What functions will a Chief Experimental Officer need to do to run hypothesis-driven experiments?
When “growth hacking” first entered the scene, it provided a fresh, practical approach to marketing. Smart engineers coded innovative solutions that generated free viral shares. Stories about Dropbox’s storage rewards for sharing and Facebook’s famous growth team became models for companies everywhere.
Eventually, though, too many marketers trying to cook up schemes in growth hacking ruined it. Startups piled on the trend, slapped the term on everything they could, and turned growth hacking into a spam technique. Coders took shortcuts to scrape addresses and send semi-personalized emails, lowering the quality of communication and letting nontechnical users further dilute the practice. Every entrepreneur who managed to put together some funding tried to hack success to justify the last round.
Today, growth hacking is at best meaningless, and at worst, a direct disservice to growth. We have too many options, not enough smart data for prioritization, and no proven processes that capture the dream of growth hacking without falling prey to its faults.
We need to transition to scientific marketing — and we need a new kind of executive to lead that shift.
A Replacement for Growth Hacking
Scientific marketing simply means applying the principals of the scientific method to marketing. Ideate, test, analyze data, adjust, and test again. It’s a philosophy-driven approach that values consistent results over the empty promises of instant, hacked success.
The only growth tactics with staying power are those that no one else puts in the work to accomplish, sometimes in channels, everyone considers oversaturated. Unfortunately, adopting scientific marketing doesn’t quite fit into the current C-suite structure. Part marketing, part technology, and part information, this new approach requires a new type of leadership to succeed.
Enter a new kind of CEO — the chief experiment officer.
The Role of the Chief Experiment Officer
A chief experiment officer will own all digital and direct attribution activities and should be agnostic regarding established channels and ideas. Chief experiment officers also own funnel strategy and lead a growth team, spending less time managing existing strategies (something better left to existing C-suite executives) and more time disrupting the status quo through explorative, data-driven strategies.
As a new breed of executive, the chief experiment officer will evaluate opportunities, find smart people, empower them, and get out of the way.
We need new officers because while the best tactics are easy to test, most are difficult to implement. Throwing everything at the wall and seeing what sticks is not a strategy. Chief experiment officers will have limited resources and therefore must focus their attention on overall strategy and the tactics most likely to accomplish the company’s goals.
Following the concepts of tools like Facebook’s PlanOut open-source testing framework, chief experiment officers can make fast, unencumbered decisions that will keep the company moving forward without sacrificing precision or due diligence.
Shifting to a Scientific Strategy
Your company and its new chief experiment officer must focus on a few key areas to succeed in driving effective marketing initiatives.
1. Plan marketing experiments around the biggest opportunities.
Chief experiment officers will operate at a high level and should not be bogged down with low-ROI endeavors. They'll create experiment-driven marketing plans based on the biggest funnel opportunities and the most significant key performance indicators.
The strategy of the experiments should follow the scientific method. Choose a specific KPI, identify the desired result, test the hypothesis with an experiment, evaluate findings, and double down on what works.
When one of our clients came to us with a limited budget and wanted to decrease the cost per install, we ran an experiment to find the lowest install times, and then removed ads from those times. Looking at data from the previous 30 days, we found that (on average) on Mondays from 12 a.m. to 3 p.m. and from 10 p.m. to 11:59 p.m., our client was receiving zero to one installs or the installs cost more than $1. We hypothesized that by not showing our ads during these hours, CPI would decrease, and we learned that optimizing the $180 daily spend by removing the least effective hours would decrease CPI by 12 percent.
Chief experiment officers will identify such areas of opportunity and act quickly, saving companies the frustration of trying ineffective strategies on well-trodden channels.
2. Write drunk, edit sober.
Brainstorm recklessly, come up with ridiculous ideas, and explore interesting new topics. Creativity does not flourish under constraints. After the creative process provides a starting point, ruthlessly prioritize which pieces to cut.
Another client of ours saw a major drop in completed cart purchases and wanted to convert more sales. The issue was that the cart popped out on the side, which provided a nice-looking user experience but didn't inspire action. When we discovered most users only bought one item, we had the crazy idea of almost forcing them to convert once they added an item to their cart. While many might find limiting time spent on your website to be counterintuitive, we tested a locked-down cart experience that redirected to a cart page with no menu buttons to guide customers' focus toward conversion, similar to Amazon's purchase path.
Conversions increased by 30 percent, validating that some of our more outlandish ideas often eventually lead to the right solution.
3. Implement a tactic agnostic strategy.
The goal of your experiments should be to create long-lasting, sustainable change. That means no juicing the numbers with fake follower counts and bots. But don’t be averse to paid marketing, as growth hackers were in the past.
While we initially viewed Google AdWords as a saturated channel, we tried it anyway and continue to land a few new large clients per month with great ROI. Remain “tactic agnostic” in your experiments to discover which approaches work better for various KPIs and positively impact growth for the least cost. Rely on your user data to refine your approach and create a library of effective tactics across channels.
If you’re still missing the people you need to succeed, put growth hacking behind you and bring on a chief experiment officer. Follow these strategies and you can achieve your goals faster through scientific marketing.