When all you have is a hammer, everything looks like a nail.

Whether you’re a carpenter or a consultant, choosing the right tool for the job is half the battle. I’m thinking a lot about this lately as we’re planning the redesign of a particularly complex system for a major client. Here’s the story.

Back in 2003-2004, my client installed a major upgrade of their Lawson accounting system. As part of the project, they replaced their home-grown billing system with Lawson’s project billing software. They were already using Lawson for general ledger, accounts payable and accounts receivable so this was basically a good idea. Using off the shelf software is almost always cheaper than building software yourself—just like buying an off the rack suit is almost always cheaper than getting a custom suit with equal quality fabric.

Unfortunately, the implementation team really only understood Lawson, their “hammer” as it were. They did find a way to meet the client’s requirement. That way, however, was very, very, very complicated. Every time the client sets up a single new project, the system requires the user to enter a total of 37 screens worth of data! And they enter hundreds of projects a year. We’ve automated the process so 98% of the time it works and the users don’t need to worry about those 37 screens. Still, there are too many times when something breaks and the fix is not at all obvious, delaying work and running up the maintenance cost. The client is expanding world-wide so we need to change our ways. It’s time to find a new tool.

My new tool in this case is just a touch of custom development. As I said above, building things yourself is generally more expensive than using a canned solution. Modifications cost money now and cost more money every time you need to upgrade in the future. But when keeping the package pure means pressing the “OK” button 37 times every time you create a new job, it’s time for a new approach. I think we can find a way to reduce the number of screens from 37 to about 10 of fewer.

Final lesson: Off-the-shelf is great as long as the processes you create can be understood clearly by your users. When a system is so complicated that it requires on-going fixes and expensive support, it’s time for different tool.

Take your hands off the keyboard and step away from the computer.

Many of us, present company included, feel we’re only during work if something is happening on the screen. Programmers are particularly afflicted by this disease. They love to be programming even if they don’t really know where they are going. So I often tell my guys, “take your hands off the keyboard and step away from the computer.” Let’s make sure we understand the problem before we start making changes.

In a larger sense, my recent vacation reminded me how important being “hands off” can be. For an entire week, I only checked my e-mail once a day and called no one. For a small business owner, this was a major victory. My mind actually cleared as I pretty much focused exclusively on swimming laps, making sand castles and deciding what to have for dinner. I came back with some much needed clarity about my business. I didn’t do any major thinking; I just clear my head. In my every day life, I work out and I try to meditate. I even get the occasional massage. But I almost never get the clear head feeling for more than a few minutes at a time.

Do you or people in your organization ever really get away from the computer? Is the Blackberry always in hand? Do you wonder what they’re doing when they’re not typing away, on the phone or in a meeting? Think about what would happen if they, and you, took their hands off the keyboard and stepped away from the computer. What might you find out.

I am the customer

I tend to repeat myself. All my employees have heard over and over again that “Our customers don’t care about technology.” For example:

• Customers don’t buy a Java or .net based accounting system. They want financial reports.
• Customers don’t want a windows server. They want a place to store their documents securely.
• Customers don’t care about open source. They want a cost effective solution

I am now a customer so my vendors need to understand this. We’re still a small company but we’re big enough that we’re a source of serious business for the right firm. I just spent $15,000 on hardware, software and a revised website in the last month and that’s not chump change for me or for my vendors. Unfortunately, because I’m “in IT”, people make the mistake that I care about technology. I don’t. I want to be a customer.

My website designer learned this the hard way. In tandem with the re–launch of this blog, I asked my long–time designer for a slight freshening and reorganization of the web site. He said it would take 3–4 weeks. It took two and a half months. He didn’t charge me more than the original budget but he spent a lot of extra time and wound up with a less then happy client. Here’s why.

I was buying a simple web site with a blog. He was selling me Movable type installed on my own hosted linux site with clean HTML code. He thought that was the best approach and I didn’t care, because again, I am the customer here and I don’t care about my website’s technology. I didn’t want to micromanage his technical decisions.

Unfortunately, he and his programmer ran into problems. My site was hosted on an old version of Linux with an old version of perl which meant that the sendmail feature wasn’t working correctly and we couldn’t upgrade movable type which meant that the comment feature wouldn’t register returning visitors correctly….. He tried to get my sympathy for all the effort he was putting in but I kept thinking that I just wanted a simple web site and a blog and that it shouldn’t be so hard. To make it worse, his programmer even suggested that I switch hosts without thinking about the aggravation that would cause with my e–mail and the further delays involved. The programmer was fixated on his beloved movable type.

I finally did my own research. Turns out that some of the biggest blog gods, including Seth Godin and Guy Kawasaki, use hosting services. I told my guy that if it’s good enough for them, its’ good enough for me. The blog, as you can see, is on WordPress. The design clearly needs to be improved but for now, I have what I wanted: a slightly revised site with a blog. My designer is somewhat wiser. And I am the customer.

When it’s out of scope, you’re out of luck.

Have you worked with the Not My Issue Consulting Group, the NMICG? You know the firm. Your legal team negotiated a fifty page contract in very small print to make sure that everyone knew EXACTLY what the firm was going to deliver. The only problem was that while NMICG had done projects like yours many times before—that’s why you hired them—this was your teams first time through such a project. So, now you’re finding that when you expect expertise from the firm, you hear, “that’s the clients responsibility.” Thanks guys.

As a career long consultant, I totally understand that scope control is the key to profitability in professional services. On occasion, though, I run into the NMICG and I just shake my head. Currently, we’re working on a software upgrade for a client. They contracted with a reputable large firm to do most of the work but have asked us for some additional help.

Some of the NMICG consultants were having problems running several of the upgrade programs. They had installed the software but the software was freezing when ever they started to run the programs. While my detailed technical expertise is not what it once was, it sounded like a problem we’d had before. I had one of my consultants, Phil, find the documentation we’d used and send our recommendations to the client. Problem solved. It’s great when you can look that smart with that little effort.

What irked the client was that NMICG said that those issues were not their responsibility. They installed the software but relied on the client for one piece, the database. I’m thinking to myself that, yes, the client was responsible but they do this kind of thing once in four years. NMICG does it all the time. Shouldn’t NMICG have at least pointed them in the right direction? Didn’t they know to double check the work? A person who had even checked the documentation would have found the problem in about 30 minutes.

My firm could not have handled this entire job successfully. It’s too large for us now. I also respect the NMICG consultants as, with this exception, they know what they’re talking about. The problem is that the firm defines responsibilities so narrowly that it defies common sense.

A lawyer can tell you who’s contractually responsible. A business person needs to know who reasonably might be able to get things done. The first is never a substitute for the ladder. Think about it when you negotiate your next large contract.

You can fight city hall.

I posted about the travails of my friend and colleague Stacey last week. (How may we not help you?) With her permission, I’d like to share an e-mail she sent in response.

I have to tell you that your blog actually helped me. On Friday I rebooted the server and it won’t load the OS. It would reboot and go to a diag screen. I tried to resolve the problem and then picked up the phone because I’m wasting too much time on a new machine. I told the vendor to replace the machine NOW! They asked that I call IBM service and I told them no; if this was a car, the dealership would take care of everything. They offered to get support on the phone for me to talk to and I said no, you are the dealership, take care of it. The vendor had a tech came out on Friday night and a second one on Saturday. The graphics card was replaced and the machine is now up and running. I ran diagnostics on it several times—all looks well. However, the vendor has been warned that if there are any problems with it this week it must be replaced. Thanks for the blog—it put a fire under me.

What can I say but you’re very welcome.

Breaking up is hard to do

Commitment is a scary thing. What if it doesn’t work out? What if I become dependent? What if it turns out they’re not what they seem to be? What if we break up? These are just a few of the questions that go through an executive’s mind when he or she’s about to sign an outsourcing agreement. You know that you can save money getting else someone to run the payroll, fix the PCs, or take out the trash. Still, while you’re attracted to the savings in money and hassle, remember that breaking up is often hard to do.

Indeed, the recent Satyam disaster should make these concerns front and center in your mind. In brief, this multi-billion dollar Indian outsourcing company stated assets of $1 billion that weren’t there. Their bad. The more interesting part is that their not just folding up shop. Their big time clients can’t just walk out the door. They’ve been saving money for years but in return they’ve lost a fair degree of independence. They can’t trust these people and they can’t just walk away or a key part of their business will fall apart. They have no obvious solution.

I’m an outsourcing advocate. I would never do payroll for my own company and have just outsourced my laptop support. It’s not that I don’t understand these areas. I’ve worked on HR/Payroll software for large companies but I just don’t see the benefit for anyone who doesn’t have several thousand employees. And while I can fix PC software, I think it’s a waste of time.

Still, when it gets closer to my own or my clients’ core business, I become wary. The savings are often obvious; the risks are not. When advising my clients, I always make sure they know how they’ll save their business if the deal goes sour. A good contract isn’t enough. I’m sure all those companies stuck with Satyam have top flight legal advice. What they don’t have is a practical way out.

So, if you’re thinking of outsourcing, remember to hope for the best but prepare for the worst. And accept my best wishes that you and your outsourcer will live happily ever after.

If it’s not on the list, it doesn’t exist.

No one ever likes to have a project go late or over budget. That’s especially true in tough economic times. I’m therefore going to give you my number one key to managing projects successfully. For all you project managers, I know I’m oversimplifying. For everyone else, I want you to learn my mantra:

If it’s not on the list, it doesn’t exist.

Say it three times. I’ll wait.

Let me explain. Consultants are rarely called in unless there’s a problem so I have a somewhat warped view of the universe. Still, I’m guessing that many of you are working on projects which are behind schedule and/or over budget. You see a lot of activity but no real progress. It can be any kind of project — a new reporting system, a software upgrade, or new POS registers in the field. In every case, you’re not sure who’s done what or where the hold up is. You ask people what’s going on and you get responses like these:

“Didn’t you read the e-mail?”
“He didn’t respond to my e-mail”
“I left her a voice mail. Haven’t heard back”
“Yeah, I guess he was supposed to talk to me last Friday. I’ll double check.”

If you’re a business person, you may think people are hiding some major technical problem. That’s almost never the case. People are the problem. People who think cc’ing an executive who gets 500 e-mails per day is effective communication. People who complain but won’t quantify. People who always have other things come up. For all these people, I have my mantra:

If it’s not on the list, it doesn’t exist.

Here’s how you solve the problem. You get every one in a room and you write down every single issue on the project. You take every issue from that software patch that the vendor hasn’t delivered to the training room you need to book and get it in “the list”. This list will be your bible. You assign dates and names. You review weekly. You let every one on the team know:

If it’s not on the list, it doesn’t exist.

Where you keep the list doesn’t matter. I like Excel but if you have the time and see the value of Microsoft Project, go ahead. People can still use e-mail, voice mail, or signal flags if they want to. They can even talk face to face. But each week, they need to know that problems only exist when they are on that list. Further, problems are only resolved when the list says so.

If it’s not on the list, it doesn’t exist.

It sounds simple and it is. Further, I acknowledge that there are certainly other reasons projects fail. But if you do this one thing, you’ll be surprised how many issues that have been open months or years will be closed in weeks. Try it and let me know how it turns out.

How may we not help you?

I’m generally pretty strict with my consultants. I tell them NOT to regale the customer with stories about all the nonsense we find when dealing with hardware and software suppliers. Our clients hire us to make problems go away and shouldn’t care about our troubles. Still, the occasional story can be important if only to remind business people that bad service is not only aggravating, but expensive.

Here’s what happened. We’re in the process of ordering some new hardware at one of my clients. As the company has grown, we’ve decided that we should have one set of hardware for daily work and another for testing and recovery in case of disaster. So, Stacey, the unix administrator who does an awesome job for me goes ahead and orders the new stuff from IBM. She specifies exactly what we need and we wait for arrival.

When it finally comes, she discovers that IBM didn’t fill the order correctly. Now, when I order something from Land’s End, I expect that they’ll send me a new sweater if they shipped the wrong size. Indeed, they’ll even take it back if it was the size I ordered but it doesn’t fit. Nothing so simple in technology.

Rather than replacing the part directly, IBM tells us how to fix their problem. It’s as if Lands End sent us a needle, thread, scissors and dye with “easy to follow instructions” to fix the problem of a bad sweater. This goes on for days, wasting Stacey’s time until it eventually gets fixed. Of course, neither Stacey nor I are surprised that this happened or how long it took to get the fix. Think about the expense involved. Stacey probably wasted 1-2 days on this. In New York, the total cost of an in house IT resource is about $75 per hour. So, for a $20,000 server, the client has spent and additional $600-$1200 in time. Ouch.

Of course, the sad part is that IBM used to be one of those vendors where you paid a premium price and received premium service. Now, they just seem to expect the premium price.

Buy a Crystal Ball

People trust software to help them avoid disasters. After the dot com crash, all public and many private companies tightened up internal controls in response to the infamous
Sarbanes-Oxley Section 404. Indeed, an entire category of software covering Governance, Risk and Compliance (GRC) grew rapidly to help companies insure that they have the proper controls. If you haven’t noticed, it didn’t help all that much. This CFO article asks why that’s so.

My exposure to Goverance and Compliance is limited. I’ve helped several clients make their system audit ready but I’ve never been an IT auditor and never want to be one. However, I have an opinion about the value of software in predicting the future. It’s not good. “R”isk software is all about making sure that what you’re doing today won’t come back to bite you tomorrow. I tell my clients time that software can be great at telling you what happened. If you’re sophisticated it may even help you identify trends. But it’s NOT good at predicting the future unless the future is similar to the past. Unless you’re selling bread or milk, chances are that your customers probably can do without you and there goes your model. Go buy a crystal ball. At least then you’ll know that you don’t have a real technical solution.

I learned this before I had my first software job. Back in the late 80’s/early 90’s I remember walking with my Dad through a shopping mall. He and his brothers had a chain of men’s clothing stores. It wasn’t a great year partially because of the economy but it was made even worse because sweaters, which had been hot sellers for years, pretty much fell off a cliff that year. Replacing a category that had made up 20%+ of sales isn’t easy no matter what a buying genius you may be.

Could software have solved the problem? They were fully automated. They knew day to day what was selling. Still, they had to make educated guesses months in advance about what their customers wanted and no software can do that. That’s still true today and part of the reason I didn’t go into retail. I’m not cut out to predict what irrational consumers are going to do 6-12 months in the future.

You need to remember that no matter how much you spend on software, you can’t escape this fundamental problem. The interface may be beautiful. The reports may satisfy auditors, executives, and investors alike. The software still needs to assume that tomorrow will be similar to today and that just isn’t so. No matter how you spent.

The end of PMD

What is PMD? Glad you asked. The first person who used the term PMD with me was the CIO at a client. PMD means pretty much done. In my business, you ask a programmer if some piece of software is finished. You’ll often hear “it’s pretty much done”. Then you ask “is it ready for testing” and you’ll get back “well, it’s not that done.” Then you ask, “Well, when will it be THAT done?” at which point they say, “I’ve got another 10% or so to go. Ask me in another week.” ARGH! our goal at Red Three is to eliminate PMD.

How can you eliminate PMD?

It’s not an easy thing. Technical projects have lots of pieces to them. It begins by dividing the work into as many little tasks so that you can check regularly and clearly that things are getting done. It requires having the right people, processes, and procedures. Most of all, it requires an ongoing dedication to continue to nailing things down every day.

We’ve been pretty good in the past at making sure that things get done. However, we’re growing quickly and I believe that ending PMD begins but doesn’t end with the CEO. Therefore, I’ve set my great team task to develop the right processes and procedures to insure that when we say something’s done, it’s really done. While I’m an advisor here, they clearly need to own the process as, more and more, they do all the actual work.

It’s clear to me that we’re not the only people interested in ending PMD so I’ll post updates as our process develops.