine skills is key. Indian firms should have been leading the automation of knowledge work. They missed the last wave where Google, Amazon, Microsoft and others have shown the art of the possible. Plenty more to be done there. And even where the model is labor heavy they need to dust off their CMM and Six Sigma disciplines they religiously applied 10-15 years ago. As I wrote here, Moore’s Law is back in the enterprise, and to offset the wage and real estate inflation that plague India, Indian firms need to become even more acutely focused on continuous improvement.

d) Become geographically agile

When I interviewed Mike McNamara, CEO of Flextronics, the large contract manufacturer last year, I was struck by his comment that they moved a quarter of their physical plant across borders. Manufacturing has become agile – to take advantage of changing US energy economics, to be closer to emerging middle classes in Indonesia, Brazil or wherever and other trends.  If physical plant can be moved quickly even with all the logistical challenges,  there is no reason why labor pools cannot be similarly agile. Indian firms need to evolve the delivery model – doing the bulk of the work in India may not be the best way any more to service clients in the throes of complex digital transformations.

Sikka and the state of outsourcing

Vishal Sikka, the new CEO of Infosys, has some candid commentary on the state of the IT services market

"It's like a treadmill of increasingly lower cost, hiring people faster and faster, from more and more mediocre places, training people less and less, putting them into job faster and faster. I think that is a wrong direction."

Problem is someone like him should have been sounding the alarm bells years ago. The comments I heard during interviews for SAP Nation showed significant customer and market watcher disappointment with systems integrator/outsourcing performance

Here are some of those comments from the book:

Chris Day, AstraZeneca’s VP of IT Transformation which is looking to revert its 70/30 outsourced/insourced model

“Outsourcing leads to a culture of generalist service and contract managers on both sides who don’t understand the technologies they are accountable for — so everything is “we’ll get back to you” as they seek out those that do understand, which is especially painful when issues cross supplier boundaries. The end result is service is put at risk with costly, non-agile project delivery and no improvement in user satisfaction scores. Any idea that outsourcing partners are able to drive down ticket volumes and raise user satisfaction scores without considerable involvement and expertise from AZ is now long gone.”

Ben Stewart, at Inteva, discussing SAP outsourcing

“The SAP partner support experience was even less appealing. We had offshored application management to TCS, hosting to HP and had a program manager in CSC. In a crisis situation, it took half-an-hour to figure out who everyone was and what they did before we could even get started.”

A customer  describing a SAP data center of one of the larger outsourcers

“I would grade it as a Tier 1 data center (Tier 4 being most robust). The disaster recovery site was only about ten miles away — not exactly comforting from a redundancy perspective. For a multitenant setting, there appeared to be too much glass to feel comfortable. Several portions of the center did not have a raised floor. And, it appeared very quiet for the suggested customer count or related support staff.”

Amit Bendov, describing upgrade project experience while at Panaya

“SaaS vendors do have the advantage of being able to upgrade their core at thousands of customers at one go. In the SAP world, the upgrade was one customer at a time….Most SIs who chased after upgrade business viewed it as a loss leader, a stepping stone to other business at the client or a way to preclude other firms from getting in the door.”

Shaun Snapp about the numerous project failures and unwillingness of outsourcers to fess up or fix things

After reading a number of articles on SAP project failure, I have noticed a consistent theme in much of the coverage. This is another example of how every time an SAP implementation goes sour, it seems as if the official publications put more of the blame with the clients

On the conflicts of interest in the services sector

“In many ways, SAP’s relationships with accounting firms grew even more complex. As an example, Deloitte auditors opined on internal controls around SAP systems, their consultants helped implement SAP software, and their advisors helped clients evaluate competitive BPO and offshore services around SAP. So, the conflicts of interest grew even more thorny.”

On business models that need to evolve

“Clearly, SAP partners were not ready for the utility computing
model in 2006. Heck, Nicholas Carr would not write his
best seller on the topic, The Big Switch, till 2009. As I researched this book, I asked several outsourcers if they offered the model today. More are starting to offer managed clouds, business process, and HANA-as-a-service, but there are still only a few customer references for such offerings.”

and

“Consultant travel continues at staggering levels even as telepresence has improved and many project phases should be done in “factories” — not at client sites.:

It’s not all negative. The book describes some of the pioneering work Appirio has done around cloud projects, the “outcomes based” implementation Cognizant did for British Gas and the low-fuss, low-cost way DeVry implemented SAP at several Brazilian locations with a local firm.

There is considerable coverage of how cloud data centers and SaaS vendors are showing how a new form of “outsourcing” which leverages automation and innovation.

In many ways, SAP/ERP are “old services markets”. There are many new ones with vertical apps which leverage mobile, social, analytical, wearable, sensory, robotic and other technologies.  There are digital transformation opportunities.

But Vishal is correct – the tendency of  the industry is to keep building pyramids of staff. Not automate. Not continually improve. Not surprisingly, the SAP economy model is 70% labor related costs.

He is right to “cry wolf”. The question is how many of his peers are willing to acknowledge the industry has many challenges. The bigger question is whether it is too late for many customers.

Vishal Sikka at Infosys

Vishal Sikka has landed as CEO of Infosys, a few weeks after a tumultuous exit at SAP.

My reaction – beyond the short term calming of jittery investors, Vishal will have his plenty of other opportunities. As I wrote last year in How Infosys lost its sheen the problems run deep.

What Infosys needs is to move application management, its bread and butter business, into the newer multi-tenancy world. SaaS vendors support 50 to 100X more end users  per support staff. Upgrades of thousands of customers are done in days not years common in the Infosys world. Given India’s higher wage inflation, the pressure to be more productive is even greater. Amazon is delivering lower data center price points every quarter. Can Vishal push that cloud mindset?

Given his recent HANA focus, I can see Vishal pushing analytics as a focus area for Infosys, He will find a newer set of talent sourcing companies have been exploring. Players like Mu Sigma and BGI have already scaled to thousands of staff in India and China respectively. They are not traditional outsourcers (BGI focuses on genomic analysis) but are showing clients there is a potentially different staffing model – MasterCard is a client of and an investor in Mu Sigma. Even more interesting is Kaggle’s crowdsourced data scientist model. Can Vishal position Infosys as a differentiated analytical player?

Given his software heritage, I can see Vishal take Infosys into more IP enabled revenue or product engineering services. Two Indian companies, HCL and Wipro have traditionally done better in product engineering, and even they have a shrinking market share as companies invest heavily in their own IP enabled products and services (In The New Technology Elite I had profiled such investments in 75 industries) . As an example, most auto companies these days have hundreds if not thousands of software engineers on staff in their product engineering/R&D groups working on infotainment, safety, fuel efficiency and other technology enabled features. Can Vishal reposition Infosys there? 

The opportunities Vishal walks into are many. One thing is for sure – Infosys’s presentations at Oracle Open World should improve dramatically. The last few have been pretty uninspiring

Indian outsourcing at the cross roads

Vivek Wadhwa writes “Why I have become pessimistic about Indian IT”

“It is not that Indian outsourcers have become less capable of servicing Western needs. It is that their customer base—the CIO and IT department—is in decline.”

He is right and wrong.

Actually there are large chunks – over $ 300 bn a year in IBM, Accenture, EDS, Cap and other outsourcing business (and over priced contracts  at incumbent Indian providers like Infosys and Wipro) for the Indian firms to still target. So, by flying low – and aiming for what HCL calls the “rebid” business which I estimate will be for about half that contract value – the short term road is still ok. When it comes to scale, IP protection etc Indian firms continue to have an advantage over E. European, S. American, Chinese competitors. But they have to compete vigorously on economics not continue with poor habits they have picked up in last few years. And dust off the continuous improvement ethos which was a hallmark of Indian outsourcing a decade ago.

In the other half of that incumbent spend and in newer markets, the challenge for Indian firms is more from clouds, crowds, cybertech and  customer priorities.

Clouds apps vendors like Salesforce, NetSuite and Workday have shown the world tens of thousands of customers can be supported and upgraded by a few hundred staff. That is quantum more efficient than the current Indian model where a single SAP customer needs 20 to 100 support staff. Cloud infrastructure vendors have invested billions in hyper efficient data centers and networks. Most Indian firms are capital allergic and have made few such investments (other than in broadband pipes to the West)

Crowdsourcing is starting to affect chunks of the outsourcing model – low end stuff like testing, but with models like Kaggle even high end work.

Activity after activity is starting to factor robots, drones, machine intelligence. We have moved into a world of cyber-physical IT but most Indian firms are still persisting with head count heavy business models.

The biggest challenge comes from growing customer investments in product and revenue centric technology. Many companies are beefing up their own technology staff to do that and looking for creative product design firms, digital agencies, contract manufacturers and logistics providers with scale. Few Indian firms fit that bill. Most have chased after Microsoft or Oracle and built relatively “safe” practices around mature, horizontal IT tools, not in emerging and creative vertical areas. 

As Vivek says “There are new $100 billion opportunities that could revitalize this industry. But from what I’ve seen, Indian executives seem incapable of steering their ships in the right directions.”

That may be too harsh an indictment, but it is clear what worked for Indian firms in the last five years will not be sufficient going forward.

The Brave New World of Software and Services

On the anniversary of GE’s launch of the Industrial Internet, CEO Jeff Immelt presented an update last  week in Chicago. You can watch his full talk here.

Couple of things stood out from an outsourcing industry POV

As I wrote during the launch last year, GE is the biggest outsourcer (and software company) most of us have never heard of. And it is delivering services not as much with armies of humans but with a growing number of  its "brilliant machines". An example is its 2.5-120 wind turbine. Analyzing tens of thousands of data points every second, the turbine integrates energy storage and advanced forecasting algorithms while communicating seamlessly with neighboring turbines, service technicians, and customers.

The other thing is GE announced Amazon Web Services, Kaggle and Accenture as three of its II partners. Amazon has pioneered infrastructure as a service and Kaggle is pioneering crowdsourced talent as a new type of outsourcing. The Accenture here is not the SAP SI of old. This is a joint venture with them called Taleris which leverages GE Aviation’s predictive analytics technology to make recommendations based on this analysis to optimize aircraft maintenance and flight operations.

Also, notice who is missing. The Indian outsourcers. That is striking because GE more than other company in the world contributed to formative growth at Infosys, Wipro, Genpath and many others . It still leverages a number of Indian providers for its internal IT but - not yet anyways – for external facing opportunities.

Finally, look at the likely impact on the software industry as its medical devices, locomotives, appliances and countless other products incorporate algortihms and commuunicate with each other and the world and allow for predictive maintenance, performance monitoring etc. A GE executive I interviewed this summer said "The greatest software companies in the world don't sell software". Indeed.

Not your dad’s outsourcer

Readers of this blog know I have been cataloging massive changes in outsourcing for the last few years – many new entrants driven by mobile, social, cloud technologies; changing business models to non-linear, usage and outcome  based; changes in delivery models and locations.

Reading the latest version of the Cognizanti magazine that editor Alan Alper sent me (you can register here to get a copy), summarized well how dramatic the changes are. The issue focuses on “everything as a service”, “the future of work”, “Code Halos”. There is no sight of SAP, project overruns, staff turnover and other dirty words that have become synonymous with last-generation outsourcing.

Just clever marketing? Not if you look at their experience and detail behind every article in the issue.

Cognizant shared with me examples of 19 BusinessCloud offerings across several verticals – proprietary IP that is available to clients to license as a service. The Code Halo case studies and methodology are being codified in a book and a mobile app (and have been profiled in a series of guest blogs earlier in the year).

Earlier in the year, I sat down with the CEO, Francisco D’Souza and he told me about his 3 Horizons – his quest to shake Cognizant  from the invariable grip of the “Innovator’s Dilemma”. The Cognizanti issue shows the many ways his vision has changed the firm – and its clients.

What have you done for me lately?

This article about IBM's large contract with customer Bharti Airtel in India provides a nice microcosm of the churn happening around legacy software and outsourcing contracts. Bharti has moved away from Lotus Notes as its email platform and is rebidding many components of its IT infrastructure outsourcing deal - one that IBM has touted over the last decade as a case study of its ability to thrive in emerging economies.

A decade however is a long time. Incumbent software vendors and outsourcers - not just IBM - have innovated little and delivered even less in the way of continous improvement as labor economics offset any productivity. Besides, little of the email adminstration, laptop refreshes, backup support etc is considered "strategic" enough today to justify premium rates.  And they are increasingly being benchmarked with non-traditional providers of storage, help desk providers - SaaS vendors, IaaS vendors, support services from the likes of Geek Squad. 

Against those benchmarks, EVERY traditional outsourcer, not just IBM, looks bloated.  In outsourcing, given it is a people intensive business, a little too much attention is paid to "relationship management". Frankly, way more attention should be going into "benchmark management".

Every outsourcing arrangement signed more than 3 years ago is going to be revisited in every customer around the world. In that sense, Bharti is just one small transaction.

A sobering moment for IT Services

Two large agencies, Publicis and Omnicon, announced a merger yesterday.  Famous brands such as BBDO Worldwide, Leo Burnett and Saatchi & Saatchi will be part of the new empire. With merged revenues of $ 23 billion, it will become the world's largest advertising firm, displacing WPP. 

In the last year or so, IT services firms have been quite excited about the CMO buying center. They hope their digital expertise in other areas of the enterprise will give them a leg up. The size of the agencies and their heritage of CMO access should be a wake up call on how difficult the CMO center will be to penetrate. These agencies are no digital slouches - Publicis, for example, has well established digital units in DigitasLBi and Razorfish.

Even more sobering are Avi Dan's comments about the merger

"Technology and digital media distribution are now more valued than the creation of content: the combined market capitalization of the new company is $35 billion, while Google's market cap is $295 billion, almost 10 times as bigger on revenues of $31 billion, with a third as many employees as Publicis Omnicom’s 130,000 staff. The world has changed and as it moves at the speed of technology, and the indispensability of scale is going the way of the three-martini lunch."

The comment about scale is also true in IT and BPO services. The large armies of consultants that the bigger firms have accumulated are a reminder of too large ERP and Y2K projects. The world has surely changed since. As with the advertising space, small, talented teams is what clients covet as I wrote here