Managing PPC can be a tough gig. Goals grow, strategies change, engines constantly “improve” their platforms, etc. It’s a special kind of madness that only mad people can love. It also requires a type of patient urgency that few have.
You have to have the patience to let tests play out, wait for new clients to come in, and to understand what internal and external influences are impacting your results. You have to have an urgency to act when you have enough information and be humble enough to adjust when new information is available.
To muddle through this enigma we call PPC, I’ve found it useful to use the 3i’s. They keep me grounded, help affirm when I am moving at the right speed and give structure to a chaotic process.
Idea
It’s not as simple, or at least it should not be as simple, as sitting in a room, thinking of everything you could do, scratching out the things that won’t impact the goal, and doing the rest. This is where over-optimization happens. This is where the illusion of improvement happens. And this is where stress/depression/lack of insight into what really impacts results happens.
Ben Horowitz, famous VC dude, wrote about this in a larger sense in a post titled Little Things. For me, having a solid “idea” is about getting close to the information, understanding what actions have been taken (internal influence) and understanding what is happening in the real world (external influence). Only once this is understood will the “idea” you come up with have real validity and a greater than random chance at success. This might take a few weeks. That’s OK.
Iterate
Guess what!? After you do all that, have an epiphany, flawlessly execute it, and kick your feet up, it’s probably not going to work the first time. You’re going to have to look and listen to the impact the idea made. Filter out the noise of factors outside your control and get to the signal that tells you what to do next.
This is about iterating your idea until it actually works. Why not start with a better idea? It’s impossible. Or at least it is mostly impossible given the constraints of time and information available. Jonathon Bendor, smart Stanford dude, tells us that we’re better off breaking big problems into more manageable chunks and iterating our way through the process. He tells us that it’s about breaking things down, then combining them in new ways until we’ve actually solved the problem.
Don’t wait for statistical significance at this point. You are looking for logic. Does it make sense that the idea as implemented should work? Can you see logical things to change about the idea that could have a big impact? Do you have new information after having some time, or getting early results, that changes some of your previous thinking? This step should be measured in days, not weeks.
Impact
How do you know you’ve actually solved the problem? 3-5 consecutive time periods with performance better than your upper control limit. Both how many consecutive time periods and how long those time periods are depends on your patience, the size of the account, and how much it matters to get the decision right.
I’d say each consecutive period should have 1,000 clicks, or 100 conversions (depending on the size of the account). If you get more than that in the campaign/account you are implementing this idea on then you can measure it by days (assuming you get relatively consistent performance throughout the week).
For ideas that are less metric driven or have super small numbers, such as a business decision to lower the price you charge for your PPC services and you only sign 1-2 new clients a month, you are going to have to consider leading indicators (a fourth i, perhaps!?).
For most accounts, this step can take 1-2 weeks before you find statistical significant success.
Combine these three steps and you have 2-6 weeks, although an initial idea was implemented at 1-2 weeks. It may seem long, but I’d argue not being patient at the front-end costs you at least as much time and you gain far less understanding.
