Brian Massey - Everyday Behavioral Science For Better Conversion Optimization

Editor's Notes

• #3: Two similar products Four star rating with 250 reviews Five star rating with 5 reviews. Which do we believe? [audience interaction] This is a 4 star product. We know intuitively that a larger sample size is better than a smaller sample size. Statistics: n for number in sample. Statistics is built in. What else can we infer from this? [audience interaction] The first product hasn’t been on the market for very long or isn’t selling well. The population of buyers is probably small. Statistics Big N=population size.
• #4: Hope: You already know how to do behavioral science. It’s built in. Likes and comments on your facebook post Decided to buy shoes based on star rating Chose that espresso machine by reading reviews Billboard charts NYT Best Seller List Rotten Tomatoes Freshness Score Laugh track on the Big Bang Effect Brains are built to react to social signals. The study of how people behave. Everyone knows how to use behavioral data. Welcome to SCIENCE!
• #5: The Lab Coat Unfair advantage
• #6: I’ve built a passion and a business around behavioral science. Corporate marketer. Created the advertising and web pages to generate leads. Frustrating job. Lots of ideas: mash them or just pick one. (The carousel) Launch and See. Most expensive I created OSOM. It helped. Made me more creative.
• #7: In 2005 the cost of collecting behavioral data went from 25% of a marketers time to nearly zero. And the data was 1000 times better. I loved it. Google in 2005. Offermatica in 2003. It is now cheaper to collect and use behavioral data than it is not to. I’m going to demonstrate that to you today, and I hope you will come to love it as much as I do.
• #8: Conversion Sciences Empathetic Personas Documented the struggles Spin story of journey Awakened their right brains Connected meetings Nothing happened. Dried eyes and fought next fire The left brain won the day. Didn’t know how to predict future. AB testing gave us a way to test brilliant ideas Composed personas and journeys Then tested Most didn’t help Rigorous creativity Exposed the limits of our left brain I am going to show you how to be rigorously creative Restore imbalance
• #9: Our brains have two very different personalities. Two sides of the brain. The right side home to: painter, musician, dancer, writer, child in us. Unrestrained by time or place. Worlds from thin air while chasing butterflies and writing symphonies. Frustrating to left side. Revels in process. Values time and order. Delights in expression. Where speech is created and processed. Catalogs past. Projects future. Worries about businesses and careers. Balances check books, then starts worrying about next month’s balance. TOGETHER Crayons and calendars Elephants and emails Moonbeams and money Farts and waving Left brain Picks through random associations of right brain and making sense of them. The system Butterflies and symphonies: the left brain excels at throwing things away. Right brain associations get tossed. This is why we struggle. Imbalance: Left brain runs business. Behavioral science will restore the balance. Isn’t “behavioral science” something left-brain could love. Think of it as raw meat for the left brain. It is a way to tell the future Left brain’s biggest problem. A way for the left brain to deal with the right brain. Butterflies and symphonies? I can test that. FUNDAMENTAL SHIFT
• #10: Marketing online N=large number of potential buyers Need to make little n bigger. Launch and see collects a large sample size. Don’t need that. When developing: Creative director n=1 Marketing team: n=5-10 Focus group or survey: n=dozens Small teams are not accurate Analytics: n=thousands Intuitively, we know this is better
• #11: Like our star ratings. Focus group, survey, taste test is one point in time. History effect. Things change. Ask a swimmer in January vs. July about beachwear. Digital data can be collected year round from our websites. Just a few lines of code. Becomes a source of ready information.
• #12: Market research takes time. Marketing studies are months old before they are applied. As cost drops, we can do more often. More precision.
• #13: Focus group or survey subjects will lie. Humans are good at rationalizing Don’t know the real subconscious reasons they do things Behavioral data observes people as they act. Asking is not necessary. We can watch
• #14: Yelp has come under fire in recent years due to fake reviews. Business hiring people Advertisers want to use our knowledge of behavioral science Facebook page with tens of thousands of likes? Grain of salt Surveys, focus groups, taste tests, get people who are like buyers Usually difficult to get searchers Search advertising better than display Behavioral data at its best is generated as our prospects and customers interact. They wouldn’t be there otherwise. Can trust our population will behave similarly
• #15: Star rating and the number of reviews = quantitative Reading reviews = qualitative Both can be helpful, but one predicts the future. Qualitative = motivations and emotions Not sure if represent broader market Quantitative = more statistical confidence * Not seasoned with human input BOTH ARE IMPORTANT
• #16: Huge apparel and food chain in UK and Europe. 10b pounds 2013: 13% of revenue from website $1.5 billion with B In 2014, launched new website $180 million 2 years The UX community felt it was probably a positive redesign. Customers didn’t agree 8% drop in sales $10 million per month Loss of brand value Best brand=visitor success We assume they did lots of research What went wrong? The Battle: Right brain vs. left brain Best practices vs. behavioral data Intuition vs. research Intuition is our ability to apply our experience to new situations. If you struggle, there’s an imbalance.
• #17: When data is expensive, intuition rules Campaign development was dominated by intuition. Data is collected at the beginning. This is primarily qualitative data, small sample sizes, strangers. Designers and developers make thousands of decisions. Behavioral data is collected at the end. Called results. Good data, campaign continues. Bad data, we start over. Classic Launch and see
• #18: When data is cheap, things change. Qualitative data is collected at the beginning. But we can afford to test components. Inexpensive behavioral data Intuition has its place, but now we have balance.
• #19: Put this to the test. Landing pages are awesome. Increase conversions Get more from ad spend Easy to collect behavioral data
• #20: Two jobs 1. Keep promise made in ad, email, post or link. If we don’t know the promise, we’re not building a landing page. Is power of landing page. 2. Get visitor to take action If we’re doing anything other than making visitor confident and comfortable, not a landing page. Keep promise, drive action Not part of navigation Not SEO’d
• #21: Create a landing page in the lab: Pure solution of web page Offer Form
• #23: Don’t start with corporate template. Page already has too many other jobs.
• #25: I’m going to talk about studies. Collect and evaluate data Small in scope. Where could we find data about offers? Where we make offers Adwords and other ad networks. Email subject lines and offers
• #26: One of our clients, looked at AdWords ads. Best performers were 20% off and $100 off. Landing page didn’t keep the promise. These are transactional buyers, afraid to spend one dollar too much. Shopping is a full-contact sport.
• #28: Why valuable? Behavioral, not self-reported Prospects and customers, not strangers Big sample size Collected over time. Quantitative Recent
• #37: The best copywriters come to you with two things: A rigorous uncover process Behavioral data from past work They are great at boiling down qualitative research into powerful messages. Market research Studies Surveys Reviews Past campaigns Competitors If your research is thin, they may ask to do a study.
• #38: Surely we can come up with some more creative headlines. Each keep the promise, but will delivery wildly different results. Let’s do a study
• #39: The website UsabilityHub.com was designed for us, and we’ll be returning to them a couple of times. Flashes page for 5 seconds. Asks visitor to answer a few questions. Great with headlines and images. This test works great on mockups because it doesn’t require a live page.
• #46: The website UsabilityHub.com was designed for us, and we’ll be returning to them a couple of times. Flashes page for 5 seconds. Asks visitor to answer a few questions. Great with headlines and images. This test works great on mockups because it doesn’t require a live page.
• #47: This
• #59: Maybe I need a new copywriter. Our page is complete. Offer, Form, Image, Proof and Trust. Notice we haven’t gotten a designer involved yet. Is there more data we could collect to help the designer create a winning page? Design is a trust builder. We want to convey authority and credibility.
• #60: Maybe I need a new copywriter. Our page is complete. Offer, Form, Image, Proof and Trust. Notice we haven’t gotten a designer involved yet. Is there more data we could collect to help the designer create a winning page? Design is a trust builder. We want to convey authority and credibility.
• #61: Let’s do an eye-tracking study. This is another source of behavioral data-both quantitative and qualitative that has come down in price.