As often happens, i was looking for something else when I came across a 2008 article in IEEE Computer by Mary Lacity and Joseph Rottman (don’t worry - the subject matter is covered in their 2008 book, “Offshore Outsourcing of IT work).  Having just set one of our clients off on the implementation of some outsourcing options to supplement their in-house resources, the list of 20 major effects of offshore outsourcing reported by project managers caught my eye.  I strongly recommend that you think of this list as a set of risks that need to be mitigated in any outsourcing implementation.  With apologies to the authors, I have sorted their list according to my highest priorities.  The in-house Project Managers reported that they:

• needed a mentor the first time they managed a project with offshore resources
• had to motivate the supplier to share bad news
• had to make offshore suppliers feel welcome and comfortable
• needed to thoroughly verify the offshore supplier’s work estimates, which tended to be optimistic
• had to provide greater detail in requirements definitions
• had to do more knowledge transfer up front
• were forced to shortcut the knowledge transfer process because of deadlines set by senior IT leaders
• had to ensure that knowledge transfer was successful by testing the supplier employees’ knowledge
• had to set more frequent milestones
• needed more frequent and more detailed status reports
• required more frequent working meetings to prevent client-caused bottlenecks
• needed to accompany offshore suppliers to all client-facing meetings
• experienced higher transaction costs which threatened their ability to deliver projects on budget
• experienced project delays which threatened their ability to deliver projects on time
• had to guarantee that the supplier followed pre-agreed knowledge renewal practices
• had to ensure that the supplier transferred knowledge about new applications or technologies to the client
• had to learn about new applications or technologies independent of suppliers to ensure that the suppliers information and bids were valid
• had to integrate the suppliers CMM/CMMI processes into their own project management processes
• had to ensure that the supplier’s employees were fully trained as promised by the suppliers
• had to fill many of the roles the the PMO should have performed

In the May (05/2009) print edition of CIO insight, there was an interesting article on “Vendor Management.” It included a list of six mistakes that tend to be very common in outsourcing where the client-vendor relationship is not as mutually beneficial as it can, and should, be.  I think it’s a good list so I have included it here with some commentary and examples of my own:

1. Failing to speak with one voice

Basically, the vendor is confused by mixed signals coming from the client.  I have seen this happen most often when  there is a power struggle between a central procurement or sourcing group and the IT team that works day-to-day with the outsourcing vendor.  This is a classic IT Governance problem and there is no right answer (although I know who I think should win).  The important thing is to resolve it unambiguously for all the participants including the vendor.

2. Skipping the homework

I see lots of examples of this but the one that really bugs me is the lack of attention to how success (or indeed failure) will be measured before the deal is signed.  Often, this is exacerbated by the need (?) for secrecy around the small team doing the deal.  The small team sometimes doesn’t have the breadth of knowledge or experience to put together a robust but flexible set of metrics for management of a contract over several years.  I always recommend including metrics experts or the team that will actual manage the contract in the negotiations AND allowing time for baseline and/or benchmarks metrics to be developed.

3. Fixating on price

Too often, I have to bite my lip to stop myself saying, “You get what you pay for.”  More seriously, it is important to think of the total profit of the vendor and savings by the client as a “Value Pie.”  If one side or the other is not getting enough of the Value Pie then the relationship will deteriorate.  It is important to develop metrics that reflect the whole Value Pie.

4. Using too few suppliers

More and more, I m coming to the conclusion that the best control for vendor management is having other options.  The simplest other option is to have another vendor ready to do the work. In a single vendor arrangement, There is a real risk of mutual dependence leading to distorted behavior leading to mutually assured destruction.  The single vendor option may seem to be the way to get the lowest price in the short term but it seems to be that any sort of monopoly is likely to become inefficient fairly quickly.

5. Only dealing with large vendors

Following on from my comments under #4, above, the best practice that I have seen is to separate potential vendors into  a few (2-4) groups e.g.  Large vendors, small vendors, Indian-based Vendors, Eastern European vendors, etc. This allows the RFP process to compare like vendors for value for money and, potentially, pick the right risk management mix of vendors based on size, politics, price, specialist capabilities or whatever.

6. Signing and forgetting

This is another symptom of having too few people involved in the original deal.  What it boils down to is that the IT team who have to make the agreement work on a day-to-day basis CANNOT ignore the vendors.  In outsourcing arrangements where I have seen the most heat and light generated, it is because a bubble of frustration builds up amongst the team as they try to operationalize an agreement that is not as strong or flexible as it should be. At some point, this bubble bursts over the small group who signed and forgot.

Back in the day, IT was synonymous with flexibility.  “Software” meant exactly that - the ability to implement things is a “soft” way that could be easily changed.  Nowadays, in many cases, software is as rigid as hardware or even more so.  Similarly, the IT group is too often the least flexible department in the organization - “You want what? When?”

Today, we find ourselves in times when businesses are looking to cut back.  Strong IT departments can flex their capacity and cost to meet their business needs.  Poor IT departments don’t have a plan for changing capacity quickly - they have a gas pedal but no brake.

I was reminded of this important quality of an IT department by a guest editorial by Edge Zarella in the ISACA Journal (Vol. 4 2009).  Edge neatly summarized the world that IT Professionals are facing:

• Intense pressure to cut technology costs
• Mandates to improve operational performance
• A need for technology to realign to meet new business needs
• A focus on cash management and a priority around liquidity

It was this last bullet that caught my eye.  It’s not enough for the IT department to claim they have “scaled back.”  They must be able to hand back cash.  How?  Much of the cost of IT departments is in the form of people.  Surely, head count is difficult to flex up and down  quickly, especially if the department is already running lean?  This is only true if there are no advance plans to flex size.  This has to be an important element of all outsourcing contracts and a key justification for starting an outsourcing engagement if you don’ t already have one.

How can we measure the strength of our IT department for flexibility like this?  How about the percentage of budget they could hand back given 3 months notice?

We can take this further.  As we climb out of recession of if a terrific business opportunity presents itself tomorrow, we can turn this IT flexibility metric around and measure how much capacity can be added given 3 months notice.

These metrics have to be supported with detailed plans that are reviewed and revised quarterly to reflect changing market conditions.

In the June 2009 edition of Baseline magazine, Rajesh Bhat and Jesse Decker have written a short and provocative article on how to partner for outsourcing success.  They start with a line that reminds me of the old saying that half of all spending on advertising is wasted - the trick is knowing which half!  Bhat and Decker’s version of this is, “Most experts agree that nearly half of all IT outsourcing (ITO) deals fail to satisfy the customer’s service improvement and cost reduction expectations.”

Of course, this is a loaded statement but it did make me read on so it worked!  Bhat and Decker argue that a robust ITO governance framework is essential to the success of a outsourcing deal to keep both parties focused on value and performance while providing predictable and transparent oversight.  I wholly support their view and would add two other words that nurture and sustain the best outsourcing partnerships: measurement  - which helps keep the focus on value and perfromance, and; trust - which emerges rom predictability and transparency.

Bhat and Decker list the following best practices:

• Start early
• Define success together
• Maintain a high profile
• Move beyond static service level agreements (SLAs)
• Pick your battles
• Understand the power of give-and-take over a long term relationship

Recent Surveys (reported in IEEE Computer, May 2009) suggest that US firms are changing their global perspective on IT outsourcing albeit slowly.  It seems that prices, tax regimes and new evidence of instability are putting companies off of more outsourcing to India at least for now.  They are taking an ambivalent view of China  - negative for similar reasons to India and positive because China is at least a reasonably mature alternative to India.

Two interesting points in the report: New outsourcing initiatives are less likely this year because of the start-up cost; and Idaho and Indiana are represented on the list of emerging outsourcing locations.  Hmmm!

While I am a true advocate for the benefits of process and continuous process improvement, there are times when its appropriate to recognize the truth in apparently heretical statements.  For example, I can all relate to a sentiment expressed by Joseph M. Hall and Eric Johnson in the March 2009 edition of the Harvard Business Review, “process standardization can undermine the very performance it’s meant to optimize.  Many process work best when they’re treated like artistic work rather than rigidly controlled.”  Hall and Johnson were not writing specifically about software although they identify software development as an area where there ideas might apply.

OK so its not that heretical.  Hall and Johnson still acknowledge that a process is required but they argue that sometimes that process needs to be artistic not scientific.  So when is that the right way to go?

Hall and Johnson suggest looking for scenarios where the inputs to the process are variable and, by extension, when that variability is not only uncontrollable but can be used to benefit the outcomes.  One example they provide is that no two pieces of wood provided to a carpenter are the same.  Despite our efforts to the contrary, I would still argue that no two sets of requirements are the same.  However, as we progress through the SDLC, the form of the deliverables from stage to stage become more similar and more controllable.  Hence, perhaps software design is an art but coding of that design can have a scientific process.  I will accept the argument that much of the coding done today is creative but argue back that this is because we continue to be weak at separating design from construction to the detriment of the quality of the code.

So what do we do if part of our process needs to be treated like an art?  The recommendation is to invest in giving employees the skills, judgment and cultural appreciation to excel in variable conditions.  Does that sound like an argument for outsourcing software design to you?

No, me neither.

Report from IAOP Global Outsourcing Summit 2009, Carlsbad, California #5

Satyam fraud - what could an outsourcing professional have done?

One session of the Summit included a great discussion with lots of audience participation on this topic.   There was some consensus around the fact that  we can probably all do more due diligence before and during the contract but that probably wouldn’t have helped in the case of Satyam.  Instead, it’s best to treat fraud as the same sort of “Act of God” for which you would have a Disaster Recovery Plan.  Of course, this extends to any unexpected event that causes of justifies a termination of an outsourcing contract.  You might have a great contract that allows you could to terminate but how, exactly, would you keep the delivery going? Without prior planning, a lot of knowledge could be quickly unavailable.

How many of us have Disaster Recovery Plans for our outsourcing?

Report from the IAOP Global Outsourcing Summit 2009 in Carlsbad, California #3

The Conference Board in CEO Challenge 2008-Financial Crisis Edition — features an analysis of the matched sample of responses of 190 CEOs, chairmen, and company presidents who participated in The Conference Board CEO Challenge 2008 survey fielded in July and August, and then took the time to fill out the survey a second time in October, following the recent economic downfall.  When asked to rate their greatest concerns from among 94 different challenges, the matched sample of chief executives participating in this year’s survey chose excellence of execution as their top challenge for the second year in a row.

But 46.7% of survey participants — up from roughly half that (24.5%) in the summer — were most concerned about speed, flexibility and adaptability to change. Global economic performance (44.6%) and financial risk including liquidity, volatility and credit risk (43.8%) were the fourth and fifth most pressing concerns, but were not in the Top 10 list of concerns in the summer survey. Business confidence also jumps into the Top 10, moving up 25 ranks from 34th out of 94 challenges to 9th. Those rating business confidence as being one of their “greatest concerns” rose four-fold from 9.1% in July/August to 36.3% in October.
Conference Board Report - Top 5 CEO Concerns Before and After Crisis

Reports from the IAOP Global Outsourcing Summit 2009 in Carlsbad, California #1:

Some interesting data points from todays sessions:

From a survey of IAOP members expectations:

Regarding existing outsourcing contracts, 43% of outsourcing will remain unchanged but 44% will have either renegotiate prices or volumes. 5% of outsourcing arrangements will be suspended or cancelled.

Regarding the number or amount of new outsourcing, 38% expect no change, 36% expect more and 27% expect less. Why? Most respondents said more savings (no change there) but an increased number said “more flexibility.” What are outsourcers going to be focused on? Greater due diligence, more off-shoring and fewer, larger providers. In other words, risk management.

A contact from a IT vendor management group in a major multi-national corporation told me that they have been instructed not to do new outsourcing contracts or renewals for longer than one year. As he said, “That takes away one of our few negotiating points for lower prices.” However, we then decided that in the current climate, it probably wont make any different and lower prices will still be available from providers worried about less work.

The Indian share of outsourcing is still the biggest but is declining. The diversity of countries doing outsourcing is increasing.

I was in Europe last week meeting with current and prospective clients.  It is interesting to see that the level of commitment to the use of function points for managing outsourcing contracts is growing.  Nobody believes that functions points are the whole answer  or that they should be implemented thoughtlessly but there is growing support for the idea that managing a software development or maintenance outsourcing contract is very hard indeed without some notion of software size.  The sizing methodology that can be most easily accepted by both parties to the outsourcing contract is function points. During the week, I was party to several conversations that illustrated how sophisticated the measurement conversation has become in SOME companies.  I was reminded that those companies who have not implemented strong software measurement plans are getting further and further behind.  Of course, they are probably not reading this blog either :-).