deal architect

Insight HCL

I am spending the day at the HCL Influencer Day in Boston today. I have long been impressed with HCL’s Product Engineering practice, which still has unlimited runway as it goes after the $ 1.4 trillion annual global R&D spend. I had cataloged the group and their work on the Boeing 787 Dreamliner project in my recent book.

The big news in the last few years is HCL has been making impressive strides in application, infrastructure, BPO and other services markets. My sources in the firm say they are being commercially disciplined (they use metrics like employee EVA to measure performance) even as they chase the “rebid” business (replacing incumbent outsourcing providers) they estimate at $ 190 bn over the next 3 years. Tough to argue with their 7X growth in market cap, even through the deep recession.

I did go into the day fresh from the GE Minds+Machines day  where they presented a data/analytics set of services with a relatively low labor component. Splunk presented to the Enterprise Irregulars yesterday their offerings around mining machine data. Over the last few years, I have cataloged on my blogs what Salesforce and Workday have been delivering in massive efficiencies in application management and upgrades, and what Amazon and other cloud infrastructure providers have been doing to redefine the data center. 

With that background, it is jarring to see in the morning slides HCL’s rapid labor growth, (impressively outside India in Brazil, Finland,  Poland and elsewhere). Vineet Nayyar, the company CEO proudly told Davos earlier this year he would create 10,000 new jobs in the US and Europe.  HCL’s philosophy is embodied in its “Employee First” culture.

Someone in the audience asked what HCL is doing for the CMO – all the sentiment analysis and marketing analytics that are exploding. HCL does not appear to have a good answer yet.

High labor, low machine services. Primarily focused on IT and CIO. That to me is the traditional outsourcer. HCL is clearly doing that cheaper, faster and will do well in what it calls the “rebid” business. Nothing wrong with that. I do hope I hear about other sides of HCL beyond just that during the day.

The Big Data laggards

In a panel on Big Data last week at Cognizant Community I cited several examples from the New Florence blog collected over the last few years – National Hurricane Center, Union Pacific, U of Washington’s ENCODE project, Social Media analytics and several others

In my recent book, Charles Zedlewski of Cloudera cited several of his customer scenarios:

• Electronics manufacturers can store and analyze usage data from
  the devices they ship so as to understand product issues in the
  field.
• Financial institutions are constructing more elaborate fraud prevention models.
• Pharmaceutical companies are cleansing and processing vast
  amounts of genomics data.
• Retailers are building machine learning models that predict individual customers’ buying preferences and make recommendations.
• Media companies are analyzing viewer data to optimize the content
  they present.

In his recent book, Rick Smolan brings out many applications all around the world.

Tom Raftery, in a recent blog post, provides several other examples from Europe.

The long and short is these projects have been on-going for years, if not decades, and are spread across industries and processes. They may not call their projects “Big Data” but their results are surely impressive.

Against this backdrop, I see so much marketing of the concept from IBM, SAP, Deloitte, Infosys and others. You would think the pioneers above have been using their tools and their talent for years now. Have they? I hope prospects look at a broader list of success stories , not just the references the vendors show them.

Even more, I hope customers do not accept the excuse “the concept is so new, nobody has references” or the proposition that customers should sign up for their MDM, change management and a variety of other products/services they push as “essential” for Big Data success. 

Meg Whitman and Ginni Rometty in 2016

We are at a historic point in the industry's evolution. Two strong and accomplished ladies are leading our biggest and most iconic tech companies, HP and IBM. As I read and heard about Meg Whitman at Gartner’s event in Orlando and about Ginni Rometty at an IBM event for CMOs in Paris last week, I pondered where they and their companies will be in 4 years.

Ginni, it appears, is on cruise control on a well paved highway and can afford to be chasing after new buying centers and making stylistic changes at IBM as Fortune describes below. HP is, in contrast, on a rocky road. Meg said last week 2013 will be a “fix and rebuild’ year. But so was 2012 and 2011.

Fortune:

You wouldn't catch Lou Gerstner or Sam Palmisano trying to smooth over someone else's faux pas. Rometty's two predecessors also are unlikely to have hosted a sales meeting in a loft, and they definitely wouldn't have described the proceedings as "neat." But they surely would have approved of Rometty's agenda that June day. She had assembled some familiar faces, the chief information officers who buy billions of dollars of software, tech services, and hardware from IBM (No. 19 on the Fortune 500), but she had also invited their chief marketing officers. (Thus the trendy venue.) Her ambitious -- and yes, unusual -- plan: Get the marketers to use IBM tools to sort their data for nuggets that will help them better reach customers and sell more stuff.

But appearances are deceptive.

What Fortune, and for that matter most of the media which follows IBM, do not point out is Ginni’s substantial challenge is not just to change the top soil, but redefine the aging core her predecessors have left her with. IBM Marketing does a good job presenting its Smarter Planet, Watson, digital marketing initiatives. But put together they make up less than 20% of IBM revenues.

The “core” is made up of 10, 20, 3o year old assets IBM continues to milk. An SAP practice which grew substantially with the PriceWaterhouseCoopers acquisition in 2002. The mainframe DB2 database which was first released in 1983. Lotus Notes first released in 1989 and acquired by IBM in 1995, Data Centers designed in the 80s and 90s with Cold War bunker mindsets.

Customers continue to utilize these products, but few of them are pleased when they see the IBM invoices for them. IBM still expects premium prices for these products, way past their prime. Ginni’s biggest challenge will be to get IBM customers to say “Neat” again.

Meg, in contrast, has an undisciplined portfolio of products that has been difficult to prune. However, she still has plenty of engineering talent important in a Polymath world where Apple, Google, Microsoft and Oracle are showcasing innovation via blended hardware/software and cloud capabilities. She also is well positioned as consumerization continues to influence the enterprise.  IBM has, in contrast, by spinning off its PC and printer divisions has been shrinking that engineering and consumer footprint over the last decade.

Also, what is considered Meg’s big weakness – its EDS heritage services and Ginni’s biggest strength – Global Services – may soon be flipped around. Meg realizes IT is still her major buying center. At Orlando, as an example she said “marketing chiefs matter, but HP will also work through the CIO." IBM in contrast, is using its IT centric services to sell to the CMO – a bit of a stretch unless its acquires some digital agencies and social analytics tools.

Both HP’s and IBM’s services group are challenged in 3 market segments – infrastructure outsourcing, application management and BPO.  The first and fast movers with rapid growth over the last five years have been cloud companies like Amazon, Rackspace, SaaS vendors like Salesforce and Workday (that are cannibalizing upgrade and application management services) and firms like Cognizant and Appirio. Granted they are growing from smaller bases, but they are growing explosively. Then there is another early trend – companies which outsourced to legacy service providers are bringing the support back in or are increasingly “churning” it. GM’s recent decision to bring 3,000 jobs in from HP may be a harbinger. Meg’s smaller footprint in the legacy service provider space may actually turn out to be a blessing.

Both Meg and Ginni have unique opportunities and challenges ahead.  I would love to do a book in 2016 cataloging how both of them have fared. No, nothing to do with the next Presidential race – but will be a good checkpoint for their performance in their new jobs and also fit well with my every 2 years book schedule.

Halftime in Enterprise Tech

Friday morning, Workday executives opened the NYSE and their market valuation quickly rose to a staggering $ 8 billion. To me, it was a fitting cap to a fascinating month in enterprise technology punctuated by Dreamforce, Oracle OpenWorld and several other events.

It’s halftime.

At least that is what I got from a conversation with Mark Hurd at OOW. Talking about SaaS vendors he said they only support a handful of the 750 business processes he presided over when he was at HP. Since he left, HP has announced plans to implement both Salesforce.com and Workday, but Mark is right. HP and other large, global companies have all kinds of complex supply chain, lab, warranty and other vertical processes (see excerpt of HP’s supply chain complexity from my recent book) so far untouched by SaaS.

Moving on to infrastructure when I asked him how Oracle would compete against Facebook and Google cloud efficiencies (also excerpted here and here from my book) he pointed out enterprise customers have not benefitted from those (primarily) consumer tech offerings. Amazon, Rackspace on the other hand, he feels, have only “mildly interesting” offerings for the enterprise.

Those are second half opportunities.

But let’s look at how Oracle and others have fared in the first half.  As Tim Chou wrote in his guest column, Oracle was a cloud pioneer with BusinessOnline as early in 1999. SAP has had a string of next gen projects and acquisitions – Project Heidelberg, BYD, SuccessFactors. IBM first talked about OnDemand Computing in 2001, and since they have had Cloud Manifestos but little Cloud Magic. Microsoft has sunk billions in its Azure data centers. They have had plenty of opportunities, but the performance can best be described as the equivalent of “a number of fumbles.”

Yes it is half time. And Clint Eastwood looms large.

There is the inspirational Clint from the Chrysler Super Bowl commercial promising things will be much different in the second half. And there is the mocking Clint from the RNC – the one of empty chairs, or in our case empty clouds.

Mark, Hasso, Ginni and Steve have plenty of reason to keep the second Clint far away. And Meg presumably does not just want to be a bystander.

It should be an interesting second half. 

Party like it is 1999 again?

Last week at Dreamforce I had a deja vu moment. It felt like SAP’s Sapphire in 1999 with Deloitte and Accenture and Wipro logos on big booths, on conference bags, on party invitations.

In the Marketing Cloud session, I wondered what value these traditional IT/ERP centric service providers would bring customers compared to digital agencies.  Likely,  they will continue to focus on Sales and Service cloud opportunities – for example, Siebel to Salesforce SFA migrations. But by now, that service should be priced much more reasonably than it was say, in 2007.

And that is where the 1999 risk comes in. These providers can hype up even a soiled brown bag.

Fortunately, we are dealing with a new CIO like Vail Resort’s Robert Unwiler:

“Because the EpicMix app deals directly with customers, it has pushed Vail's IT team to think more like a commercial software company. That means working closely with the chief marketing officer and her team. Urwiler's using outsourcers much more selectively now, and he's adding people with software vendor experience to his team. And Vail is benchmarking EpicMix against the likes of Facebook and Foursquare in terms of design and ease of use.”

1999 expectations, say hello to 2012 realities.

Small $$ for Big Data

Rick Smolan, former National Geographic and Time photographer, has used a crowdsourced model (of fellow media professionals) to create a series of “A Day in the Life of” books.

He is now tackling the impact of Big Data/analytics on many aspects of our lives and the book should be out later this year. Fortune has a gallery of some of the examples and how they are impacting farming, emergency services and other aspects of our lives.

I found the farming example particularly poignant

“The Climate Corp., a company founded by former Google executives David Friedberg and Siraj Khaliq, provides crop insurance for farmers by analyzing 22 data sets for weather every six hours, calculating about 10,000 scenarios that could happen to a grower during the next two years. They run 34 trillion different simulations that can be used to price hyper-local insurance rates for atmospheric calamity. When erratic rain caused one farmer's corn crop to fail, Climate Corp. automatically compensated him $45,000 for his losses. The farmer didn't even need to file a claim.”

In contrast, I am amazed at pricing for analytics projects in the enterprise. SAP HANA and consultants around it behave like it is still the go-go ERP days. Contingent pricing and payouts? HaHaHa.

I shudder to think what Watson will cost WellPoint over the years. I wonder if they went to Apple and asked for a custom version of Siri or to Google for Now. Or Yahoo! which has been the biggest contributor to the open source Apache Hadoop tool which the IBM Watson team heavily leveraged.

But even more so, the Smolan crowdsourced model shows even for complex analytical projects, there is enough talent around the world to leverage. No need to pay $ 500 an hour for hyped up consultants. 

Hush, Henry!

Henry Blodget gets some snide comments for just waking to the large software revenue that defines IBM.

Actually you cannot fault him.

If you see IBM’s TV commercials, or their social marketing or their pitches to media and investors, they are much more eager to be known for their Smarter Planet and Social Business offerings, even though they are only about 10% of revenues.

They would rather keep under the covers the 20-30 year old DB2, Tivoli, Lotus, WebSphere software. And Cold War era data centers. And the old SAP outsourcing practice. Those products are meant to just grind away and generate billions in revenue and margin, not publicity.

Shame on you Henry, for bringing them attention :)

Multi-Tenancy: Where are the outsourcers?

I was reading Phil Wainewright’s nice note on multi-tenancy “The cloud operates within a service-oriented architecture, where providers achieve economies of scale by delivering their services in volume and operating them from pooled resources.” (italics mine)

I drifted off wondering why the outsourcing world has not aggressively pursued – and passed along to clients – the benefits of economies of scale and pooling when it comes to talent.

It looked so promising a while ago.

In 1996, at Gartner, I wrote a research note, which introduced the concept of Shared Solution Centers (SSC).

“An SSC enables consulting clients to derive economies of scale and to leverage dedicated resources. Clients can achieve this by allowing their consulting firm to perform certain phases of the project at a solution center, or by remote dial-in access by the staff at such a center…. Good candidates for SSC-based services include configuration of software parameters, writing of custom code, performance benchmarking, testing, development of reports, and documentation development. A dedicated team of testing specialists can keep up with a variety of testing tools and techniques - across multiple application vendor projects.”

I even defined a maturity model:

In Stage 1, an SSC is primarily a "showcase" for software demos and training. In Stage 2, the SSC can be used for "jump starts," while the client builds its own infrastructure. In Stage 3, certain project phases such as report development are done at an SSC. In Stage 4, more project phases including testing are done at the SSC, and the SSC reaches "scale" - at any given time 20 or more clients are being serviced from an SSC. Finally, in Stage 5, there are phase-based "factories" where testing or data conversion specialists and tools can be leveraged across multiple projects, whether on SAP, Oracle or other software.

In 1998, I described Andersen Consulting (now Accenture)’s SSCs in another research note

“AC's ERP-focused SSCs (11 in the United States, 11 around the world) have allowed its clients the following flexibility:

• A U.S. AC client used the Las Colinas SSC for R/3 Basis support, Chicago for mass-report development, and San Diego for end-user training development.
• A South American client moved its design team to a U.S. SSC. It reported improved productivity by being closer to the specialists instead of having them fly to its location.
• Several clients have outsourced "post-live" SAP operations support to the AC Cincinnati SSC.

The SAP model is being leveraged to other ERP initiatives (e.g., PeopleSoft) and for other IT services. Today, AC SSCs house a staff of 2,000, expected to grow to 8,000 by 2000.”

In the last decade I have seen the emergence of huge outsourcing campuses in Bangalore, Beijing, Budapest and elsewhere. `They would appear to be the realization of what I had described – huge economies of scale and pools of specialist resources.

But go into the buildings, and you see floors or sections of floors dedicated to individual clients. They are secured tight. Restricted access, PCs with USB ports sealed off – all in the name of protecting client specific IP.

While there are some shared economics – in training, in physical and telecom infrastructure for the staff, the big sharing opportunities I had envisioned have not happened. 

Outosurcers say clients tie their hands. Sure, if you are doing proprietary product engineering work for a client, they absolutely want it kept away from other companies. If you are supporting a custom legacy application, there are not many sharing opportunities. But when you are supporting hundreds of SAP or PeopleSoft  environments there are plenty of commonalities across the clients that can leverage pooling concepts.

I am actually surprised cloud vendors don’t use this more in their positioning against on-premise vendors. Look at how few staff Salesforce.com or Workday need to support thousands of customers. In contrast look at how many staff IBM or Infosys need to support each customer in on-premise mode.

Insourcing – more than a political slogan

Insourcing is on US TV every day thanks to back and forth between the Obama and the Romney campaigns.  Outsourcers may chuckle, smirk and discount the term as it is used in an opportunistic, nationalistic way. But they should be wary as Randy Mott, now CIO of GM, has announced his own version of insourcing. He wants to move from 90% outsourced to 90% internally staffed in 3 years. Now many will even discount what Randy says – I have already heard he came with an “agenda” from his last job as CIO at HP and GM’s largest outsourcer is EDS, now part of HP. Or there is ServiceMaster and its insourcing from IBM.  You could argue those are exceptions,  but I started hearing from several CIOs 3-4 years ago they felt they had “outsourced too much”. That led me to write in The New Polymath

…most enterprises spend 80 percent to 90 percent of their IT and telecom budget with outside vendors. That is a dangerously high dependence on a supply chain. ..Even in automobile and aerospace, which have much more mature supply chain management, companies such as Ford and Boeing get around 70 percent of components from suppliers. In those industries, supplier consolidation and tiering makes sense because Ford and Boeing are dominant customers and can dictate design standards and justifiable economics.In technology, by contrast, buyers have little leverage — even companies as big as Ford and Boeing make up less than 0.5 percent of an IBM ’ s or Oracle ’ s revenue.

Many outsourcers are sowing their own seeds of discontent. As I wrote last week, many of them seem bored with their current work assignments. Still others have eyes bigger than their stomachs. I was talking to a salesman at a Tier 2 offshore firm and he told me he is getting invited to deals at many big brand customers. “They are being neglected by their incumbent providers. The Wipros and the Infosyses want IBM and Accenture size commitments from their customers”.

And there are new competitors. In infrastructure services, players like Amazon and Oracle. On customer/revenue opportunities, digital agencies.

So, chuckle if you want when Joe Biden talks insourcing.  But when a CIO mentions it, while he may not bring it all back in, he is likely looking around.

And it will not end in November. 

The elephant in the IBM Cloud

Gery Menegaz’s bio says he is an “Executive IT Architect” for IBM. The small print says his columns for ZDNet including the recent one on clouds don't necessarily represent IBM's positions, strategies or opinions.

Except that they do.

His first point says “Data center migrations are disruptive, costly, and complicated. Tightly coupled mission critical systems that make up the minority of the data center take the most time and expense to migrate”.

I have heard some version of this point in countless IBM proposals to prospects, presentations to clients, its ads.  In fairness, I hear it from many other legacy vendors, but IBM being one of the biggest enterprise hardware, software and services vendors shows up in my consulting and my research quite often.  IBM was one of the first to talk about “on-demand” computing (starting back in 2002), it acquired the ASP pioneer, Corio ( in 2005) and the cloud integration tool Cast Iron (in 2010). It has been flirting with clouds for a decade.  

Its point boils down to this: “the status quo is safer”. It may be so, but can the status quo not itself improve, evolve? Why does it need to be so expensive and onerous? Why over the decade have those migrations not become mindless?

Here is what has been happening in the industry in the last decade:

a) The Data Center has been a beehive of innovation. In my two books I have described countless improvements in the Microsoft Azure, Google and Facebook data centers. Location in obscure places, cooling efficiencies, vanity free servers, tax breaks – I could go on and on. Amazon and Apple, in their own way, are taking such innovations to another level. IBM is one of the biggest data center outsourcers in the world. Why has its status quo not improved dramatically?

b) SaaS vendors like Salesforce, NetSuite and Workday are showing remarkably efficient shared services when it comes to talent to support thousands of customers. Upgrades in their world are an absolute dream. IBM is one of the biggest application managers in the world. In the last decade it has dramatically grown staff in low cost pools like India, but should it not be emulating the SaaS vendors and moving to highly efficient and automated application management? Why has that status quo not improved dramatically?

c) Software maintenance and release management has become much more automated and cheaper with offshore resources. IBM is one of the biggest IT infrastructure software vendors with DB2, Websphere, Tivoli, Notes etc. The efficiencies it should have garnered in maintaining those relatively stable 20+ year pieces of code should be in the tens of biliions in the last decade. Why has the status quo in the form of its charges to customers not improved dramatically?

d) If you look at the projects in the services ecosystem around Azure, if you look at Appirio’s countless projects around migrations to the Salesforce, Google and Amazon clouds, they are efficient and safe. IBM is the largest systems integrator in the world and has had plenty of experience with its own Corio, Cast Iron and other migration/integration tools. Why has that status quo not improved dramatically?

Next time you hear IBM say a move to the cloud is disruptive or unsafe, ask why IBM has not made it less so? But even more specifically ask why its status quo has not become much more efficient in the last decade.