Consulting objectivity

The bottom line on bias is exactly this: a consultant can only truly be unbiased about a choice between two technologies if neither he nor the client have ever used either one.

Suppose your company wanted to hire a consultant to advise on appropriate computing infrastructure strategies for the next five years. Suppose further that you're down to two choices distinguished mainly by their answers to the question: are you biased toward one technology over its competitors?

One says that he isn't: that he has a wide range of applicable experience but reviews client requirements without technology preferences. The other says that he generally prefers Sun's SPARC/Solaris combination with Sun Ray desktops, but bases client recommendations on what he thinks will work best for them.

How do you interpret their positions, and what should they mean for your selection recommendation?

Here's the first bit of a two part answer by frequent contributor Dave.Leigh - with the quotation coming from last Thursday's blog entry outlining the basics of the SAB (Solid American Brass) information architecture challenge:

Finding the right software

"Now, as you can imagine, my first reaction to the IT challenge is going to be based on finding the right ERP/SCM software to run this business on Solaris with Sun Rays and Mac laptops..."

Of course, since this is a brand new business with no established infrastructure, you should determine the business processes that will realize their goals, and finding the right software in which to implement those processes. Then find out what hardware you need to run it. Of course these things are to a degree interdependent, and some iterations and adjustments are necessary, but those are the priorities.

Before any of this legwork, you've already decided to sell them Sun Rays and Macs. You're not working for the customer, you're working for the vendors. If I were the manager I'd fire you and take on another consultant who was better at sublimating his biases in favor of my requirements.

He might also suggest Sun Rays and Macs, but I'd have confidence that it was to my benefit.

And my immediate answer:

I see it differently..

First I'm asking the question: stating a bias and saying that's my first reaction, not the necessary decision. People who really are working for the vendor assume the vendor's environment and never ask the question.

Second: bias is often indistinguishable from experience. If you hire a consultant who claims to be unbiased, he's either lying or incompetent -and either way you don't want him. Why? because it's impossible to work with two technologies and not form opinions about which is better. Thus the guy you want is one who says .. my first reaction is X, but now lets discuss options... because that combines experience with openness.

In retrospect, that was only part of the answer - a more important part would have been to recognise that in a new business the requirements aren't known; so, as I said yesterday, the right software is going to be a package incorporating lots of useless functions because that shotgun approach gives us the best chance of having what we need when we discover we need it. On a personal basis I would now point out, too, that most consultants rule out the Mac/Sun Ray combination because their expectations of client bias against that solution mean that they have no applicable experience of their own - in other words, they ignore the option they don't know because theirs is a self fulfilling bias on selling what's selling rather than what the client needs.

The bottom line on bias is exactly this: a consultant can only truly be unbiased about a choice between two technologies if neither he nor the client have ever used either one.

Look at Dave's second response and I think you'll see the chickens and eggs co-populating this issue - here it is:

Biases aren't for public consumption. Not at that stage of the game. Certainly, when I walk into a consulting gig I lean heavily on my biases as well. But that's not for discussion with the clients until I've done the legwork to let them know that their unique needs supercede my preconceptions.

Note that it doesn't matter whether their needs are actually unique. They might be just like the last ten clients I had, but I don't know that until I've asked the right questions. At the most I may say, "Yeah, I've got a few ideas, but let me ask you a few things about your situation first." Then at least I know whether I'm likely to be giving them solid advice or whether I'll have to backtrack and undermine my own credibility. He may not have even given thought to these prerequisites yet. In which case he needed me to bring them to his attention.

And though "my own credibility" sounds self-serving, it's only a fallout of doing your best to make the right recommendations to your customers. I say this so often it may as well be my values statement: "It is always in your best interest to work in the customer's best interest."

I happen to think that being good consultant requires me to sublimate my biases early so that I can -- as objectively as possible -- evaluate my customer's situation. THEN I can put my experience into play. Hell, I might conclude at that point that I'm not the guy for the job! If so, I'll recommend the right guy if I can and move on to another gig that doesn't waste the customer's time and money and ruin my cred.

If you can't cultivate a "shrewd naivete" (by which I mean an open-mindedness tempered with practical intelligence) in the requirements gathering stage of a project, you're almost as big a risk as the guy with no experience at all. I say "almost" because you might actually get something to work after a fashion, though it may be over-budget and late.

Good consultants are like scouts or guides in the Old West. Their customers look to them to recommend the right path. But you need to know something about the wagon train and their destination before you can do that, and you need to be aware of changes since the last time you passed through this neck o' the woods.

I think he does a great job here - clearly stating the need for objectivity in the consulting approach and contemplating the issue of what happens when the client's needs don't match what the consultants are selling. As a consultant I've done exactly what he recommends: say nothing upfront, review the client's business strategy and available skills, and then consider the client's skills as a co-determinant in making my recommendations - but I've never been happy about it because nine times out of ten the only recommendation the client will accept is to continue doing what ever he's been doing - and more than half the time that's simply wrong.

So my bottom line is simple: I'd hire the guy who genuinely has experience with all of the major systems available and says, right up front, "look, nine times out of ten I recommend the Unix solution - why? because it's usually both better and cheaper - so now lets see what makes you different..."