IT: The In-house or Outsourced Debate
Editorial by Bill Bittner, President, BWH Consulting
Many companies are struggling with their IT budgets. Some have literally chosen to give up on all or a portion of IT and turn it over to outside firms.
At the same time, IT specialists are practically, if not literally, becoming a-dime-a-dozen as high schools, technical schools, junior and four-year colleges, as well as graduate schools, turn out thousands of new workers with the skills necessary to operate, connect, and build computer applications.
The focus within IT has shifted in recent years to the design of service-oriented architectures (SOAP) that are enabling applications developed by different organizations to speak with one another through the web.
Much of the effort involved in today’s applications is centered on interoperability. In order to assure an application will run on every single platform, it has to be tested on various browsers, web and database servers, as well as operating systems. New applications must be able to communicate with other applications running in their own environment (the piece covered by SOAP). Application design has to be based on a developer’s view of the business processes and the challenge of overlapping functionality between various applications must be addressed.
What if, instead of trying to make all these disparate pieces work together, a company simply said, “Let’s design it the way we want it?”
Instead of buying an application that is capable of running anywhere and on any hardware, suppose companies wrote one for their business and the hardware they plan to use.
Instead of having to understand the terminology used by different groups of developers, applications would be designed to meet the company’s needs with its terms and processes in mind.
Instead of requiring developers to understand a variety of computer languages and design methodologies, companies could use a single methodology for their business. Software vendors may even begin to offer “par-developed” applications that leave the detailed user interface portion to the business user so they can better fit it to their environment.
Maybe this idea is not so farfetched. I actually had this thought before I read that Sainsbury, the large British retailer, has made this very same decision.
Moderator’s Comment: Is it time for companies that have outsourced IT to bring it back in-house? How big must a retailer be to consider customizing a
purchased application or building its own?
I really believe Sainsbury might be onto something by bringing its IT back in-house. I’m not advocating companies no longer buy applications, but rather
manage and modify them in-house.
You could look at applications the way you would a house or a car. The developer or manufacturer builds the basic unit (the application), but you use your
own interior designers or “motor pool” to customize and maintain it. As a result, you can have the benefit of both worlds. The bulk of the heavy lifting is done by the software
vendor to create the applications, but companies would use internal personnel to create a custom look and feel that fits its way of doing business. You don’t have to continue
paying maintenance contracts and, if you like someone else’s solution, in a few years you can always change. –
Bill Bittner – Moderator
I would keep it out, unless it is a core competency or core capability for the retailer. My past experience says internal IT people put on every bell and whistle and make sure it works for everything. And it takes twice as long to do it outside. Oh, and they can’t ever explain to you why they need to do all the adds, etc. So I say keep it simple and keep it outsourced. Pay them just for what you want it to do. In fact, you should look at outsourcing any work that is not core to the vision and mission of the organization. This is the way of the future!
It’s a question of ‘core competencies’ and efficiency. Whether a company is better off ‘in-sourcing’ may be determined by considering the requirements for maintaining a top-notch internal IT force vs. the main-mission of their business. Example: Does an auto parts company get better by having full-time employees in IT?
Today, the IT service community provides efficient teams of subject experts in software applications. Any good service company then works to ensure the team develops expertise in the client business to properly support the business mission. Done well, the outsource team should provide ‘seam-less’ support for the client (essentially the same as an internal team).
If outsource teams are less competent than full-time staff; or if they are more costly, then the case for taking IT back inside is clear. Otherwise, companies should focus on core competencies.
This is the age old controversy over make or buy, combined with the economic “wine and wheat” theory. IT long ago proved that specialization and outsourcing (when done correctly) maximizes the return for each dollar spent, while minimizing the TCO (total cost of ownership) as seen by the organization. With the rapid growth of both hardware and the compatibility needs of software (and the necessary interfaces) it is almost always better to purchase standardized software. This ensures that the product is robust, under continual development and has the necessary focus to address future technologies. Few organizations have the resources or competencies to create the software which addresses their needs, works well on future platforms and is safe and secure for their organization. The key to success in modern business is to focus on an organization’s core competencies and outsource to fill your organization’s needs.
Michael Richmond got it right, I think. IT professionals – one of whom is my son – usually have supersized egos and very high opinions of themselves. While good if channeled, who inside a supermarket company will have the requisite combination of management skills and computer development chops to earn the respect of the IT department? At some point up the chain of command, someone will be in charge of the IT folks but will not really know what they’re talking about. That’s when internal IT departments get out of control. Pride of authorship becomes paramount, egos emerge, and there’s no one to apply the brakes.
At Fleming we were required to complete a course entitled “Dealing With The Politics Of The Data Processing Department.” Seriously! Internal IT (or MIS) departments are “voodoo” departments. Most company managers don’t understand what IT does, and the IT department likes it that way. Working with outside suppliers adds some accountability to internal IT, and removes much of the NIH (Not Invented Here) syndrome that inevitably appears.
Supermarket companies are classically chauvinistic – not with “male chauvinism,” but just plain ol’ chauvinism as defined in the dictionary as “prejudiced belief in the superiority of one’s own group.” Within supermarket companies this is expressed as “we don’t do it that way here” and “we know our customers and that doesn’t apply to them.” If the supermarket way of doing things expanded to other industries, novelists and network newscasters would develop different versions of their work for all of the various U.S. geographic areas. A move by supermarkets to return to relying more on internal IT departments is greatly driven by this (misplaced and unrealistic) chauvinism. It’s a step backward.
To outsource or not is a company by company decision. It is also a “make versus buy” decision at its core.
When a firm chooses to completely outsource an operation it is generally because it is more efficient or less costly to outsource that it is to “make.” Ask why.
When a company decides to bring an outsourced function back into their business, that means it is more efficient or less costly to “make.”. Ask why.
When you look at this on a company by company basis you find very different and complex reasons. The most common reason to outsource an entire function is that it is not running well under current management. The most common reason for bringing the function back in-house is because the customer is not happy with what the supplier is delivering.
Accenture says it doesn’t think loss of the Sainsbury contract will have a “material impact” on its financial results… Well what about the $500 million charge that Sainsbury suffered?
The issue here isn’t about in-house or outsource – it’s about what happens when the consulting contract becomes so large and complex that it causes all parties involved to lose their way.
Fast-moving consumer goods companies need fast-moving information systems that support quick, accurate decision-making on a daily basis. That requires IT investment and a continuous commitment to managing the program. For large firms, it seems inevitable that this requires a blend of outsourced and in-house competencies.
Supermarkets are not in the technology-creation business. They are in the business of efficiently moving goods and information about those goods with the goal of maximizing profitable sales to end-users of those goods. To get the job done, they must acquire and manage just enough technology to automate routine activities and to support decision-making.
Reading between the lines of the news item, it seems like Sainsbury may have gotten swept up in utopian thinking – led by its trusted consultants.
Outsourcing and ageism is influencing kids to stay out of computer related careers. Spending more money on math and science education will not result in more graduates. These kids have seen what the dot com bust did to their parent’s lives. Who wants a career that ends at something less than 40?
Requirements elicitation practices must change if IT is to provide value. Past practices emerged from an environment where costs drove a focus on development efficiency, not operational efficiency.
The bells and whistles end up in application, because programmers abstract away from the requirements. All media involves the beating of the media with the message. The message can’t help, but ends up constrained by the implementation media, in this case software. This means that the automated domain is mixed with the automating domain.
To control costs around implementation, use Agile practices to ensure that you get exactly what you want. Most shops are not Agile shops. Agile shops deliver requirements in very small packages. You don’t have to wait for the whole application to be built only to find yourself surprised.
Agile will also help you to get the requirements correct. But, the requirements elicitation process as it is doesn’t deliver correct requirements. It delivers development efficient requirements. The mindset of the requirements analyst is one of the problems. The insistence on a single definition of concepts is another.
Outsource is a managerial focus issue, not a cost control issue. Any cost reductions are secondary to regaining and focusing managerial time and effort on the offer. If IT is mission critical, it can’t be outsourced. If IT is in the offer, it can’t be outsourced. And, if IT is outsourced, you shouldn’t be managing it. You should be buying it. If you insist on dipping into what should be a black box, something you have no visibility into, rather than just dealing with the interface, you will not save any money outsourcing. You have to push the outsourced elements out into the value chain and manage them with your supply chain management system. If you haven’t systematized your supply chain management system, don’t expect to get all the cost savings you are looking for. And, while you are fixing your supply chain management system, realize that you have to outsource more than IT to make that investment pay off, because it is going to eat into your organization’s managerial focus.
There is no right answer to outsourcing or not. You have to outsource the right components at the right granularity.
Shaw’s Supermarkets had this as a long practice – design your own platforms, modify them as needed, keep all IT in house. Everything ‘turned on a dime’, which is almost everything in retail. Only downside is that some programs depended on one or two people if something went awry. But pride in work plus company loyalty generally had the programmers/support available as needed. That was while Sainsbury owned Shaw’s (wonder if that’s where Sainsbury got this best practice idea?). Things are different now with a different parent company… large expensive pre-packaged programs, IT partially outsourced, programs not as ‘pinpoint’ responsive and the company does not ‘turn on a dime’ as they once did. Too much time spent on integration failure rollouts instead of internal IT development of more efficient company software response.
It will be interesting to see if Sainsbury can figure out how to implement what worked very well for one of their chains and adapt that to a much larger environment. Bet they save money in the long run if they are successful.
The choice of outsourcing alternatives and structures is so wide, and the expertise so commonly available, that there is no reason I can see to bring IT in-house, once it’s outsourced already. Generally, the best systems people work for systems-driven companies, just as the best lawyers generally work for law firms, the best physicians work in practice groups or teaching hospitals, the best logistics people work for logistics-driven firms, etc. Yes, there are many exceptions, but why fight the trend? If a retailer has a great in-house systems department, the advantage of outsourcing might be minimal or a negative, but once the group is outsourced, if there are satisfaction problems, they can be worked out or another outsourcer can be easily located. Much depends on the written contract standards and the how the culture of the outsourcer meshes with the retailer’s culture. I’ve negotiated and administered outsourcing relationships totaling over 9 figures. The only major regret I’ve had is that I didn’t do more.
Outsourcing means losing all chance of genuine understanding and loyalty, handing strategy over to people who don’t know your business or its objectives or customers, and paying for the privilege. Keeping IT in house means accepting the need for training and communication, getting all departments on board and cooperating with one another. Not an easy task, as we well know. But, in the long run, one that makes more sense than sending bast quantities of sensitive information out to people who may or may not have your best interests at heart when handling it.
I have seen a few medium to small supermarket companies financially injured because they spent millions on IT systems and programs and did not know how to use them. I’ve even seen VPs in charge of IT who needed their secretary to help them turn on their computer. One time at a former employer, we did an employee satisfaction survey and found that the most disgruntled employees were in IT. Back in the 1990s, they were job hopping from one company to another. Because of the superior knowledge over their supervisors, they pretty much got to make their own rules, set their own hours, disregarded the dress code, etc.
Hiring outsiders to do IT can be risky. Often I have seen them tell management that the problems were big and would require thousands of dollars of work when, in reality, they weren’t. But senior management had no choice because they barely knew how to turn on a computer anyway. I’ve also seen other departments hire IT consultants and spend thousands of dollars over several weeks to prepare and manage identical projects that other departments in the same company were doing internally in a matter of hours.
A good in-house IT department that understands the needs of all departments is the way to go, in my opinion. And it is also important that the VP in charge actually has some real knowledge of IT.
I would still be more aligned towards getting IT Outsourced because you don’t really have to know how your engine works to reach destinations. The important thing is to know whether your system works or not. By depending on outsourced services, I would still be safegaurded with the Guarantees / Warrantees that come along with service agreements. Without outsourced services, I have to be sure about my own capabilities, which in this case is IT; not the primary tool in my business.
To bring IT in-house will only limit the scale of developments to the company’s present foresight. An in-house developed application may not be able to comprehend the possibilities that are unforeseen. It makes sense to pay for that extra layer to make it complete and robust.
But then, at the end of the day, it really comes down to…How do I save more $’s? If the outsourced service is something that’s eating up more and more with time, I guess only on those circumstances should the company look for ways to in-house operations. Even in that case, I think, the process of setting up in-house capabilities be outsourced before finally managing it in-house completely.
As we all know, it takes two hands of our body to clap, but each hand works differently to other and not all are ambidextrous.
Outsource the standardized, the normed, the cost, the non-differentiators. Insource the differentiators, the proprietary.
Outsource not for cost, but to preserve managerial focus. So if you outsource, really outsource. Don’t think about managing it. Let the vendor run their operation and expend their managerial focus. But, manage their performance, not how they do it. That’s their business. They should be doing it better than you could, because their managerial focus is there and yours should be elsewhere.
When looking at IT practices, the change in economics, in cost, has not changed those practices. In what ways does cheaper IT translate into better IT beyond cost? Start with requirements volatility. If you didn’t have to agree on requirements and every functional group could retain their terminology, models, and meaning–their culture, then requirements volatility would go away, and your applications could be delivered quicker and work more intuitively. Agreement is usually politically enforced. The need for a powerful executive sponsor originates in the need for power to make and enforce agreements. But, requirements volatility is what happens after power. So cheaper IT could mean more than just cheaper. It could actually mean cheaper and better. But, to get there, you have to buck the industry practice, outlook and culture.