Let's look at some simple analogies to see what needs to be considered. Rather than think forward as a project manager would, let's trace the process backwards and assume we didn't plan for quality at all.
Suppose you are the Coyote working in the Acme bomb factory and want to test the quality of the bombs (let's just say that blowing up is sufficient). You will need to spend project budget on a bomb proof room so the project team isn't blown up. If you test 100% of the bombs, you will blow up your budget, timeline, and full product line testing quality. So a better approach may be to take every hundredth bomb (or a statistically significant sample) off the production line and make sure it blows up. That assures you have plenty of bombs to sell and you will just need to make sure that the cost for the 1 in every 100 that is unavailable is included in the selling price.
Now you are happily blowing up every hundredth bomb, but the client complains they are not getting enough bang for their buck. You will have to expend money on figuring out how to measure the bang and equipment to do that. Your time spent on this effort causes project delay. Delay that could have been avoided had you proactively planned for quality.
So now you find out that 50 out of every 100 bombs tested do not have enough bang. You are going to have to inspect the bomb requirements, design, and other specifications to determine how the components inter-relate. You need to take apart a few bombs before they are blown up and check the quality of the explosives, the casing, the fuse, etc. to see if they are within tolerance. You now find yourself negotiating with TNT Explosives Company, LLC to improve the quality of their product so your bombs will have a bigger bang.
But wait! Suppose you didn't find anything wrong with the components. Now you are going to have to find a bomb expert to review the design specs to make sure all the components were properly combined and assembled to get the right level of bang. More cost and more delay. Or you inspect the client requirements and find that there is a conflict between the casing they insisted was required for the bomb and the amount of bang they wanted. You are going to have to sheepishly explain to the client at this late stage of the game that their requirements were faulty and you need to change the bomb casing. The casing costs 200% more and you are going to have to pass along that cost.
At this point, your client is going to call your competitor to find out if they can build the better bomb. So what have we learned?
- It usually pays to plan and build quality in from the start, because the further along the project gets in time and phases, the more expensive it is to fix a problem.
- Time and cost need to be considered in developing the quality plan -- if you blow up all the product to test it, there is nothing left to sell. If you perform exhaustive tests, you will exceed the timeline.
- Quality itself is a cost, and you won't be able to determine the fair selling price that permits a profit without knowing the cost.
- The right resources with the right training is important. QA cannot be learning the product on the fly, since personnel is also an expense. You need to get the right tests the first time to be sure the quality is sound.
- Since time, cost, and quality are like a triangle, a short cut in any side will force the possible lengths for the other two sides. If you demand the product be produced quickly and cheaply, you can only get the desired result by cutting back on the quality.
So to get the best bang out of your project, quality needs to be planned up front and meet reasonable parameters. You will need to negotiate this balance with the sponsor, the client, and other key stakeholders.