Integrating multiple technologies such that they work as a single technology can be one of the most difficult and frustrating things we technologists do. Who has not suffered from the “it works on my machine” syndrome?
One day back in 1992, I was working with an analyst who needed to produce a one-page trading summary report after the markets closed each day. The report went to all the brokers, traders, and management summarizing the day’s activity. Some of the information for the report came from their trading system, some from the accounting system, and finally a small graph from their market data system. The core problem with making this report was that none of the systems supported any type of end user data export capabilities. In fact, one of the systems could only be accessed by its own terminal. What to do?
Her self-service solution certainly solved the problem: First, she printed separate reports from each of the systems. Then she got out her scissors and glue stick. Then, through the literal process of cutting and pasting, she glued the report together by cutting out pieces from each of systems. After it was glued together, she took this “Franken-report” over to the copier and made a copy. From there she used whiteout to cover all the cut lines. Once the whiteout dried, she had the master from which they cranked the required copies of the trading summary report. Twenty minutes later: requirements met!
The funny part about this is that I was there as part of a team to bid on the automated creation of this report. At the time, I remember thinking, hey the market data system has its own terminal and printer, how on earth are we going to automate that? Is this really a good idea? The resulting price tag to automate this single report was quite steep.
The ultimate solution was a custom report rendering application which picked up data dumps from a batch process that ran on each of the individual systems. However, they still ended up having to manually enter the market data directly into the reporting application each and every day. This data was used to reproduce the market graph as part of the printed report. At the end of the day, one could argue that the core requirements were met, but was it worth it? The analyst had to pick up a new daily task of entering the data for the market graph, which took about ten minutes each day. The old self-service solution took around twenty minutes end to end. When you rack it all up, the whole project ended up saving one person ten minutes a day. Excluding the cost of ongoing changes, which were frequent, the break-even for this project was easily north of a hundred years.
No matter how you slice it, integrating separate technologies is difficult or impossible without a common denominator that both the integrator and all the technologies can operate against. In addition to knowing how to integrate the technologies, the integrator must also have some type of operational canvas upon which they can express, validate, and execute against this understanding. The accumulation of these factors: shared understanding, expression of intent, and means of execution form the ecosystem that I call the empowerment food chain.
In the case of this trading summary report, the physical printed page was the highest common denominator across the systems that met the criteria for empowering the analyst. Today almost all enterprise systems have some type of interface which makes integration like the above quite a bit easier than it was back when this story took place. The scissors/glue solution seems so caveman in light of what we have at our disposal today. So, what if we were faced with same scenario today? Taking into account all the advances of the past twenty years, can today’s typical analyst automate the daily creation of that same trading summary report via self-service?
There is no question that today’s manual solution would certainly be more efficient with regards to things like digital cut and paste as opposed to physical cut and paste, but without a means of automated execution and without a means of expressing their intent, the highest common denominator remains at the clipboard. How far have we really come?
This tells me that the impact of advances in technology are limited by their accessibility. We don’t need more technology as much as we need to enable the self-service integration of the technology we already have. Our charge as technologists is more than just solving problems and fulfilling requirements on behalf of those who can’t do it on their own. Enabling accessibility is key. Our charge should include fostering an empowerment system, so next time they can do it on their own. That is how technology makes leaps as opposed to takes steps.
This brings me back around to why I think InRule’s business language is so key to the next generation of empowerment. Once the vocabulary has been created within InRule, it is more than just having self-service access to these systems – it also feeds the ecosystem from which new technology can evolve.
Suppose we lived in a world where the typical analyst truly lived at the top of the empowerment food chain? Suppose they could take a clever idea and were empowered to try it out via self-service? Suppose every analyst could save themselves ten minutes a day on their own? The return on this investment could be measured with a stopwatch. That is an investment I would figure out how to make.