Pay-per-click is a channel that’s easy enough to understand, but difficult to master. I know because I’ve been wrestling with PPC campaigns my whole life. That’s actually not true. More like the past 5 years. The former sounds better.
Regardless, I’ve developed my own philosophies for PPC management over the years, and I hope you find them useful.
1. Pay-Per-Click Can Be Intensely Personal
When you sit down to manage a pay-per-click campaign you have to remember that there’s someone, somewhere, doing the exact same thing. You two are taking actions that affect each other in a weird advertising tango.
Because advertisers compete in an auction where the advertising space is finite—every bid and strategy you decide upon directly affects the performance of everyone else’s campaigns. The inverse is true as well—the most beautiful campaign that produces conversions and makes you look amazing can turn to shit when a competitor decides to outbid you; it’s hard not to take that personally.
2. Bidding On Brands Is War
Think about the this way works: someone’s searching for “Joe’s Ice Cream” and you show ads for “Sally’s Ice Cream.” Bidding on other people’s brand name is a jerk move. This is what experienced pay-per-click managers will think when they see brand terms being bid on.
First, typically when someone’s searching for a specific, proper noun thing—they want that thing and only that thing, so functionally these campaigns usually don’t perform very well. The basic thesis of all pay-per-click is a tight relationship between the original query and the ad.
Second, you increase the cost of advertising for someone to bid on their brand terms, which are typically the cheapest of the cheap when we talk about cost-per-click rates.
3. Seasonality Matters
Every industry and business has an aspect of seasonality: a period where demand is different for a type of product. Seasonality means you sell jackets in winter and shorts in summer. More importantly, it means that most pay-per-click campaigns can’t function year round. If you do have a campaign that produces all year, it likely needs significant adjustments to budget or geography to work.
4. Some Factors Are Beyond Your Control
Just like you can’t control the weather, you can’t control certain factors that affect your account and campaigns. Some verticals are prone to bad quality scores, bad weather, violently fluctuating bids, and shaky competitors.
At some point your campaigns experience a turn. Notice “a” turn instead of “the” turn. It’s the point at which things go sour and what was a shining beacon of performance has turned into a trash pit where money goes in and clients go out.
5. Forget The Settings, Focus On Keyword Intent
People get lost in the details of AdWord settings, much in the same way that people get hung up on technical elements of search engine optimization. All trees and no forest. What matters, what really matters, is keyword intent.
First, keywords are just words. Words that represent ideas, needs, and desires. Projecting yourself in the words that people use and getting that right is one of the most challenging aspects of pay-per-click.
“Keywords are just words. Words that represent ideas, needs, and desires.”
Second, some keywords are bad—not because they’re intrinsically a terrible choice, but because they don’t fit the task. Whether a keyword is too broad or doesn’t get enough clicks, some keywords can’t be forced into working.
6. Impressions Are The Lowest KPI of Success (So Find A Better One)
In my experience, there are people that believe that impressions are some measure of success. Most of the time, it’s not. Producing impressions is one of the easiest things to do in pay-per-click since even terrible campaigns filled with junk keywords and low-position ads can produce impressions.
Unless there’s a heavy branding component, celebrating impressions is like brushing your teeth and acting like you won an Olympic medal for curling.
You need key performance indicators that hit closer to the project goal: conversions, interactions, telephone calls, form submissions, etc. If you don’t have those, then you start eating through campaign data to find a metric that proves you’re getting closer to make conversions happen.
7. All New Campaigns Are Awful
There’s an unwritten rule in my mind that new campaigns are awful. They are awful because most untested ideas (and your campaign is an idea) are shaky. It’s one of those “if at first you don’t succeed” things.
But there are a few structural things that work against new campaigns:
- New campaigns often come with harsh quality scores.
- The tools are limited for gauging competitiveness and cost per click.
- Landing pages are fickle. So is branding. So are prices. So are products. If those are jacked, then the campaign is as well.
Iterative change is what we’re after. Rome wasn’t built in a day.
8. If There’s An Option, Use It
I like to sneer my nose at new features. A bunch of hooey! But a lot of new features have been rolled out throughout my career (e.g. ad extension, keyword insertion, URL parameters, etc.) Every new feature is an opportunity to experiment—a chance to go from ambiguous moneypit to successful, profitable campaign.
Turn on those sitelinks. Use every client as a chance for experimentation (hell, don’t even tell them). Move fast and break things.
9. Think About Customer Segments + Geolocation
Some businesses are easy to understand apropos geography. If you’re a baker and you sell doughnuts in Raleigh, NC then it makes sense to target Raleigh and the surrounding area. A local business for local goods.
But imagine advertising apartments or real estate or attorney services. You might have campaigns that target potential customers locally, but regional campaigns, national campaigns that seek to draw in various customer segments. In thinking about the real estate example, you may have a local buyer that recently got a promotion and is looking to move to a bigger home.
“Some businesses just fit with a certain model of geolocation.”
In another case, you may have someone in California looking to buy a second home on the coast. Each customer segment that you’re able to verify (through campaign performance) gets you closer to not just good campaigns, but entire models of customer behavior.
10. Cut The Fat
There’s fat in every campaign. Keywords, ads, campaigns that just don’t work and are allowed to thrive overtime. Like overgrown weeds in a garden.
They say good design is taking away all the things that don’t work or belong. I’d argue the same for pay-per-click.
There’s always the dream of making of the perfect campaign that doubles your client’s return-on-investment and has 10/10 quality scores. The reality is that pay-per-click is about accounts vs. single campaigns. Single campaigns get good by taking away the things that don’t work.