How can I measure the performance of different teams in different departments?

Question

I was hired by a start-up company as a manager to run their operations and their offshore office. There are different teams under me including internet marketing, accounting, IT, etc. My background is IT, and this is the first time I have managed that many teams with different backgrounds.

Since I'm not an expert in marketing or accounts, I was not assigning any tasks; instead, I was giving them targets and projects with high-level details. The team didn't meet my targets; they gave excuses and raised issues. I sense that the team is not being honest, and that I have to do something else to make them do work.

What should I do in this situation so that the teams work on auto-pilot, and I can measure their performance? Keep in mind that I cannot set the KPI of something I'm not an expert in.

Answer 1

You can't compare efforts between marketing and accounting; accounting is non-competitive. I suspect it is an error to try to compare efforts which are dissimilar. The problem set before you is difficult and potentially wicked. On the other hand the only option you can rule out is to do nothing.

I suspect that if I were in your shoes, I'd ask the teams to select and predict their own KPI, and measure them on whether they are in control/able to hit the target they are aiming at.

Ask each department what their goal is, and what their confidence interval is; that is the target, and you can meet regularly to track progress towards that goal. update OP subsequently explained that this is a startup, so I need to modify. Normally the manager has access to the last several years worth of reports, and can calculate the maximum, minimum and median for whatever target they set. The team should be able to explain why their target is different from the median. Since this is a startup, that information isn't available. In many cases reference data is available. Sometimes it is low quality data, but that just affects the confidence intervals. If we were running a mature shop, I would expect the KPI to have a confidence interval of 10%; if we're running a startup, then I might expect a 50% confidence interval. Although I didn't state it earlier, the KPI should be relatively easy to calculate based on available data. You should be able to meet periodically (daily/weekly/monthly as appropriate) and review the current state of the KPI and compare it to the anticipated state. In many cases you should be able to adapt the assumptions & formulas from Earned Value Management. The staff will tell you that their KPI is a special snowflake that has a unique projected value curve. (they should tell you this; if the EVM curve is linear, then they're overpaid). That is fantastic news - that is the opportunity for you to learn their task/business/activity at an appropriate level of detail.

If they can't tell you what their metric is, if they don't have a KPI, then that is a tremendous opportunity; your management will be grateful for any evidence that they're doing work, and you have the opportunity to train them to actually manage their work. That is a problem that is hard, but clear, measurable, relevant and high reward.

Answer 2

I completely support Mark C. Wallace on his suggestions. Once you have the goals set though, you can't just let them go on "auto-pilot". If you set a goal for the next quarter, and then you don't check in with your teams until the end of the quarter, odds are those goals are going to be missed. It's just basic human nature and why, despite it being field that often is not liked, project management exists.

That's not to say you need to hire a project manager, or be a project manager. You just need to be an engaged manager.

I fully recommend the Manager-Tools.com Basics Podcast series. It's about 20 podcasts, though the key for you right now is the podcasts on One-on-Ones. In a nutshell, have a half hour one on one with each of your directs every week. Reschedule instead of cancel if at all possible.

The format is a bit counterintuitive, trust me though, it works. The first 20 minutes is just for the direct. Let them talk about whatever they want, even what fluffy did this weekend. The last ten is for you. Here you can ask general questions that can help you to keep a high level track of their deliverables. Instead of going for the "what's the status on KPI #1" try a more leading/helping question, "Need anything from me for KPI #1?".

That's the 30 second basics. I recommend listening to the Manager Tools basics series. Listen to those and then decide if you want to learn more. They have ten years of free and valuable podcasts.

Answer 3

you could have done nothing to avoid this. after this one, however, you should be more closely involved in their job, so you must understand their problems. you should participate in their daily activities in the following phase, so you can see whether they were right or not about their excuses and issues.

besides this, you'll need to work on unifying processes through all the departments to get comparable KPIs in the future.