When to pull the plug on projects

At one point or another every business will try something new; a new product, a new service, a new process, a new marketing idea. If you don’t I suspect you’re not really growing as a business. And you’ve got your doors locked so no salesmen can come in. And you’re not listening to your customers, either. And you’re not reading the newspaper, listening to the radio, or watching television, reading the Internet, or looking at anything else around you. You probably live in a cave, selling crumbs to cockroaches.

The point is, at some point every business will try something new. That’s a good thing. The only harm in trying new things is a) taking too big a risk with too little forethought, b) not knowing when to pull the plug on the experiment, or c) not knowing what the benefits might be and therefore not recognizing a successful project when it occurs.

Today I’m going to discuss only the second problem: not knowing when to pull the plug. This is when you let an experiment go on and on, burning up time and resources until it start to either drag the company down, or uses up resources that should have been routed elsewhere that would have benefited the company.

The first step in knowing when to put an end to a trial period is to know what benefits you are looking for in beginning the test in the first place. It’s generally pretty hard to notice what you’re not looking for, and therefore quite easy to not notice you haven’t seen it yet, even after a reasonable period of time. Clearly identify what you hope to gain from trying the new idea and how you might see those benefits appear.

The second step is regularly gathering and reviewing results. Are you seeing the benefits you are looking for? Are they meeting expectations? Are the benefits sufficient even at this point to continue? Is there something you can adjust or do differently to boost results? If you’re not periodically studying and discussing your results it won’t be easy to know when you’ve reached the point you need to stop (or to make the changes permanent).

Finally, you need to have some basic idea of how long results should take to appear. You may need to give things a little more time sometimes, but you should at least have a an end point in mind where you know you should have sufficient data to decide. At that point have the guts to decide. The project is either working or it’s not. If not, start shutting it down before it does more harm to the business.

Let me give you an example. In my company we sell and rent used video games and movies. Our stock of movies until recently consisted only of whatever movies our customers traded in to us–very little was more than half a year old. We had no interest in dealing in new release movies. We had a competitor down the street that did that–and did it better than we ever could.

But recently our competitor close their store. Suddenly we had a flood of customers hoping we could supply their demand for new release movies. We didn’t but we promised to think about it. When we did, we decided it might be worthwhile to try it for awhile and see how it went.

We did none of the things I just told you to do. Well, almost none. We did regularly look at how we were doing and discuss whether or not to keep going. But without a real clear idea of what results we were looking for or how long it would take to see them, it has been difficult to decide whether or not to pull the plug. Every week for the last month we keep buying new movies. Every week we take a look at the figures and decide it might be time to put an end to it. Every week we either then see a few customers come in, or we think of a new benefit we might gain, and we decide to give it another week.

We should have identified up front what benefits we expected to see. We should have identified how long it should take to see those benefits materialize. Without that we’ve been flailing around a bit. Granted, the experiment has not been all that expensive, but it could add up after awhile if we remain indecisive about our results.

It’s not too late for us. We can still sit down and identify our expected results and draw a line in the sand as to when we should be seeing those results. In fact, I’m putting that on the agenda for tomorrow right now.