Back when we were all and green on process change management being a key aspect of IT business performance management projects, we were referred to an Oil Exploitation Company to develop a few key performance reports using data extracted directly from the company’s financial and operational al databases.
About mid-term in the project we were discussing the progress with the chief sponsor and expressed a concern that buy-in was taking too long. He laughed and said,
That moment of truth allowed a project low profile project to bring about successful business change that removed reducant processes and added significantly to the business growth. It was a turning point to learn how we determine the delivery value management of similar projects.
When large retail bank called us to tell us they were falling behind the market, it was not to a request to come and build some reports. They knew the key information they needed, to react to a much faster responsive market was not available until way past the use-by date. That was directly threatening their ability to make business decisions and remain competitive. “Busines is walking out the door and we don’t know it until is too late,” the CEO told me and we need to fix the processes so we get our information fast
Be it a bug infestation or blocked arteries shows how information, when throttled, becomes is almost valueless. This picture shows response time value of the week to week business data of the Bank when it is available 3 weeks into the next month.
Best Response Time Value
The Bank had what seemed a quite a different problem to our Oil Barons, who had bad data that constrained timely usefulness of reports. In the end, the Bank was just not getting it fast enough so they could plan.
This is the story about how we approached their problem.
When we looked at what the Operations and Executives used to manage the business, we found the just used lists of products with performance variances totaled by category, segmented for each branch and channel of the business.
Nothing fancy, no charts, no dashboards, no blurb, just simple columns of comparative product detail.
Each row had items showing month and year to date values other with budget and latest forecast values.
It looks very detailed, but once you see the patterned style of reporting larger % stand so it is easy to spot issues and get a sense of what as going on.
Simple as Simon says, isn’t it? the COO said. “And we use it to pull some levers on hot spots, identify non-performers and fix poorly trending items to get them back on track.”
When we asked how their latest forecast update process worked, the COO said, “Managers re-assess product balancing weekly and do month
and quarter updates to show how they will achieve their target. They do this by writing the new numbers on items on the page, then their assistant keys that information into the system. and it takes less than a day a month” “Simple isn’t ,” he said with a smile, “
To make any suggestion otherwise seemed like it would be borrowing his watch to tell him the time.
He went on to say, “Things like optimizing competitive price and position are local issues that on-the-spot managers gauge well. They are very skilled at using the on the ground information to update the forecasts. Plus we have benchmarks and intelligence in our database they can also refer to .
Tactical planning, however, is driven highly by competitive positions which need up to date data. Hence getting the lists out earlier is our issue. In a pinch, we can make so with the walk around, but in the end, we need the data to remain on top if it all. And then we can to plan well and take decisions to stop erosion before it starts.
Most in the retail banking world understand it is also as much about the social environment as it is about product and price. Once customers desert in significant numbers to fall below critical mass, the Empty Store Syndrome sets in and death is almost impossible to stop.
What we want is good data early enough to react to it – no more no less.
Before our COO meeting, we had been with the Finance team who prepare all the reports. They told be us their process to deliver reports meant extracting data from the data warehouse, then cleaning it up by sorting out processing errors before updating organization information.
Once done they merge all this with offline comparative data including the budget and the previous year to date in summary by segment and product.
This was all done in aggregation spreadsheets that had them to prepare and send out profit statements and the compiled lists with of all the relevant comparative product performance information. This took until the third week of the month, and they told me a lot of the time was rework for organization changes that the system had not caught up with was needed.
Earlier in the day, I had also met with two key Business Unit Execs who told me their frustrations with accessing the often incorrect transaction data in the data warehouse. Their concern too was waiting too long for information, that validated what they suspected but arrived after the horse has bolted so they had created their own process.
Roughly the AS-IS process looked something like this:
The solution to fix seemed clear. It needed an analytical tool to pull it together fast in so everyone could look at it early in”
Te COO respose was sharp and succint based on streams of BI salesmen had obviously been in there over time trying to sell them BI projects that would fix all their ills. “We are not going to spend millions and take months on a BI project. We need this now and anyway, the issue is not the reports, but the processes compiling them”.
My response was equally succinct as I could see their confusion
“Any competent person with the basic script to query the data warehouses should have something running in a couple of weeks to start managing improvement”.
So a full BI project is not always needed. What is first needed is the data pulled into an analytical cube and connected some form of analytical off the shelf reporting tools so everyone can see the results. Drill down to go to the details is important and even the transaction sometimes. With such capability reports can then total and do variances along with and other comparatives all in alignment. That is what makes it all work so well, not the glitter and glitz.
It will not fix all the problems, but what is so powerful is they show up. Then everyone sees all the issues at once and will work to fix them to get fast results. And then all the rest will just follow.
We knew from our Oil company that the when lights go the cockroaches run so you can get busy solving the real issues.
Once we started we knew if we live by the sword we could die by it too”, but was fun to see the stuff that runs out from under the rocks as the soon very excited managers who could see their data before it got touched. Miraculously too , like the Oil company data, it magically started to bi-pass the bottlenecks and cleaned itself up.
For our Bank, their new process looked like this in a matter of weeks. And just take a look at those faces :
Before buying a business intelligence system; first check with your IT. The chances are you already have one and don’t know it. But even if you have to go out and buy one, go get one that can start for dummies first and grow with you as you learn how to handle analytical format so it is not a big deal.
In the end, you will need something that can also answer questions, or let you ask them, as part of you total performance management. All that good stuff will naturally follow without any fuss if you get it started with some early results.
Gordon Wood is Managing Director at Sherwood Group Consulting www.shernox.com