December 4, 2012
Someone really smart once said that smart people learn from their mistakes and wise people learn from the mistakes of others. I’ve hit the ball out of the park my fair share over the past 12 years but I’ve struck out swinging, too. Here’s some take-aways from those mistakes.
1. It’s getting A LOT harder. CPC’s go up, CTR goes down and conversion rates are getting worse. The good news is that the tools are getting a lot better. Anyone remember the Google Content Network a few years ago? Quality was horrible and the tools at your disposal were meager. The Display Network is much improved, easier to use and in my experience, provides a great ROAS than the system of ole.
2. Properly geo-targeting can cut out thousands of wasted clicks, spend, poor quality leads, etc. It can take a fair amount of time to get good data regarding what Regions work best for you so many people keep on eye on it for the first few months of a campaign but then eventually forget to keep checking it. I’d say you should have about 3,000 clicks or 50 conversions from a Region before you decide if it works well or not. It will depend on your account though, the important thing is to find statistical significance before making the decision to pull back or push a specific geographical region.
3. Completely shutting an account off for periods of the day (i.e. day parting when your call center is off) has an adverse effect on your PPC during your peak hours. I find that sure, spend is pulled back, but that you lose conversions from people who come back after work to purchase or who are in the research phase and will buy tomorrow. Instead, I choose to bid down by 25-50% during off peak hours. That way the ad still shows and can pull in those researchers who are up late at night surfing but not ready to buy.
4. The more things change, the more the stay the same. I was talking with our CEO the other day about a strategy of creating top performing campaigns for a select set of keywords that drive the most conversions and he said, “I called those Heavy Hitters back in 2004.” The exact tactics and tools we use change but the strategies that worked when I got into the industry, seem to generally still work even if it takes a bit more elbow grease to get there.
5. PPC doesn’t work in a vacuum. Phone calls, email follow-ups and organic clicks all improve with a heavy focus on PPC. Also, there is information that your sales department has about the quality of leads, number of fraudulent orders and extra conversions that don’t fire your pixels that you should be using to optimize your campaigns. The only way to rectify this is to create systems to track activities outside of Google Analytics and AdWords, adCenter tracking codes. Then you run reports and also talk, often, to your sales department to hear what they are seeing.
6. You can’t spend enough time on achieving goal clarity. ROAS, ROI, CPA, CPL, budget, etc. surprisingly mean different things to everyone. It’s frustrating, but it’s true and if you take any metric of KPI for granted you’ll likely run down a path that takes you the wrong direction.
Good clarifying questions include:
- How have your goals changed over time? What causes them to change?
- What’s the formula you use for calculating ROI/ROAS?
- If you could only watch one metric, which would it be?
- How frequently do you analyze goals/performance? Besides reports we pull what are other ways you determine the success of your PPC campaigns? Can I have access to those reports?
7. No matter how well you set goals, expectations will shift. A.K.A. have an iron gut. No matter how well you do, your boss or client will want more. It’s easy anyway, right? Be ready for the curve balls, the spontaneous reports that shatter what you thought was working and the immediate shift in goals after you hit them. It’s part of the PPC game. Learn to enjoy the challenge!
8. Determine Upper and Lower control limits to primary metrics and manage to those. This means that don’t just manage to fixed number but manage to a range that indicates the type of performance you are looking for. This reduces stress around a bad day or two and helps you spot issues quickly that you might not notice if you look at too long of a rage. This can be very complex. But you can also simply determine the range of daily volume/CPA and accept that range as variance from goal in any given day.
9. Never say no and never lie. Seems obvious but those two aren’t always easy to do at the same time. It’s important that if your boss, a client, your significant other asks something of you that you go to “yes” first and then objectively determine the obstacles in the way. Once you understand that picture you can come back with the “Yes, we can do that if….” and lay out the conditions necessary to achieve the desired outcome.
10. Ad testing is usually done wrong or not done at all. Some account managers start tests and never conclude them. Others test ads once or twice a year. Even more do tests, conclude them and start new tests without actually learning anything from the test itself. After enough clicks, impressions and conversions have passed I like to download my ads and not only find those that perform the best, but find the individual messaging that performs the best. I do this by concatenating the description line 1 and description line 2 and then doing a SumIf function to see which message that I am testing has the best overall impressions until conversion. I can then report that info to my boss and client so they can use it across other marketing channels.
11. It’s not usually about what you do but it is about how you do it. A bid change can be disastrous, neutral or making a significantly positive impact on an account. What matters is not that you do a bid change but how you do that bid change. Did you take in to account seasonal performance from last year’s data or Google trends? Are you looking at multi-channel funnels to ensure you’re not cutting off volume from an influencer? It’s easy to accomplish a task but it is harder, yet more effective, to focus on the actual intended outcome.
12. Best practices don’t always apply. I’ve seen campaigns that you did what all the experts say to do and have a complete failure. I’ve also seen accounts with one campaign opted in to search, display, mobile and desktop and then have nothing but broad match keywords perform excellently. I’m not saying do that but I am saying that every account has its unique personality and trying to force it into being something it is not can tank performance.