2 of 5:Top 5 Mistakes Collecting SPM / EIM Project Requirements
It is possible to collect requirements out of order, and it often happens. Just like painting a wall, there is some strategy involved. When you paint, you typically want to start at the top and work your way down. The reason is simple, drips. Even the most experienced painters aren't going to be able to prevent all drips, streams and splatters. As you coat the wall you inevitably get streams of paint running down via gravity. If you are painting from top to bottom, it's no big deal. You just paint over and smooth out the drips as you work your way down. But if you are painting from bottom to top, you have a problem. In order to smooth out the drips you have to go back and re-brush some areas that have already been covered. This amounts to a lot of wasted time and effort.
When things get done out of order, it's usually for good reason. Human nature is to do what you can, as soon as you can, with the least amount of effort. When you painted your first wall you probably started at eye level. You don't have to bend over and you don't have to climb a ladder. When collecting requirements, it's typical to start with the easiest and most readily available information. This wouldn't be so bad if everything was fully documented and available from Day One. However, in most case this just means you schedule whoever is available and forthcoming. Beggars can't be choosers when it comes to getting client subject matter expert (SME) time.
The problem is this could be anyone. Maybe the IT guy was available Monday morning, maybe it's the Director of HR or maybe it's a compensation manager. It's a crap shoot. You may need to talk to all these people, but the order is important if you want to avoid wasted effort and rework.
For example, here's a typical week:
Monday: You have a 3 hour session with the IT guy. He goes over the transaction data you will get and you identify of a few extra elements that he needs to add to his list. You leave the meeting happy and fulfilled.
Tuesday: The Compensation Manager clears the calendar for you. Here you finally get a good understanding of the comp plan documents you've been studying. A lot of questions are answered and plenty of new questions are brought up. She identifies a couple of metrics they'd like to pay off of but can't within the current system due to data limitations. You make a note that you'll need to schedule another meeting with the IT guy to talk about these things.
Wednesday: The Finance Director has an open slot so you grab it. She identifies her wish list of analytic reports. Most of these are out of scope, but everyone agrees that you should â€œbuild the system with future considerations in mind because you'd hate to have a lot of rework later.â€ You make a note that you'll need to schedule another meeting with the Comp Manager to talk about these things.
It's starting to look like Week Two will be a repeat of Week One. You're not alone.
Let's start by saying you're never going to be able to completely eliminate rework. The people you're talking to may not have all the answers or they might change their minds. However, there are strategies for minimizing rework.
Here's a basic order to follow:
1. Compensation Calculations - You need to walk through the current calculations as well as future desired metrics. This probably mean visiting with Compensation Administrators who do the current calculations as well as Compensation Managers who will have insight into future requirements (or current designs they'd like to change with the new system.)
2. Reports - Once you have the compensation components down, you can ask intelligent questions about reports. If a step is skipped, you'll just get some reports that look exactly like their current set with some minor tweaks.
3. Data - Take your requirements to the data people and work out a plan for getting it. If there are elements that absolutely can't be had, then you will have to go backâ€¦ but it won't be for a lack of trying. Sometimes the key is to ask. I cannot tell you how many times I've seen Compensation Plan Designers grumble about their data restrictions. And then I later only to find out later that what they want is doable, they just never actually asked for it.
4. Wish List - Wish list items should not drive the requirements process. The reason is most high level requests could be satisfied in a variety of ways. Here, you want to capture the fundamental need (or wish). And then go back and figure out a way to satisfy it within the realm of what is already being designed. Obviously, you can go back and make some new requests if you can't fill the need. But this occurrence should be rare.
This is very high level and simplistic, but the premise should be clear. It's in your best interest to get your meetings scheduled with the right individuals in the best order. The ideal scenario is to run your meetings conference style with multiple groups represented during the relevant discussions. For example, if the comp team wants to pay off of a new metric it might be nice to have the data experts in the room to verify the possibility.