Lawson’s New Office

April 22, 2009

If you read this blog at all, you know that I hammer over and over again at a single theme: the middle-aged application. What is a middle-aged application? It’s one whose basic mode of interacting with data was invented almost 20 years ago. It stores data in tables; you interact with it by pushing data into those tables or taking the data out. It was a great model, 20 years ago. Today, in my view, it’s fading.

What do you do if you’re a vendor, and you have an app that was built on a paradigm invented 20 years ago? Well, one thing you can try is a new interface. That way, even though you’re still basically storing data in tables, you’re doing it in a way that’s brighter, more colorful, and possibly more useful.

Unfortunately, this is kind of an obvious idea, and the landscape of ERP systems is littered with “new” interfaces that turned out to be not so new after all. According to the vendor, every single one of these modernized the application completely. But all they really did was put lipstick on a pig–and not much lipstick on an exceptionally hairy pig.

The latest such attempt (and surely one of the most creditable) is an interface demo’ed by Lawson at its recent CUE conference. I got a chance to take a look at it (if not a look into it), and I would have liked nothing more than to say that it changes something important about the application.

Unfortunately, that’s not possible, but not for the reason you’d expect. It isn’t that I looked at it and saw, “Oh, it’s just like all the others.” It’s that Lawson made it impossible for me to tell one way or the other. This is not a good thing, of course; if Lawson really has something great, they should show it to people. But it does say something important about the software industry.

First, let me tell you what I know. Lawson told us at the conference that they’re offering two new things that might change the way users work with the application. The first is a search function, which allows you to search for records via their contents, rather than getting at them through the menu system. The idea of this is really good; if you had to name just one thing that makes the old ERP apps middle-aged, it’s the fact that you can’t Google through them in any way, shape, or form. The second is something called Smart Office, which puts a Microsoft Office-like skin on Lawson and lets you interact with Lawson through Microsoft Office products, like Excel and Outlook.

The search function is particularly exciting, because the problems with search are built into the way 20-year-old ERP systems work. ERP systems store their records in a database; to get the records out, you have to use the tools the database provides. Search isn’t one of those tools. (It is too slow.) Lawson gets around this problem by manages pushing all the records in the database out to an appliance, a separate virtual machine, and then using a tool that wouldn’t work on a database, the open source search-engine, Lucene.

From what I could tell, there are some really thoughtful things in the way they’ve implemented the search. Particularly impressive was the fact that you can search back on all the work you’ve done, which makes it easier for you to stop in the middle now, go back to something when somebody asks you about it, or correct mistakes.

Smart Office, alas, seems to be a little more ho-hum. Some of what SmartOffice does was actually built into PeopleSoft more than 10 years ago; much also looks like a knock-off of Duet, the much-vaunted SAP attempt to integrate with Microsoft Office. Still, when you take a quick look at it, it doesn’t seem too bad.

If either product had been introduced ten years ago, I think I would have left my reporting at that. Lawson had come up with something new in the interface and demo’ed it. It wasn’t really very clear exactly what. But who cares. One needed to give them the benefit of the doubt. Surely it was better than a green screen.

Ten years later, though, I don’t think that’s right (even though that’s what most of the analysts did). Ten years later, there’s been a lot of experience with “major” interface improvements, and, as I said above, it’s been a disappointing experience. Most of these improvements have been at best, an injection of Botox, doing nothing about the basic problems with an aging interface, and possibly making things worse.

What this means for Lawson is that the burden of proof has changed. If they’re going to come up with a change in the way users get at data and claim that it’s a compelling, potentially transformative difference, they’re going to have to prove it, some way or another.

Certainly, if the language of the demos is any guide, Lawson does think that they have a compelling and transformative innovation. To listen to Dean Hager, their inimitable and endlessly enthusiastic spokesman for the product, Smart Office and Enterprise Search are, at the very least, what people were offered in the movie, “Seconds,” a brand new body and hence a second life.

So did they prove it? Nope. Yes, there were a few things in the demo which were encouraging. Some of the examples of how to use Search, for instance, were pretty insightful. And while the SmartOffice demo couldn’t be made out by those of us watching on the web, some of the claims sounded good, and the customers who were video’ed were enthusiastic.

But there were also a few things that were discouraging. Since I couldn’t actually see the demos (Lawson doesn’t invite me to their user conferences, even though I don’t ask them to pay my way), I can’t point to anything concrete. But I can say that I found the fact that Smart Office is a client technology (not a thin browser technology) provided by Microsoft less than heartening. (Anybody who doesn’t understand why that’s a problem should take about a ten-second look at a phone that runs Windows Mobile, then take a 10-second look at an iPhone.) Also not so heartening was the fact that Lawson’s broadcast of the event for “remote analysts,” that is, people they don’t think are very important, was incredibly web unsavvy. (Unless you use Microsoft products, you basically can’t even get a feed.)

Add up the discouraging and encouraging, though, and do you get proof? Certainly not. Both the positives and the negatives are way too vague.

And that’s precisely the trouble. It’s 2009. People have been working on these products for 20 years. There’s a great deal known about them. And today, there’s a presumption, born of bitter history, that when a vendor announces something fabulous, but most of what the vendor says is hand-waving, that what they’ve got isn’t really that fabulous.

So what would it take for me (or any rational person) to be convinced that the latest Lawson innovation is one that matters?

Well, let me tell you, it’s not easy. Even if I had full access to this application, something that there’s not a snowball’s proverbial chance that Lawson would give me, it would probably take me a day or two–I’m not kidding–to figure out how much of an improvement it is. To see why this is, let’s just take search as an example.

Lawson says they’ve made search of an application’s records as easy as Google search and that they’ve done this, as I said above, by putting the records on a separate appliance and using an open-source search engine. Certainly, doing this solves the problem of speed, which has been a serious problem.

But the trouble is that solving the speed problem does not make the record search as easy as Google. To make it as easy as Google, there are a lot of other problems you have to solve, too.

Here are some examples. You need some way of ranking what is sent to you in a way that makes sense for your uses; otherwise, as with Google, you’ll get tens of thousands of records when you make a search, and you’ll have to page through them. And you need to solve the ontology problem, the fact that you want to include synonyms in what you found, as well as exact matches. (Sometimes, these “synonyms” are nothing more than making sure that IBM, I.B.M., and International Business Machines are all included in the search results.) You need some way to use metadata descriptions. (Can you easily–and I mean easily, easily–restrict the search to sensible populations, like AR records, or sales orders from two years ago.) And above all, you need to deal with the problem of bad data. Many of the records you’ll grab in any search are records that shouldn’t be in the system. You need some way of filtering them out.

(Enterprise applications that have been in use for any time at all have this big, smelly layer of refuse in them, just the way lakes do. On the whole, I think, the bottoms of lakes smell better. Unless carefully managed, search has a way of stirring up all that stuff on the bottom, not something you want to do.)

If an enterprise application search product doesn’t deal with those problems, well, it isn’t much good. You can say it’s as easy as Google. But when the customer uses it, they just won’t have that experience.

There are still other problems, and one of them’s a real biggie: it’s what I call the transition problem.

You see, if the search (or Smart Office) is really going to change the way people interact with this 20-year-old app, people have to use it. So, in addition to providing a brand-new very hot, ultra-cool search capability (or Smart Office capability), Lawson has to make it possible for its existing customer base to take this capability up and take it up in a way that prevents the costs from overwhelming the benefits.

This transition problem is a thorny one for all the providers of middle-aged enterprise apps; I’ve seen versions of it at SAP, Oracle, and Infor, as well as Lawson. The provider figures out some way of getting an app to sit up and sing, “Rule Brittania,” as the old phrase goes. But then they tell the users that they’ll have to go through a massive upgrade and retrain all their users in order to make the trick happen. Then they’re surprised when most users opt to dispense with any stirring displays of patriotism.

To prove that Enterprise Search or Smart Office actually does something important, it seems to me that Lawson has to tell people (you, me, the customer, etc.) in a clear, simple way just exactly what it has done to deal with all these problems, and then it needs to provide us (you, me, the customer) with enough access so that we can genuinely see for ourselves.

Has Lawson done that? Well no, of course it hasn’t. I’m sure the thought of doing something like that never even entered Dean Hager’s head.

Not only is this unfortunate, it also says something important about the enterprise application industry and about the ability of Lawson (or any of the other middle-aged app companies) to modernize. The fact that Lawson (or SAP or Oracle or any other apps company that does this) doesn’t even realize that the burden of proof is now on them shows how far behind they really are. In a way, it’s puzzling. In the era of the iPhone and FaceBook and the collapse of GM, shouldn’t any company like Lawson be worried? But the answer seems to be, “No.”

Why don’t they think so? I”m not sure. What it comes down to, I guess, is that middle-aged apps and middle-aged apps companies are a lot like middle-aged people; it seems obvious to them that they’re still young, and so they don’t realize it when people start laughing at them. They think that they can give a cute demo at a user conference, have a lot of salespeople go out and tell their customers that it’s the latest, greatest thing, and then the IT staff will put it into the plans. That’s how it was ten years ago, and that’s how it is today. After all, the customers are paying maintenance, and Lawson is taking the maintenance money and working on things that keep the product up to date. What more is wanted?

Well, what is wanted is evidence. Without that evidence, the presumption has to be that applications like Salesforce (enterprise search three or four years ago) or Workday, which uses Flash, not Microsoft, and doesn’t use a 4GL, are now carrying the innovation banner.

Of course, the middle-aged apps vendors are unconvinced, just as all middle-aged employees are. We have the experience, they say; we have the know-how; we have the functionality; see, we can still dance a jig.

They say it, but in my view, they are wrong. Viewed by anybody who uses an iPhone every day, these apps just don’t work all that well any more, relative to how much they cost. They badly need to be brought up to speed. Ten years ago, we could assume that the companies were trying to keep up and we could accept mere assurance. But ten years later, that is no longer enough.

Yes, it may take the customers longer to figure this out than it does the analysts. But eventually, even the most ardent fan realizes that Elizabeth Taylor weighs a little more than she used to.

Oracle Layoffs

April 17, 2009

April 15 apparently saw a fairly good-sized layoff at Oracle; on LayoffBlog.com, the usual background chatter suddenly swelled to a roar.

I’ve been canned once or twice, and I once quit in advance of a layoff, so I know how bad it feels. For anybody who’s experienced this, I’m sorry.

It makes me wonder, though, what is a person’s useful life in this industry? We all know technology changes, and we all know that an existing technology can only be pushed so far. But for people who have trained as programmers or salespeople or application consultants (or analysts, for heavens’ sake), the feeling was that you had something–but what?–that transcended the technology, so that you could grow and adapt even as things changed.

Is that really true, though? Maybe, once you spend 5 years putting in Oracle EBS 11i v6-12 Accounts Receivable, the need for you in the next 5 years is not going to be all that great. The installations that are out there are settling in and working, so they don’t need as many consultants. The number of people out there who are buying and installing the product is going down. And really, what is it you know that’s worth $120,000/year (or $240,000 or whatever)?

I don’t mean to be mean when I ask that. I’m serious. I don’t know for sure now, but it used to be there were all sorts of date problems created by the way Oracle handled multi-geographies and accounting periods (two separate issues). Let’s say you’re an expert at writing reports that cover up those problems. Sorry, this is not expertise that will be valuable into the 22nd century. On the other hand, let’s say your expertise is in something far more general and transferable, like ways to raise cash on invoices. Yes, you can get a job somewhere else. But it will take a long time for you to learn the new systems and the new businesses, and at your high salary, that’s a lot of investment for people to make in you.

In my world, the place this really hits home is in marketing. I have any number of friends who had senior marketing positions at some application company or other and are now out with the trout. They’re talented people. But I wonder whether much of what they’ve learned in the past 10 years isn’t now outdated? It used to be that you were marketing innovation, so you had to teach people about your app and persuade them that buying this innovation would make a huge difference to them. But these days, apps are mostly a commodity, and the job in commodity selling is different.

So maybe, what you think is that you’ll hold a job in this industry for an entire working life, advancing from post to post and area to area, as you learn more and more, and eventually retiring with a big nest egg from those stock options. But maybe that’s just now how the technology business works. Maybe you get in and what you know is only worth what you’re paid for 10 years.

After that, you’re out.

Talk to the typical SaaS executive about the complexities of customer service in Cloudland, and they give you a blank stare. What’s the problem. We provide our service. We guarantee it. We do a better job than

Well, yes. But that’s not all. When you’re getting cloud services, the services are often coming from multiple providers. Each of them can think they’re doing a good job or the right job. But even so, delivery can fail. And when that happens, it’s a nightmare for the customer. A nightmare.

Here’s a simple example. I have a Blackberry Pearl. My web site and domain name are hosted by XO. Two SaaS providers. I get my Blackberry e-mail from XO, which forwards it to the Blackberry domain name (tmo.xxx.com).

All of a sudden, this weekend, I started lots and lots of new spam–you know, the usual, sex stuff, plus a lot of spam with a Facebook return address, etc., etc., stuff that had been filtered. The new spam appears on both my Blackberry and in the e-mail delivered directly to my POP3 Mail account on my Mac. But much more new spam appears on my Blackberry.

Obviously, something happened to one or more of the spam filters that are provided to me. I know of at least two of these filters: one at XO and one on my Mac. Is there another one at RIM? Who knows.

I’d like it to stop, but I don’t have the time to troubleshoot this problem. Did XO change its forwarding policy? Did RIM install a new filter and screw up? Who knows.

If I just had one provider, I’d have some chance of solving it. But when I have two, I just have to sit and wait. The worst of it is, it might be the case that neither provider will ever realize that there’s a problem.

I wonder if the iPhone has a spam filter on it?

The point? Well, cloud service providers need to start thinking about this–and taking responsibility for the actual delivery of services, not just for making those services available.

Acres of Servers

April 13, 2009

Brian Sommer offers this astonishing story about a bad implementation. The story comes from early in his career.

“I worked an entire year getting the first mainframe General Ledger (and other vendor apps) in production at a Texas client. I have never before or since worked so many hours in one year.

After that project, I was off to another installation of the same vendor. When I arrived, I found a team that had been diligently preparing for a conversion to take place in approximately one year.

The users wanted to push this software product to the absolute extremes of its functionality. The software was going to be configured with virtually every conceivable option.

Having just lived through the Bataan death implementation, I knew there were some practical limits for this software. Within 24 hours of being on-site, I shared with the client my disk space concerns for one of the validation tables: a reverse code block look up table. By my calculation, this table alone would require 14 acres of IBM 3350 disk drives. That’s
correct:14 acres.

The client didn’t believe me.

Two weeks later, the client executive director invited me to listen in to a call he’s having with the vendor’s CEO. Everything went well until we got to the subject of the lookup table. The software CEO confirmed my math and told the client that his software was ahead of its time. He said that hardware hadn’t caught up with his software and that the client should “either scale back the functionality or do a hostile takeover of IBM”. Needless to say, that didn’t go over well with the client.

Now the client believed me.

They also believed me when I told them that the product wasn’t ready to run in IMS DB/DC. It took another 12-18 months before the technical platform was ready.”

Thanks, Brian. But one question. Who was at fault? The client? The vendor?

“Fault. It was the project team. They went along with everything the client wanted. Sometimes, in the long run, you need to show clients a little tough love.”

Oh, and one other. What is an IMS DB/DC?

My Worst Implementation

April 6, 2009

Vinnie Mirchandani has been asking his friends to write about the way technology has changed their hobby. He’s gotten responses on everything from chess (from his daughter) to repairing old cars. Now he’s got a column on bridge, a hobby of mine that has been transformed by the multi-player gaming technologies that got their start, among other places, in the old Project Athena.

You can read the post on his Innovation Blog.

But I always think, one good idea should beget another. Here’s what I suggest. I will ask friends of mine to send me a short article called, “My Worst Implementation,” in which they briefly describe a failed or completely inadequate implementation that they’ve been associated with personally in some way.

My single worst? Well, there are so many. But it was probably the implementation of SAP that the project manager later described as, “Like going through the gates of hell.” I kid you not. The factory, which made telecommunications equipment and was located in Westminster, Colorado, first bought QAD, which is why I got involved, then dumped QAD and bought SAP. $600 million later, the implementation worked–better than some I’ve been involved with–and the project manager was justly proud of what he’d done. But by that time, the business had completely changed, and I believe that almost all of what they made is now made overseas.

Ooorgh.

Got an oorgh of your own? You don’t need to be my friend. You can just send me an article. Write me and inquire; you can get my e-mail from the web site, but I don’t want to post it in a blog.

A few ground rules, required. Tell the truth. Don’t name names, except the names of application companies. Explain why things went wrong, but recognize that these things are never one person’s or one company’s fault. I haven’t done all that yet, but this is already too long.

Follow

Get every new post delivered to your Inbox.