Limiting confirmation bias – the biggest threat to having your biggest year ever.
I know of no correlation between what you chose to drink to ring in the new year, and your level of success in the following year.
It’s safe to say, though, that the Kool-Aid you drink to inform your ongoing strategy and execution will always affect your results. And one cocktail, in particular, can be deadly.
The biggest threat to your campaign success in 2012 isn’t “the economy.” It isn’t a rise in energy or labor costs; rising prices for clicks; large competitors, small competitors, or shady competitors. It isn’t click fraud, hackers, or the weather.
The main threat is within you and your organization. It’s a deadly combination of a losing mindset and fixed beliefs – in other words, the combination of destructive thinking and confirmation bias – that will drag you down, if you choose to sip that nasty beverage.
Especially if you’ve been in business awhile, you’ll often form a priori beliefs and working hypotheses, and then subconsciously work like hell to be proven right. That’s confirmation bias.
If what you seek to confirm largely centers around the assumption that mediocrity is normal and circumstances are bleak, your fixed mindset will be compounded by being locked into an “I can’t” theme. According to author Siimon Reynolds (“Why People Fail“), destructive thinking causes a range of problems from poor business performance to bad health.
Let’s diagnose this compound problem, and look for an antidote. Consider:
- The perils of prejudging data. Last fall, I met an advertiser who had never done PPC until 2011. She insisted that buyers of her company’s building materials were always home users and never industrial or business buyers. She declared that the PPC campaign manager’s keywords and ads would need to be edited to eliminate the “wrong ones.” Unbeknownst to her, that same day, those very (B2B) keywords and ads had produced her company’s first sale from PPC – for one of their higher-ticket items. Why prejudge the data that you’re about to pay Google good money to generate?
- The costs of accepting pricing “norms.” The tens of thousands of media buyers willing to shell out $10, $20, and even $30 CPMs when the vast majority of their communications needs could be satisfied by comparable and available inventory that is (still) selling for $0.50 CPMs. Since everyone at the IAB conferences thinks those are fair prices, that must be the range, and $10 must be the low end of the range. Right? Are you with me?
- The rigidity of making generalizations. How many campaign managers do you meet at conferences who either “always” or “never” use Conversion Optimizer, broad match, the display network, Facebook, dynamic keyword insertion, third-party automation, etc.? Consistency is the hobgoblin of…something-or-other. In any case, that hobgoblin might just be gobblin’ up your profits in 2012.
- The inertia caused by seeking acceptable “industry” benchmarks. Many marketing managers ask around about what a good average CTR or CPA is for a “business like theirs,” to check up on whether they’re going to get in trouble for being too far away from the norm. To cite Avinash Kaushik, “aggregate” is not the name of a single one of your customers. And if you can easily do better without breaking a sweat, then your current number is terrible.
- Confusion caused by low energy levels or impatience in test interpretation. A competent but stressed campaign manager – let’s call her Ursula Underconfident – has been spending a few too many 3 a.m.’s trying to reorganize a messy account, and has lost all sense of hope about the future of it. “I’ve never worked on a stranger account. The ads have no pattern as to when and why they work. There is no logic to this.” Chances are, it’s no stranger than any other account, and you’ll discover the patterns through the same process of creativity and testing you always have. Maybe you’ve gotten too close to it.
Now consider instead:
- The popular online retailer that has grown 30 percent each year over each of the past six years, and was set to grow about 20 percent this year. Instead of pulling back like everyone else, they turned the projections upside down by enjoying a huge growth surge for the end-of-year holiday season in 2011. Industry-wide, e-commerce holiday revenues reportedly increased 15 percent over 2010. Yet these guys grew 71 percent year-over-year that same period. So much for conventional wisdom! If it’s a battle for market share, then sticking cautiously to the 15 percent growth average for the sector could be a formula for losing mindshare and ultimately slipping into higher costs and negative growth.
- “AdWords has gotten too expensive.” So say some “PPC dropouts.” Most costs (rent, shipping, software, etc.) look formidable when your business isn’t cranking on all cylinders. And certainly AdWords is very expensive if not highly optimized. But after all these years, 40 percent of the above retailer’s online revenues were directly generated via AdWords this December, even counting a lot of direct traffic and repeat business. You can already factor a lot of this new customer acquisition into next year’s repeat business. So an optimistic picture of year-over-year growth for 2012 is already forming in our minds.
So what are some relatively simple ways to ward off “limiting thinking confirmation bias”?
Try the following:
- Let the data speak to you. Don’t tell it what to think. Don’t turn things off because the rules say you should.
- Realize that if something “doesn’t work,” it’s possible that it “doesn’t yet work.” A keyword that starts out with a $40 CPA can be reduced to a $30 CPA with a more conservative bid. Throw in ad testing, negative keywords, further bid tweaking, Quality Score improvement, and further analysis, and feasible next stops in the downward CPA trend might be $30, $15, $14, $9, and finally, $7. Does that scenario sound typical to you? No? Yes, this sets the bar high, but not impossibly so. Having some visualization of the potential is helpful in breaking out of limiting beliefs. Of course, if something is putting up goose eggs for an extended period of time, you do have to give up on it.
- Have a colleague work on a part of an account that is frustrating you. Chances are their new ideas may combine well with the base you’ve built, and the account will move closer to its goals. Sometimes it’s as simple as the second person not assuming something won’t work. Maybe conditions in the business have improved. Maybe the combinations of match types, ads, and bids they try lead to a better initial outcome and faster learning. Sometimes, what you don’t know can free you up to succeed. (Don’t confuse the above with second-guessing or campaign management by committee. That just leads to a mess and obsession over small failures.)
- Use tools. Let a tool (such as bid automation or a keyword tool) take the load off you so you’re not so worn down by rote work that you fail to take the time to analyze the strategy. Why not work with the power of machine-learning and automated data crunching rather than trying to prove you can come up with everything manually? Take the AdWords keyword tool, for example. Maybe you’ve developed an irrational pride in your ability to use 28 third-party tools in place of Google’s. By all means, use a range of techniques. But use Google’s keyword tool constantly. It’s serving you both positive and negative keyword information based on tens of billions of real-world queries – and may even incorporate user behavior after the click. (That being said, don’t let the tools take over. You’re in charge, and accountable.)
- Frame your professional efforts with inspiring or informative background research. When was the last time a dentist, stockbroker, or member of the clergy felt alone in their profession? These professionals are steeped in, and ideally inspired by, their “literature.” “Zoom out” by making yourself aware of business case studies in your sector. Not so you can worry about competitors, or lazily chase certain benchmarks (Tony Hsieh probably won’t be discussing his CTRs and Quality Scores, anyway), but so you can inform your own unique approach. If you’re in e-commerce, for example, you may find inspiration from the Zappos’ founder’s book, “Delivering Happiness.” This book reveals that the legendary company had many struggles, nearly went bankrupt several times, drank vodka on the rooftop when things went right (and sometimes, when they didn’t), and had a different way of calculating customer lifetime value than most retailers do. Zappos won by staying obsessed with the needs of its customers and developing its own unique culture and imprint on the business world – not by following someone else’s formula.
- Use an experimental method that is falsifiable and “real world” enough to address complexity. In your spare time, run AdWords Campaign Experiments (scientific A/B tests in real time) to transform speculation and hearsay into a definitive knowledge base of how complex campaigns truly operate.
- Use imagination to think forward. There are hundreds of ways imagination comes into play – in crafting a subtly better call to action, for example. You can also use it to give you hope. Looking forward to the fast seasons and thinking about what customers will suddenly be hot and heavy for is a good way to keep you planning and working for that time. Getting bogged down in “poor” results in off months, and thinking up ways to get defensive about them, is a waste of energy. For slow times, do your best to replace debate and subjective assessments of how hard you and the team are working with dry, solid facts and graphics that refer not only to recent trends but year-over-year patterns.
- Have fun, most days. Associate your pursuit of goals with rewards, not punishments.
Our job is to tame chaos, not use it as an excuse.
Our job is to achieve growth and greatness in 2012.
To achieve that, we have the luxury of team collaboration, tools, and especially enormous amounts of rapid market feedback at our disposal. When the evidence allows us to make a better decision, we should take that as a gift, and we should keep doing that again and again. Relentless optimization gives us slightly better odds of succeeding each and every time we activate that lever.
Stack all those “slightly better odds” in an orderly pile, and you’re left with much better odds.
It never fails.
Originally posted on ClickZ – Don’t Drink This ‘Failure Cocktail’ in 2012