Skip to content


Why Improv didn’t succeed, Part Deux

Seems like a dredged up some "old" (you know who you are) Improv fans with my earlier posting, so I thought an additional comment or two might be in order.

First off, lest it be misunderstood , in no way do I feel anything but totally proud about Imrpov as a concept and a product. I hope it is not inconsistent to feel that way, and yet agree that for a variety of reason it was not successful, as a product.

The concept of an alternative (not replacement) for a spreadsheet for real financial modeling is, I think, quite sound. (Yes, I can do page layout in Microsoft Word, but if that's my purpose, PageMaker is better. You could make an analogous point about PowerPoint and PhotoShop.)

By the same token , if what one is building a real financial model (i.e. a business plan or an integrated set of financial statements) then a product like Improv would be far superior. I say like Improv, because that was over 10 years ago, and it certainly had room for improvement at the time.

I received an email from Peter Murray, CTO of a company called Quantrix. They have created a new application called Quantrix Modeler in the spirit of Improv, which has continued to improve and model. Here is what Peter said of my posting [slightly edited for length]:

Dear Pito,

I was interested to read your response to Adam Bosworth's talk @ ICSOC'04.

The Improv model is not "ancient history" yet! At Quantrix, we believe there is significant value in the Improv inspired approach to building complex models - and our customers are proving that out. There is no question that the free-form two-dimensional-grid based approach is useful in many cases. However, as soon as the calculation begins to get complex or the model is utilized through time in many iterations, our customers find that Quantrix is a more powerful tool.

Our theory on the reasons for the demise of Improv follow more the line of thinking in your final paragraph, rather than the idea that the product is somehow inherently flawed:

1) Lotus was positioning Improv as a spreadsheet replacement, rather than a specialized tool to better perform an important subset of tasks currently performed with a spreadsheet.

2) Lotus was in the throes of a heroic battle for survival against Microsoft's Excel - causing undue pressure on the company to make its product portfolio clean and understandable, and to organize all resources behind the flagship product. Introducing new technology costs in a scenario such as that.

Quantrix is employing a patient approach which identifies niches where the pain on traditional tools is great and the relief from a more scalable, transparent solution is palpable. As we progress, we are listening to users and building in new / better functionality - and thus making the product more attractive to a larger audience. In fact, we have folks from all kinds of disciplines who have sought out an alternative solution to the traditional spreadsheet - financial planning / budgeting professionals, equity analysts, business consultants as well as engineers, scientists and policy researchers. We even have customers working on genome analysis with Quantrix.

So, rather than ancient history, we prefer to think that - in this case - "what's old is new again".




And just to close the loop, here is what I responded:

Overall, I agree with that analysis.

Of course, I would hope that the Improv model wasn't ancient history! I was referring to the particulars of my experience at Lotus. I can be critical of people who live in the past and relive long past experiences so I am always careful of that myself. That is a totally closed chapter for me and I wouldn't have even thought of it until I read Adam's story. My feeling was that at a very high level he was hitting the nail on the head.

The notion of something being "inherently fatally flawed" seems relative to me when applied to a product. For example, if I tell you I have designed a great new televeision but I don't include a screen, well it's fatally flawed. But if I told you the same thing was a radio, well, you see where I am going 🙂

Predicting the future is hard


_" Scientists at the RAND Corporation have created this model to illustrate how a "home computer" could look like in the year 2004. However the needed technology will not be economically feasible for the average home. Also the scientists readily admit that the computer will require not yet invented technology to actually work, but 50 years from now scientific progress is expected to solve these problem.

With teletype interface and the Fortran language, the computer will be easy to use" _

[From 1954 Popular Mechanics Magazine, from Peter Grossman]

Late breaking news: the whole thing is an urban legend. No such photograph ever appeared in Popular Mechanics. Thanks Peter and Larry for helping me preserve my journalistic integrity.You can read all about it on the wonderful Snopes site. The big steering wheel should have been the tip-off.

Adam Bosworth on KISS

Adam Bosworth's speech

Adam Bosworth's an interesting guy, and very smart, and has been deeply involved in web services and many related technologies for years. He really knows this stuff. And it's interesting, because one of the key Web Services technologies is SOAP, a loosely coupled, "webbish" way of doing remote procedure calls.

[Aside #1 : If you feel your eyelids getting heavy around now, I appologize for not labeling this entry as "GEEK" but I thought it was border line]

[Aside #2 : I wonder whether the term "Web Services" is now dated, and what the more au courant term is. Perhaps "service oriented architecture". If so, please accept my appologies for not being fully PC]

[Aside #3 : I see that Aside #1 and #2 are both appologies. I should have labeled them Applogogy #1 and #2. I appologize.]

Anyway, if you've followed SOAP and the related and simpler standard known as XML-RPC you will have noticed a big difference, particularly in the degree of complexity in each scheme. And by the way, SOAP has spawned a whole series of related formats or standards, one more complicated than the one before it.

Anyway, Adam Bosworth, closely connected to the origins of SOAP, in this speech gives an empassioned plea for what we know as the "KISS" principle. Where I went to school, KISS stood for "Keep it Simple Stupid" (yes, I didn't get out much 🙂 but Adam Bosworth uses to mean Keep it Simple and Sloppy.

This argument about how simple is too simple is an important (and not new) one in our field. Adam Bosworth's article is a good piece because it collects together many of the traditional points as well as clever phrases which are used in this debate which we see in our field in many different guises, and it makes you think.

I don't have the answer or really a clear opinion. You can rattle off standards / protocols / formats / languages which are easy to understand (for example RSS 2.0) and explain and those which are thoroughly inscrutable and confusing (for example XML Encryption.)

I could cite a dozen other examples , but I won't.

When I say I don't really have an answer or clear opinion, here's the reason.

When I read something like the XML Encryption standard, I get totally lost. My default conclusion is that it's just over my head, and whoever designed it understood the the subject matter better than I, and I just am thankful for good libraries to call. But on the other hand, I have seen plenty of examples of totally overengineered systems and architectures to know that unnecessary complexity is definitely a disease that inflicts us in this field.

So I am not ready to say one way or the other that simpler is always better, or that one approach (for example SOAP) is clearly inferior to another one (for example XML-RPC.) Adam Bosworth's piece won't answer this question, but it is thought provoking (as you can see, because it provoked me to write this little piece)

Why Improv didn’t succeed

It's been a while.

Which is why my heart beat just a little faster when I read a reference to Improv (not altogether flattering) in the Adam Bossworth piece I just wrote about, where he said:

Consider thespreadsheet. It is a protean, sloppy, plastic, flexible medium that is, ironically, the despair of all accountants and auditors because it is virtually impossible to reliably understand a truly complex and rich spreadsheet.

Lotus corporation (now IBM), filled with Harvard MBA 's and PhD's in CS from MIT, built Improv. Improv set out "to fix all this". It was an auditors dream. It provided rarified heights of abstraction, formalisms for rows and columns, and in short was truly comprehensible. It failed utterly , not because it failed in its ambitions but because it succeeded._

I happen to know a little bit about this 🙂 I was the guy , for better or worse, who came up with the idea for Improv and got it built and shipped at Lotus.

Let me start by saying I have neither a PhD nor an MBA (and proud of it!) And I am almost positive there was not a single one of those on the Improv team. But other than that, Adam makes I think an important and accurate observation.

This is so many years ago now so I don't think about it much anymore. But at one point I was often asked and thought about what happened with Improv and why it failed. And while I had more complicated answers at the time, I have to say that there is a lot to what Adam says.

It's easy to get all philosophical about what makes a spreadsheet a spreadsheet, what it's core essence is. And I have. It's hard to argue that one of the keys is the maleability of the spreadsheet as a medium. The fact that a spreadsheet can grow organically, be modified and grown in a kind of an Improvisational manner. But when spreadsheets get complicated they get messy and error-prone and this is what Improv set out to address.

In the end it didn't go anywhere, probably because in setting out to improve on spreadsheets, Improv lost the essence of a spreadsheet and in doing so lost the market.

Innovators Dilemma

There is an interesting Innovators Dilemma kind of perspective here, referring to the now-classic book "The Innovators Dilemma" by Clayton M. Christensen. To summarize one of the key theses of the book extremely briefly:

  1. that a new technology might initially have a fatal flaw which makes it unsuitable for the purpose it was intended for. (For example, Electric cars don't have the range or speed to meet the needs of today's drivers.)

  2. An established maker of cars (e.g. Ford) has a strong disincentive to invest in Electric cars because it doesn't meet the need of it's existing customer base.

  3. An upstart competitor could identify a different market where the apparent flaws (speed and range) are actually 'features' not 'bugs' (for example a car specifically made for very young drivers) and use that market as a spring board to be able to perfect the technology.

  4. Eventually the upstart has years of experience with the new technology (the batteries and drive train) and improves it to the point where it actually does meet the need of mainstream customers, while the established player has been ignoring things and gets unseated.

I am not sure it applies, but one could argue a parallel here with Improv. In particular this would lead you to the conclusion that the key strategy mistake was to try to market Improv to the existing spreadsheet market. Instead, if the product were marketed to a segment where the more structured model was a 'feature' not a 'bug' would have given Lotus the time to learn and improve and refine the model to a point where it would have satisfied the larger market as well.

Who knows. Ancient history.

What is a Blog, redux

I just came across a good summary on one of our favorite topics from KM Magazine. It was written by Bill Ives of Portals and KM fame:

" Blogs are both a technology and a way of communicating. On the one hand, they are nothing more than simple Web sites, often published in the form of a journal, that provide original content, links to other relevant content, simple editing features, comment capability, archiving, search, and security options. This ease of use provides a vehicle for the writer's unedited personal voice, which is considered a defining aspect of blogs. While many uses beyond personal journalism have emerged, the individual voice remains a central characteristic, even in the corporate setting."

I'll forgive Bill for not mentioning BlogBridge 🙂

Does this map look familiar?

Have you seen this map before? Yes?

Actually it's the map of states where the Cosi Restaurant chain has stores. I don't know what it means, but I thought it must mean something. Amusing.

(from a private mailing list I am on)

[Note to those who don 't obsessively follow US Politics and culture: the map above looks a lot like the map showing U.S. states where the Democratic candidate for president won in blue and where the Republican candidate for president won in red. However the map actually shows where a certain sandwhich restaurant chain has stores. The relation between the two is a mystery to me too.]

What is a blog, again?

My answer to the question: " What is a Blog", which I get asked often, goes something like this:

"A blog is a just a web site with articles or notes that get posted there, generally by a single person. There's nothing especially new or complex about the technology of blogs. What's new and compelling is the blogging phenomenon : For whatever reason, individuals with unique and important points of view are choosing to publish those in blogs."

Actually the point of this post is to point you this great and I would almost say moving account by Kevin Sites, keeper of a photo blog that was very active and interesting earlier this year and the year before.

Kevin Sites is a reporter who has been covering the Iraq war from the inside. You might have heard about the video tape of a marine shooting a prisoner which has made quite an uproar both here and in the Middle East.

It is, to me, amazing to be able to read, in Kevin Sites own words, what he saw, what he thought about it at the time and now, and how it has unfolded for him.

New BlogBridge Alpha

Hey, sorry if you aren't a BlogBridge user yet, but I just can't resist some news here.

First of all, we've released a new Alpha, just today. You can get it here. The way we work it is that if after a week or so of use there are no show stopped bugs, we will promote the Alpha to a Beta. The general release model is described here.

This new version has a ton of new feature stuff, plus major stability and performance improvements. I dare say, we are getting close to having a Release 1.0 product. Here are some highlights:

  • The BlogBridge Service gets a major new capability, which is RSS Discovery Services. The change to the user experience is minimal but I would say pretty significant. When you want to add a Feed, you don't need to hunt for the RSS Feed URL, which our non-technical users won't even understand. You just enter the URL for the Blog itself, that is, the one you read with a Web Browser, and BlogBridge will attempt to 'discover' the RSS Feed. Behind the scenes it spawns a process to talk to the BlogBridge Service to resolve it. Try it, it's pretty cool!

  • The second major new capability is the Feed Discovery Service. In it's initial form, there is a new command on the Feed Menu: "Discover new Feeds". Again working behind the scenes with the service, BlogBridge will analyze all the articles in a certain Feed, to see whether any of them point to Blogs that you don't know yet. If so, they are highlighted right in the article (let's say in Red.) The user can right click on the link and in one step add that new Feed to their subscription list.

  • Among the smaller changes are, much better (although not yet perfect) support of Mac OSX, a minor change to the menu structure and terminology based on customer feedback, and numerous bug fixes and performance improvements.

Here is a link to the announcement email we sent to our users.