After our first round of product testing, we decided to make big changes in the look of our Infovark product. This is the first of an occasional series of posts about the design of the Infovark Beta.

We use the Infovark mascot — we call him “the Vark” — to represent our company. We use his head for our avatar in twitter and as the favorite icon for web bookmarks.

I like the simplicity and abstract look of the icon. It has a certain Scott McCloud less-is-more aesthetic to it. As Scott explained in his book Understanding Comics, the less detailed an illustration is, the more it invites the viewer to use their imagination to fill in the gaps. The more detailed a drawing, the more it represents a particular thing. We wanted Infovark to be about you, the knowledge worker, and not about us, Gordon and Dean, two guys making software.

We also chose an animal mascot for a reason. The Infovark name was partly an accident, but it stuck because it was a useful metaphor for us. Infovark does a lot of work in the background for you. It watches files, catalogs email, indexes content, and other useful things automatically. We liked the image of an animal digging through your information, trying to find interesting items and making associations between files, contacts, and email. It made our software feel friendly and helpful, like an enthusiastic pet you’ve taught to do neat tricks.

For round two of our product design, we wanted to try giving the Infovark just a little more personality. But not too much personality. Ever since Microsoft introduced Clippy to the world — and to much popular scorn — software developers are mindful about overeager assistants. Does anyone remember these guys? It’s a fine line that we’re walking here… 

So after much consultation, this is our attempt at something a little cuter, with a touch more detail.

What do you think? We’ll use this new character on our redesigned blog theme, and possibly in the product itself. Leave a comment and let us know your thoughts!

Continue Reading

I recently finished Keeping Found Things Found by William Jones. It’s subtitled “The Study and Practice of Personal Information Management” but I think it ought to read “The Study of a Practice…”

The book is intended as a broad survey of information management practices and principles that individuals use to manage information. This is not a how-to manual or a self-help guide. It’s written in the style of an introductory textbook, which is not surprising given that much of the content was developed as part of the Keeping Found Things Found Project, run by the Information School at the University of Washington.

The academic style of the writing might appeal to some, but I found it a slow read. The footnotes and citations often got in my way as I was trying to absorb the material. Each chapter begins with a lengthy introduction and ends with a point-by-point recapitulation of key messages. These are probably useful for a semester-long course, but it’s overkill for anyone wanting tips and tricks.

Your mileage may vary, of course. I’m not the target audience for a book like this. If you want a primer on the personal information management field, Keeping Found Things Found is a decent place to start.

Key messages

We live in a world awash with information. Some of it is useful to us, but much of it isn’t. The point of information management is to keep the useful information easily accessible while ignoring the unhelpful bits.

This is an extremely challenging task, because we use information in a variety of different ways depending on our current situation. It also requires us to make some guesses about the future. How likely is it that we will need to recall or retrieve the information again? Is it worth memorizing it? Writing it down? Saving a hyperlink or adding a bookmark to our browser? We can’t predict our future needs exactly, so we often wind up storing more information than we need.

And that leads us to develop strategies for organizing it, indexing it, and searching it. And because we’re all different, there are almost as many different methods for keeping found things found as there are people. Some of the strategies — like the post-it notes stuck to your monitor — are obvious. Others are subtle, like keeping a mental list of go-to people that are experts in a particular topic so that you can ask questions as needed.

How this relates to Infovark

Having recently worked at a firm making records management software, I was familiar with much of the material in the book. The main reason for reading it was that records management focuses on large-scale organizational methods used by government and corporations to store, track, manage and dispose of information. I wanted to find out whether there were specific strategies that work well for individuals or small teams. Were there methods or tools that might help us refine Infovark’s design?

Well, I didn’t spot any, though some of the sources cited in the book might be worth checking out.

One of the book’s themes that connected with me was that we are really operating in a different age. The amount of information available to people is orders of magnitude greater today than in the past. Many of us now have music collections on our iPods that would have put independent record shops a few decades ago to shame. We have vast amounts of movies, television, and other video available to us on DVDs and streamed across the Internet. And then there’s the mountains of text and reams of financial data. All this content means that we’ve outgrown many of the concepts that guided us and tools we used in the past.

Gordon and I saw this clearly in our old job. It seemed silly — ridiculous, even — to expect a large organization to get a handle on its information when most of its employees have difficulty dealing with their own email inboxes. And companies can’t afford to train all their staff in the ways of research librarians so that they can manage the stuff on the corporate intranet. We need new approaches.

And we’ll keep reading, and talking with people, and attending conferences, and scouring the Internet, until we find them.

Continue Reading

Somewhere, there’s a dusty old librarian who thinks Google is a messy and irresponsible way to find things, and gets very annoyed when you  don’t put the index cards back in the right place in the catalog. I found this one under ‘P’ for Process…

I spent some time with a client the other day who was trying to determine the best way to manage how information flowed in their organization. They were having a difficult time, because they had laid out a series of discrete steps that needed to be followed by every person in the organization — first the document arrives, then the document is entered into a tracking system, then it is registered as a case in the case management system, and so on. There are thousands of these kind of processes being discussed in meetings all over the world while we drink our morning coffee. I must have sat in on hundreds of meetings like this. 

The problem within  this organization was equally commonplace. The staff weren’t following the process. The software was too weird/hard to use, and it forced them into working in a way that they weren’t used to working. So, because they wanted to get on with their work, they were going around it. Composing documents outside of the system, and saving them on their local machine or network share. Getting things done despite the fact that they were aware it was ‘the wrong way’.

Typically the way these meetings go is like this: Once it’s determined that people are doing it wrong, somebody insists that they need to start doing it right. And that it’s “Their Job” and they should “Do what they’re told” and then the meeting usually decays into “Change Management” and how best to force acceptance of a system that nobody really wants.

There are numerous problems at work here. The centralization of authority forces people to rebel. That happens all the time. There are probably other leadership issues, and poor planning, and issues of control. There’s a lack of willingness to admit defeat, or worse, good money thrown after bad. There are loads of other subtle personal and political issues that are also affecting the outcome. But you know what? The work is actually still being done. And it’s not like the workplace is ever going to be free of these problems — they’re part of human social behavior.

Process is not our only product

The whole time, while this conversation was playing out, I was thinking about Infovark, and how we are trying to solve this problem in such a completely different way.

Infovark encourages the user to do whatever works for them, and then publishes all the output (or lack of output) to their team. We don’t control “the process” at all. We simply measure the output and let folks figure out what works and what doesn’t. With an open and transparent system, the best methods ought to rise to the top. Why not let the best processes emerge as memes within an idea economy?

Is this irresponsible? Perhaps. Bad eggs could try to game the system. Clueless bosses could incentivize perverse behavior. And certainly, in some businesses, overall consistency might be more valuable than quality or efficiency.

But frankly, rigorous approaches to process control aren’t common in most businesses. It requires discipline and training to enforce strict guidelines. It costs time and money.

The Strict Process Control approach to productivity and efficiency might make sense from an academic or manufacturing perspective. But nothing pollutes a good process like a bunch of humans. Now, where do I put this index card again?

Continue Reading

Here’s a philosophical topic to start the week. Information Aesthetics, a blog about design and data visualization, posted two videos from Maya Design. The first discusses the term information, the second architecture. As a software developer, these are terms I use all the time, but I often have a hard time defining precisely what they mean.

It’s a bit of a challenge for our marketing and communications strategy. The core API for Infovark, our product, is fundamentally nothing more than an information architecture. We use this core to create solutions that help knowledge workers get stuff done. But if information and architecture are tricky to define, you can imagine the confusion involved in combining the two! But where was I?

The two videos do a great job describing the terms information and architecture from the classical point of view.

• Information is the essential message being communicated. It is separate from the form the communication takes.
• A design is a plan for making a thing. An architecture is a design for making plans.

These definitions are classical in the sense that they rely on Plato’s concept of an idea being separate from a particular physical representation. There are other approaches. For example, Marshall McLuhan believed that “the medium is the message” — that you can’t separate the meaning from the representation that conveys it.

Did I mention this was going to be a philosophical post? I did warn you…

Getting real

Even though the topic is a bit academic, there are real-world debates going on about representations and forms right now.

How many differrent representations of you can be found on the web? There’s your LinkedIn profile, your Facebook page, your  StackOverflow account…
Much of the buzz surrounding the open stack technologies is due to there being a way to finally associate a person with all the stuff about that person on the web. Wikipedians talk about having “canonical URIs” — a web page address that points to the master record for an article, rather than a particular translation of an article. And if you support or build software that uses the REST pattern, you’ll face this issue. What is the base address for this web page? How do I get to the HTML version, the XML version, the PDF version, or the MP3 audio version of the page?

There’s not a “right” answer to any of these questions, but we usually adopt certain conventions for dealing with them. Libraries had to figure out whether to keep Braille editions of books together with the print editions, for example. After all, they’re the same book — they contain the same message — but the format is different.

We’re working out those same conventions on the Internet right now with regard to privacy, data portability, text translations, podcasts vs. screencasts vs. print, etc.

Continue Reading

ConnectBeam pointed me at a study conducted by HP of Twitter connections. It suggests that people interact with very few of the contacts in their social network regularly. ConnectBeam suggests that  ”Social Network Analytics” — figuring out which contacts are more important than others — will become increasingly valuable as we explore social networking within the enterprise.

These findings are interesting from a social perspective (Hey, it turns out that most people with 1000 friends on Facebook don’t have more ‘active’ connections than you do!) but when it comes to leveraging your network inside the corporate firewall, they are much more important.

Dean and I have always known that some people within the organization are more important than others — as senior technical consultants in our last company, TOWER Software (which, in a nice twist, ended up being bought by HP), we were constantly in demand. We provided our expertise to sales, services, support, pre-sales… In short, nearly everyone in the company depended on us for at least some part of their job. We would get calls from people we knew on behalf of people we didn’t, and be brought in on everything from sales calls to marketing events. Our manager spent a lot of his time ‘running interference’, when we actually had work to get done. The nature of enterprises tends to create these information hubs and dependencies between each other:
 

Here we can see that there are two discrete information pathways between Sales and R&D, and two between Marketing and R&D, all based on ‘who knows who’. Sometimes these channels are formal, often they’re informal, and the one thing we’re learning here, is that it doesn’t really matter, because the information will flow anyway!

Like most human social systems, these relationships tend to break down when you exceed about 200 people. As long as everyone knows everyone else, these systems operate wonderfully. When your organization grows bigger than that, you end up with two major problems:

1. Where can we get the information? People often know they need more information, but they don’t know where to get it. This is the classic discoverability problem: you don’t know what you don’t know. In our example above, Bridget in Sales isn’t getting any information directly from Charles — the only possible route is fourth-hand information via marketing, or third-hand through Aaron. Is this lack of information affecting Bridget’s commission? Who knows? Bridget doesn’t.   

Addressing the discoverability problem from an information-only perspective i.e. the way Google fixes the problem doesn’t appear to be ideal within the enterprise. People can misinterpret information and corrupt its original intent. Information-only solutions can lead to situations where Bridget quotes technical specs from an R&D product that doesn’t exist to a potential customer.

It’s much better to be able to get the people involved in this information transaction connected than to pass raw data around. That’s why a lot of this enterprise social networking stuff is so interesting. Having done that, we still have the second issue.

2. How can we evaluate the quality of the information source? How can we determine that information from one contact tends to be more correct or useful to us than information from another contact? This is a much harder problem. Small, subtle errors can have cumulative effects that degrade productivity.

Here at Infovark, these are the kind of things that we lay awake at night pondering.

Simplify, simplify

The way we attempt to solve this problem is to make a few assumptions.

First, we assume that your enterprise is generally successful at whatever it is you do. We figure that the current mode of operation within your enterprise has led to some kind of success. Once we assume this, we can watch how people work within the organization, and  build a pattern of communication pathways — just like the ones that HP studied on Twitter.

Having an idea of how people communicate with each other is the beginning of solving the problem, but we still need to address the quality issue. How do we know which are the best ones? What characteristics define best anyway?

Trying to find an answer to those sorts of questions is impossibly hard for two guys in a basement. Hence our second simplifying assumption: we figure that “repeat business comes from previous success”. Instead of trying to teach Infovark how to evaluate different sources of data, we simply watch what you do with the information. If you consistently re-use an information pathway, we assume you must have a reason for that. It must be more useful somehow. So we strengthen that relationship in our calculations. “Stronger” isn’t the same as “better”, but we’re banking on people to be lazy and gravitate to the pathways that require them to do the least amount of work for the most reward. It’s kind of an Invisible Hand principle.

So, the individual needs to be presented with all the options — the strong and weak connections. People can try out the other paths, and if they become more used, we assume that they are ‘better’, and adjust them accordingly.
 

Now Bridget can see that it would be best for her to connect to Anthony, Charles or Brenda — but most of the Sales/R&D communication seems to go to Anthony. Armed with this information, she can choose to follow the same path, or connect directly with Charles, or even talk to marketing. The awareness of the most used pathways, and the available contacts, give her a better idea of how to go about solving her discoverability problem. 

Over our experience with Infovark we’ve found time and time again, that the best way to improve productivity is not some huge mathematical scary algorithm to divine hidden trends, but simply to watch the way that people work and how they work with each other. 

Social Networking Analysis? Or just remembering what works?

Continue Reading