The holiday season is now over, so it’s time to pick up where I left off.
The previous two parts of this short series looked at the nature of social networking and what business process management does.
In this section, I briefly consider what software does, how we work with it and the kinds of system that result. I then look at how they interconnect. The diagram below sets out the flow of logic. Click on it twice to see the whole picture full-sized.
The four kinds of software
Computer software, broadly speaking, is of four types.
- Type 1 enables the computer system to run. This includes operating systems and software for database management, system management, and networking.
- Type 2 helps human beings use the system. Examples include programming languages, graphical user interfaces (often provided with the operating system in personal computers) and World Wide Web browsers.
- Type 3 solves the problems that computers create, such as exchanging data among dissimilar systems, converting among different protocols for data communication and file creation, and managing security.
- Type 4 does actual work, on its own or with people. Examples include accounts packages, spreadsheets, and programs for word processing and production control.
To the computer industry outsider, there is an astonishing amount of time, effort and money devoted to the first three categories. It astonishes (and dismays) many insiders, too.
For the user organization, those first three categories all count as overhead (‘burden’) or on-costs. Reducing their drain on corporate resources is an important objective for any organization that uses computers.
My concern in this discussion — as in Office Jotter generally — is with Type 4 software. Call it application software, line-of-business software, applications, applets, apps, widgets or whatever you like, it’s the software that ordinary employees grapple with to do what they’re paid to.
The three ways of interacting
When people use application software, they deal primarily with one of three realms — data, people or processes.
In the first — data — users interact with some form of electronic record. It could, for example, be a database, directory, document, spreadsheet or video recording. Software within this category includes BI, CRM, EDM and ERP systems.
Social networking activities such as shared editing, wikis, voting, podcasting, blogging and tweeting fall into this group. (Twitter direct messages are personal, so belong to the next group.)
In the second — people — users communicate with an electronic version — a simulacrum — of another person or other persons. Examples of this kind of interaction include instant messaging, texting, emailing, conferencing (text, voice and video) and telephone calls.
Interacting with people is the basis of collaboration software; it’s you connecting with me or them. It involves moving data about the place, of course, but that’s neither the primary intent nor the users’ perception.
The final realm — process — is where a person deals mainly with computer-based logic. This could, for example, be as project management, workflow, scheduling (in the proper sense, not just a diary), BPM or e-commerce transactions. Telephone call-routing systems are another example. Again, the transfer of data is implicit in the interaction but is not the main purpose.
Most comprehensive social networking products enable or mediate interactions with data and people. Tools for process handling are also starting to be incorporated into social offerings. How that might happen is what I consider in these four articles.
The two kinds of system (says AIIM)
AIIM is an American-based body representing suppliers and users of systems for image and document management. Its supplier members include Autonomy, Eastman Kodak, IBM, Microsoft and Oracle.
In October 2010, AIIM published a report from one of its working groups, called Systems of Engagement and The Future of Enterprise IT. In the group’s view, computing in business has, up to now, concentrated on making, keeping and updating data records. I think most of us would agree with that.
AIIM calls these systems of record. As the report says, their main focus has been on transactions. Their main value resides in presenting a “single source of the truth”.
Organizations are now adopting social networking systems, which AIIM calls systems of engagement. Here, the focus is on interactions; the value is in providing an “open forum for discovery & dialog [sic]”. In other words, these systems allow the communal search for the truth or truths.
You can read more about the topic in this presentation. Both publications repay reading; they’re not long. The working group plans to release this coming November the results of further work on the subject.
The three kinds of system (say I).
AIIM’s review is helpful and its terminology apt but I think its model needs refining. In my view, its scheme should comprise three kinds of system, echoing the three kinds of human-computer interaction. The new, third group would result from splitting off process-based systems from those of transactions.
I don’t know why AIIM conflated them in the first place. The organization has a long history of involvement in workflow management, so should understand the distinction between the two kinds of system.
We might call the third group systems of action. Their focus is on movement and their main value consists in assuring progress to a truth.
Linking the three kinds of system
A major task for user organizations and suppliers is making connections among the three kinds of system. The links need to be robust, cheap in monetary and system performance terms, and easy to make and use. If they can be generalised, rather than being specific to each pair of supplier’s products, so much the better. (This is Type 3 software, by the way, as described at the start of this posting.)
Each cell of the table below comments on the kind of link or links that apply for that particular “from” and “to” combination. (Click to enlarge.)
Where the link is between two or more instances of the same kind of system, such as record to record, I have labelled this “native”. It’s reasonable to assume that such a link meets all the requirements set out in the previous paragraph.
Links between systems of record and systems of action are well established, so should meet at least most of the desiderata. I’ve labelled these “extant”.
The relative newness of systems of engagement means it is uncertain how they might link to the other two kinds of system. Where the link is to or from a system of record, these might be:
- Arising from interpersonal communication and going to data stores — tagging; communications transcripts & analysises (perhaps by text analytics); activity reports; location data.
- Supplied by data stores to aid interpersonal communication — BI-style feeds; search; user profiles.
I have put question marks where systems of engagement intersect with systems of action. How they might interwork is a matter of current debate. Will, for example, the same kind of link be needed when going from social software to process software rather than vice versa?
AIIM has made up its mind. “Clearly, systems of engagement need to operate on top of and in touch with our existing core systems of record”, says its white paper. Other people, me included, are not so sure this is true. I discuss possibilities in the final article.