The Saga Continues….. (or Open Source, What a Concept!)

Things are starting to calm down, at least a little bit. Perfect timing. My owners came to me a day or so ago and said “OK, enough with this new system, we have got to get down to business”. Well put. I have personally done NOTHING else for the last two months but this project. It is not like my regular duties just went on hold while all of this has gone on. Monday I hope to start sliding my IT hat off and put my manager’s hat on.

Our company is pretty well all over the map. In some ways, it defies description. We are a direct marketer…no, we are a face to face sales organization…no, we are a light manufacturing firm…no, we are a retail store… no, we are an eCommerce player…no, we are…what? This is partly due to ownership’s attitude that we must be all things to all people in our marketspace. It is arguable that this is a pretty dangerous strategy. So, when it comes to implementing an ERP system, there are inevitably some compromises. Our new system is definitely targeted to the multi-channel direct marketer. We do some light manufacturing, but really it is to feed our marketing channels with products. There is no denying that the products we manufacture or assemble ourselves offer very attractive margins. Managing this element of the business with the new system is a case of trying to shove a square peg in a very round hole.

We finally waved the white flag on this issue in the last couple of days. We have decided to create a new business unit that will handle our manufacturing and assembly operations. Basically, if a product requires any processing before it is ready for sale, it will be handled by this unit. Of course, this excludes packaging, bar-coding and the like. That operation will still be handled by our WMS group.

This new unit will run on its own ERP platform.  In the wide world of manufacturing, its needs are pretty simple. We started our thinking in the traditional “Let’s se what is out there, do some cost studies and get some demos” approach. Once again, our new Open Source advocate offered us an interesting solution. Riding on the success of vTiger, we were much more open to this idea. Certainly at the level we intended to operate at, the price was right…FREE.

What he proposed was a web-based application known as Web ERP. At first blush, it looks pretty rough around the edges. We looked at some other ERP systems out there. We wanted it to have a web-based interface. We would have preferred it use MySQL as the underlying database. We needed it to have a fully functional accounting element. I looked at quite a few that fit the bill, but I kept coming back to Web ERP. Certainly cost was a key factor. You can imagine that this last project had eaten up my budget.

We will see. I have given a tentative agreement to create a test environment with this application.

I’ll keep you posted.

Going….Going…Gone

The vendor’s on-site support team left today. Sniff. It has been most excellent to have their experience right on hand, available to answer questions, put in fixes and have a more than direct line to the development and hosting departments back at their home base. Did I mention that this system is virtually hosted? What we have on site, as far as the application is concerned, is a fancy router and two bonded, redundant T-1′s. Everything else is at the vendor’s data center.

This is a total departure from the models of all our other mission critical applications. Everything else has been the traditional local servers and direct access by the vendor for support. This application is the mirror of that. We have total access via the above mentioned T-1′s and they have all the servers on their site. While we were going into a brave new world with this configuration, once we got past the “we have always done it this way” thinking, it really made a lot of sense. This application runs on about six different servers and climbing. For us to maintain six servers, keep them patched up, backed up and configured, we would have had to hire at least on full time engineer. With this virtual hosting, we pay a simple (and quite small) percentage of the gross run through this system and they do EVERYTHING else. Our cost studies showed that the new setup was slightly more than our current costs, with no significant capital investment. If we protect our connectivity, we should be fine. We have really loaded the system these last two weeks and my biggest fear, latency, did not occur.

I will say that much to our surprise, we learned we were the largest number of users they have ever run in this environment. A couple of glitches, the worst of which was about a four hour downtime did present themselves. The vendor stepped up, doubled the number of Citrix servers at our disposal and life has been good since. Between the massive downloads / uploads we had during go-live and our 70 users hammering away, I am going to make an educated guess that we have done a pretty good job of stress testing the configuration. Famous last words? I hope not.

Anyway, the team left us today. I am sure they breathed a sigh of relief when they pulled out of the parking lot. It has not been easy for them. For every two strengths the new system has, there is one weakness (or at least a compromise). It is in the long term that this system should really show its benefits. Of course, the family that owns the business are pretty short term goal oriented, so this has been a struggle.

But now we are up and running. So we bid a fond farewell to Tammy the Round, The Big “C”, Crystal Clear, Jeff the Patient and Anita Can Do It Remotely. We wish them well.

Still Standing (but limping a bit…)

Hard to believe that it is the end of Wednesday already. Seems like I just walked through the door on Monday. To say it has been hectic is an understatement. I have been running all over the place. (And for a guy with a bum hip and a rainy day, that ain’t no fun) But, in the main, it has been the types of issues you would expect. Still a bit of nervousness, still of bit of denial, still a lot of newness. The trick is putting one foot in front of the other, convey a sense of control and solve the issues.

It’s human nature for everyone to think that their particular problem is the most important. This is where the issues management system, vTiger, really shines. Yes, everyone expects their tickets to be solved immediately, but with my core IT Team and the vendor’s excellent staff, we can all hop on the system, pick out the “low hanging fruit” and then concentrate of the really thorny ones.

Our main concerns were down to four major issues. We actually got through three of them today and the fourth may be solved early tomorrow. Hurray for our side. Many of the other issues fall into two categories:

1. Permissions and instruction, meaning that the system will do what they want, but they either forgot how to do it or they don’t have the proper access to do it.
2. Things that really ain’t gonna change, at least in this release. We have created a category called “Wishlist”. There are audible groans when someone’s ticket has been “wishlisted”. But there it is, it just won’t work the way they want it to.

I have come up against some real lack of understanding on how software works. Requests like “Well, can’t they just (move, delete, change, display) this a little bit?” Sigh, sometimes it is but most times it is not a trivial change. I grow fairly tired of getting ping-ponged about between our senior managers and the vendor. I feel like I am “Between the Devil and the Deep Blue Sea”

We have started to spend some time in the actual data structures. Fairly clever, definitely database programming 2.0 (or whatever they are calling it). Never the less, we have developed a strategy that should allow us to get to whatever we want. Complex as hell though.

Tonight, I am going to have fun testing out  rather significant patch the vendor wants to roll out. While it has apparently passed Q/A on their side, they have (once again) put the onus on us to sign off on it.

Go-Live +7 days…

Mission Accomplished….

One of my team said that today. The only image I could conjure up was Former President George Bush standing on the aircraft carrier in his snappy flight suit declaring that as far as the Iraq war was concerned, it was “Mission Accomplished”. Didn’t work out too well for him. I think we are in the same place.

There is no question we are up and running on the new system. There ain’t no going back now. We accomplished a hell of a lot in the last week. We are doing business on the new system. And the new system is finally starting to show some of the promise that I saw when I originally was demonstrated it over two years ago. It is but a faint glimmer today, but I really think we are on the downside of the slope.

Speaking of slopes, very early on we discussed the concept of the issues slope. In a perfect world, it would look like this:

The Black vertical line is when the Test Environment becomes available, the Red vertical line is when we cut to the Production environment and the graph is the number of issues per day. In this perfect world we would have addressed most of the issues before go-live since everyone would be diligently testing the system and finding all their concerns.

In reality, it looks more like this:

Same scheme. The Test environment is available late. OK, lots of reasons for that, not the least of which is the problems with the product feed mentioned in an earlier post. Never the less the Test environment is available in increasing stages of completeness for over two months. In spite of a Herculean effort to set up scenarios, answer questions and emphasize the importance of testing the system, many key players do not. As a result, what are apparently key issues are neither identified or resolved before go-live.  Two days AFTER go-live, we start seeing issues spiking way up.

Sigh.

I wish I could say that I am totally blame-free in this situation. To a certain extent that is true. Should I have pushed my peers harder? Should I have put in their faces the issues that I knew would occur? I guess that is a hopeless exercise. We are where we are.

For the first time in over a month, I am going to take the weekend off. I think we all need a chance to recharge our batteries. I am sure I will not be able to resist the temptation to fiddle around in the system, but it will be in an ad-hoc, exploratory way. Not the crushing, deadline mania things I have been doing.

We learned a lot from the last time. We knew what we didn’t know. That is powerful.

What are the lessons learned so far?

• Be sure you completely understand the scope of the work the vendor expects you to do. We were frankly stunned when we learned just how much was expected of us. And we learned it just prior to the project launch.
• Resist the temptation to press on even if some deadlines are not met. If you lose time, you will NEVER get it back.
• Think small. As a senior manager, I am inclined to think big-picture or birds-eye. You must analyze every single daily process that occurs in the business. You may have made decisions years ago about a particular process or procedure and the reason was lost to history. You might lose some of the benefits of the new system if you model your procedures based on “Because we have always done it that way”. As painful and time consuming as the exercise is, the potential upside is significant.
• Communication, communication, communication. Keep as many people in the loop as you can handle. Make everyone a stakeholder. Do not ever give them the impression that people at almost any level do not have some sort of voice in the process.
• Do not rely on preconceived notions of how given individuals will participate. I was totally surprised that my ideas of who would embrace the system and who would be a handful were, in some cases anyway, upside down. Yes, some of the players I knew would be a handful certainly were. But there were just as many that I was pleasantly surprised at their commitment.
• Test, test, test. You must get something very close to the final production environment available to users as soon as possible. Create specific scenarios for each department to execute. The closer you can offer what their regular workday will be like, the better the chance of finding issues and procedures that need to be addressed. Once you find them, test again. If I were to repeat this bullet four more times, it would still not convey the importance of this.
• Don’t let senior managers “get away” with not actively participating in this testing. This burned me badly and caused much unnecessary stress in the final go-live.
• Be sure that EVERY customer facing operation is completely dealt with for the actual go-live. There may be delays that cause orders to get queued up. Be prepared with backup methods of capturing and entering these orders as quickly as possible. It is amazing, for example, how having to key in each web order that used to flow into the system automatically is a huge time sink.
• Defuse any negative energy as quickly as possible. Don’t let managers validate the naysayers. They must aggressively keep the process upbeat and positive.

Well, this post is long enough. There will be more as we fight our way through the final tweaking of the system.

Go-Live (complete) +1 day….

Houston we have a problem….

Well, we finally made it. As of this writing, we are 100% on the new platform. It was not without  struggle, though. The political dynamics of this business can be daunting at times. Anytime you work with a closely held, family company that has many family members actively involved in the business, there are bound to be issues. I am also a “shirt-tail relative”, so I probably fall into this category as well. Still, I witnessed a discussion today between the Co-President of the company and the Vice-President of Marketing that would absolutely curl your hair. C’est la vie. It is what it is.

Still, we were on hold with the website over issues that impacted perhaps 5% if our users. This meant that 95% of our users were still getting substandard service with the jury-rigged method we were using to process orders from our old website while we tried to sort these out. Sigh. In the end, we finally reached a consensus to bring up the new site.

Other than that, it was really a matter of managing issues. Nothing major. If fact, today was mostly spent on procedural issues, rather than major operational issues. I am not surprised by this. I knew going in that this platform was going to cause us to re-examine almost every aspect of our daily operations. There is no change without trauma. I have tried to be as diplomatic as possible and open up all discussions to all parties involved. There are times, however, when in order to escape getting mired down in endless discussion, there comes a point that it has to be said “This is how we are going to do this. Please make your plans accordingly”. Some of my team has to learn this lesson a bit better. The latter is at the end of the discussion, not the beginning.

Shucks, just checked on the website and the thing that advises them to recreate their password is missing.I will hear about that.

At any rate, I think we are going to make it. The Vendor’s team has been outstanding. Our team has been terrific.  My team has been above and beyond. It just goes to show that if we all work together with a little honesty, openmindedness and willingness, great things can happen.

Now we can begin focusing on the great part of this system. The slick operational tweaks and reporting that will allow us to fine-tune our company like a fine Swiss watch.

I will probably post a couple more times here specifically on this subject, especially if anything “interesting” happens.

Full Go-Live Zero Hour….Now!

We are holding for launch….

We were supposed to go-live today, but didn’t. Really not a catastrophe, just not quite ready. Given our last experience, I think everyone is a bit (no, a LOT) more cautious.

I am not too upset by this. Ultimately, I am the one who made the decision. My biggest concern now is that customer facing operations have been shutdown for almost five days. Also, anyone who is nervous about this adventure now has one more day to think about it. So much of this psychological element is solved by just getting them busy.

The vendor invasion happened today. We now have a total of five people on site from the vendor. Not just warm bodies, either. Each can act as front line support but also has a specialty that fits with our project. Looks like a pretty good crew. For sure, it is going to take a huge load off of my core staff and I. Well, tomorrow will be the test. We were still in a holding pattern today.

The website was a bit of a disappointment. We previewed the site on Friday in the test environment and were pleased, as I said in an earlier post. Unfortunately, what we saw today on the live site was a step backwards. Probably some easy fixes, but broken image links and suspect functionality did not exactly inspire us. So, it was that coupled with a late start at getting regular operations running that caused us to push. The upside was that it gave us some more time to test…test…test.

Zero more days to go (I had hoped to say Go-Live +1) Sigh.

On your mark, get set……

This is it. Tomorrow is the big day. Our last day before going live was pretty calm. Sure we had our share of glitches, access issues, balky hardware and last minute gotchas, but by and large, we seem pretty close. The first few hours tomorrow will be hectic, to say the least. Confidence is high.

I am really tired. (Just feel asleep at the keyboard for a minute). But it is a “good” tired. We are as ready as we will ever be.

I am fortunate that I am surrounded by good people, motivated people, people who can and will get this job done.

Enough for now.

Less than 12 hours to go….

Sad day, but Cautious Optimism

I learned of the death of a friend today. Long battle with cancer finally caught up with him. This guy was really a mentor for me in my life as a Sports Timing producer. We used to be the agents for the Sports Division of a big watch company in Switzerland. You saw their name flash at the end of every race at the recent Beijing Olympics. My friend sort of took me under his wing when I knew nothing about the business and guided me to some success. A bit of a scoundrel, too. I am sorry I lost touch with him in the intervening years.

I have finally spent some quality time inside the application itself today. I remember now why we made this decision in the first place. It is an awesome tool. Slick interface with an incredible amount of information in front of the people who need it. Outstanding communication tools. Powerful CRM and Marketing applications.

As I mentioned in an earlier post, there are a ton of configuration issues. Nothing that is a show stopper, but there is a huge gap between what is just OK for Go-Live and what we really want the system to do. Well, first things first.

I keep running mental (and digital) checklists over and over, trying to figure out what I know we must have missed. I would like to say that we are covered, but look at our collapse on Friday.

Monday is the last day for us to get ready. I already have a curve ball. Our current website, which was supposed to have the communication severed to our servers so it would just queue up any orders taken, has apparently been left up. That means that any orders taken from Friday at noon until sometime on Monday when we do get it fixed have been cheerfully posted into the old system. We will have to determine which orders they are and jury-rig a data entry procedure to get them in the old system. Sigh.

Still, I think we all have done a good job. My team, the department heads and the rest of the employees on our side and the team from the vendor have done yeoman’s labor to get this thing right. As far as I am concerned, I am in a pretty good place. It was a “bet the farm” move, especially in these troubled economic times. When we first started working on this thing, the financial picture was quite rosy. I don’t think we have suffered more or less than many businesses like ours, but that is really not saying much. It is still pretty grim.

“I think I can, I think I can” The Little Engine That Could, 1910

2 more days to go…