Don't sign any deal before reading this!

Not a particularly customer-friendly message. Yet that’s precisely what is implied by the behavior of so many integrator firms. “Drop $40 million in the corner of the room. Shut the door on your way out. You’ll see us again when we need more business. In the meantime, we’ll slide project artifacts under the door to you. If we encounter a problem, you’ll hear from our lawyers.” Of course, that’s the worst case. Integrators are a critical cog in today’s IT arena. This is especially true during the current IT labor “triple witching hour;” most IT shops now face economy-driven budget cuts, competition-driven demand for new systems, and a severe IT labor shortage.

Properly utilized, integrators can be a powerful complement to an IT organization. Integrators, disciplined by direct, IT-level competition, must constantly enhance their skills base with new tools and methods.

Also, integrators must be passionate about forecasting-type cost metrics. And they can utilize human capital management techniques (like offering market or partnership equity) to attract and retain quality folk. Finally, they can assume some degree of employment risk by leveraging employee expertise across clients.

IT groups, on the other hand, frequently under-invest in new tools, focus on “steady-state” types of cost metrics, and are constantly battling with corporate human resource groups about really important stuff like job titles and formal hiring strictures.

Projects typically are the “products” sold by integrators. By selling projects, integrators and clients can reasonably bound expectations, price points, deliverables, time frames, etc.

However, the project metaphor creates its own problems. A project emphasis tends to drive complexity up (bigger projects, bigger fees). It tends to emphasize artifacts delivered over knowledge transferred (“we code, we don’t operate!”). It minimizes opportunities for mid-course corrections (“this is what we agreed to, that is extra”). And, most importantly, it vests overall relationship control with the integrator (e.g., “shut the door on your way out”).

So, how should an organization manage integrators during these tough times:

Never let integrators “own” requirements. If you must employ a consultant to write requirements, make sure that they are working directly with your business folk – and try to utilize a different firm to implement the system.

Force integrators to break projects down into smaller, discreet chunks that allow for reasonable points of integration and mid-course corrections – and can set the stage for a new round of bidding, if the integrator is habitually messing up.

Work only with integrators that demonstrate an appreciation – a passion, even – for knowledge transfer into your organization.

Never let an integrator tell you that one of your people is screwing up, and that they have the perfect replacement in mind. A real test of knowledge transfer is if the integrator takes the steps necessary to make your people heroes. Under no circumstance, though, should an integrator be allowed to place a fox guard at the hen house door.

Unless the integrator truly possesses domain knowledge that you can’t otherwise access, assume that they’ll be 30% more expensive than your own people. After all, the integrator has to pay sales folks and return something to firm owners.

Never, EVER give up control of your infrastructure. Outsource the pieces that aren’t business critical, but maintain end-to-end manageability.

The test of an integrator’s capability shouldn’t be a demonstration of affinity with new and groovy tools, but rather the ability to utilize new tools to build systems characterized by ease-of-operations.

Assume the worst. Build in safeguards to protect your interests – and that of the integrator’s. Only then can a positive working relationship be established.

The integrator business model has been developing cracks for some time now. I’m amazed that the integrator industry, which has been touting the revolutionary qualities of e-business, CRM, etc., seemingly haven’t taken any steps to incorporate new productivity-enhancing business practices.

But think about this. Every industry, ever, sees product costs drop as buyer and seller experience appreciates and competition dominates. Except for the integrator industry. A company deploying an ERP system today is likely to pay double what a competitor paid for essentially the same system five years ago. This is an economic aberration that will not last.

Peter Burris is a Meta Group Fellow and independent analyst. He can be reached at peter.burris@twigresearch.com.