“Computers are useless. They can only give you answers.”
Does anybody else have the feeling that we just have to absorb too much information lately? Between work, the Internet, media, etc. it seems that we’re subject to so many topics, facts, trends, analysis, and live coverage that it’s a bit like trying to get a drink of water from a fire hydrant.
For example, the MSN home page shows the five most accessed search topics daily, and then lists the top five suggested searches. Today’s offering, for your consideration:
Five most popular searches:
The top 5 suggested searches, and I kid you not:
Yikes. What ugly, twisted mind at MSN dreams up these suggestions? Do we really need this much to think about on a nice day? There is such an overload of information that the sophomore girls at Central High, with my daughter at the forefront, now play a game called “Googling” where the goal is to enter a string of search words that produce exactly one result. Hard game to win.
Anyway, this brings me to the serious topic of the day, information management at banks. At Cornerstone, when we start a system selection process, or a planning process, one of the things we always, and I mean always, hear in our initial meetings is that the bank is unhappy with information management. This will encompass the amount of information available, quality of data, ease of report writing, ability to analyze, ability to share, etc. At the end of the day, the result is that senior management has a distrust of the information they need to make decisions and a belief that there is a real opportunity to reduce costs if the entire information management effort can be improved.
This is despite a very concerted effort by vendors to provide information management solutions. The legacy mainframe vendors all have designed SQL or Oracle data warehouses and port flat file data to them for reporting and analysis. The iSeries and Server-based vendors all tout the relational databases in the core system and have put various report writing tools in front of them. Everybody flaunts richness of information and ease of use that result.
So how do we get serious about fixing this disconnect? Somewhere in between the easy tools and the unhappiness with data is the hard work of making information management better. I propose these ideas:
Four Goals for the Vendors
1. Re-examine the report writing tools.
The “we have a relational database and a Crystal” pitch is meaningless to the client. I have sat through many demonstrations of information management solutions and those “easy, intuitive” report writing packages aren’t easy at all. They range from lousy to OK. Most of the time, if a system presentation actually includes designing a report from scratch, anybody outside of IT or maybe Financial will walk out frustrated with how hard it looks to link tables, write and/or condition statements, format reports, put them into production, compare current numbers to a prior period, do any trend analysis, or perform many other steps that are at the heart of their needs. And I agree with them.
I think every vendor has to challenge the quality of its front-end report writer to alternatives available in the marketplace. Short of that, one way to help is to build templates for simple list/total needs. Another is to build some templates that can be re-used to automate formatting. A third, if you haven’t done it, is to change field names from COBOL-based terminology to something people understand – no more “sddbal” stuff.
2. Push the building and sharing of report libraries harder.
Every vendor talks about how it facilitates the sharing of reports that have been designed by its clients, or ones that it has developed, but I’m always surprised at how seldom this seems to resonate at the desks of the bank users. The reports are there to share, and an intranet is custom made to facilitate the sharing. I think this is really a matter of more proactive selling of the benefit to clients
3. Make Excel a seamless import/export.
Forget what anybody says. The reality of report writing tools is that they don’t come close to Excel in terms of formatting look and feel. There is an incredible amount of work being done using report writers simply to get data to Excel for sorting, changing field width, shading, italicizing, font changes, and printing. And your users are hooked on Excel like a ‘60s Berkeley student was to the water pipe. If your demo shows how this transfer is automated, you’ll get supporters in the room quickly.
4. Help banks figure out calculated/derived fields.
There is a huge frustration among users who want to calculate a field and use it in a report. Example: the core system has a beginning date and a term for a CD but the bank user wants a report that shows remaining term. Dead stop. Nobody can do this, and anybody who could figure out how to use your tools to accomplish this type of task is already in Houston working for NASA. Why not develop those calculated fields in the database and make them available, rather than forcing users to use the reporting tools themselves to accomplish this? You’ll be heros.
Four Goals for the Banks
1. Get serious about the number of reports floating around the bank.
This is a broken record for us, but there are far too many reports in every bank. This is the result of the afore-mentioned Excel junkies in each department coming up with their own slightly altered versions of sales or balancing reports that are often created for reasons that are less than essential. It would be a great idea for someone in IT to referee a gathering of such reports. I’ll bet the total pile will be higher than you might think.
2. Set standards for the quality of reports and enforce them.
Maybe my start-as-the-teller background is showing, but we need to get back to the premise that REPORTS HAVE TO BALANCE TO SOMETHING. The total of loans outstanding on the sales report ought to look the same as the GL number, n’est pas? Banks need to get serious about enforcing quality control on new reports. Take the review role loan committee plays on a new large loan deal and apply it to a group that reviews a new report that will be circulated. One meeting a month that allows reports to be challenged for accuracy/quality would probably pay big dividends – and that same group can begin challenging the creation of what are essentially duplicate reports.
3. Stop using Excel and mini Access databases as systems of record for data.
There is far too much data stored only in Excel, not in any database. Because of the calculated field problems in most core system solutions, most people use Excel to create them. They then use Excel to store them. Often, users enter data into Excel from a non-core source and that becomes the final resting place. For those who are more energetic, Access has become popular for building a quick database that solves one particular reporting need. My partner Carl Faulkner has related several instances where he has visited banks and seen dozens, if not hundreds, of Access databases in various departments without any overall knowledge at the bank of what they are or what they do.
Has everybody heard the “islands of data” term? Nothing will cause this problem more than the same data stored in multiple Access databases and Excel spreadsheets. With IT in the lead, banks need to start determining how to build single, common, data sources that people can access and download to Excel or other desktop tools. It will reduce work and allow Excel to return to a serving as a tool for formatting and analysis rather than a de facto database.
4. Get thy sorry butts to training.
No elaboration needed.
And One Joint Goal Figure out who will take the lead in allowing users to access multiple databases/systems more easily. For example, many employees want to do analysis that requires information from a core system, a branch platform system that has data never sent to core, a loan origination system that has data never sent to core, and profitability information in a system in financial that is entirely stand-alone. Nobody has really figured out how to do this well. The mainframe vendors that have developed stand-alone data warehouses may be a little ahead, because it is easier to build tables that download from multiple systems and link tables, but we have never seen this eliminate a lot of work being done manually when reports require data from multiple systems. Vendors and clients need to jointly examine how there can be a better combined effort to automate reporting that involves multiple sources.
Solving the information problem isn’t about the big vision anymore. It’s about addressing difficult issues that will require hard work to resolve one by one. But at the end of the day, the senior managers who sense a big information and productivity payoff if this is done right are onto something.
“Call out the instigators, because there’s something in the air.”