How tha fuck nuff metrics do you report ta yo’ crew on a weekly basis? Do you know which one(s) you’re actively tryin ta improve, biatch? What’s yo’ process fo’ tacklin dat challenge?
Enactin chizzle wit data is easier holla’d than done, so we’ve collected all dem thoughts on how tha fuck ta be data driven as we crafted our latest in-depth guide ta optimization. Our thugged-out asses have access ta a tremendous amount of data, with dozenz of metrics n’ KPIs at our disposal. It aint nuthin but tha nick nack patty wack, I still gots tha bigger sack. Despite our dopest efforts, collectin data n’ reportin on metrics can still leave our asses guessing.
“Data-collectin don’t mean data-informed, n’ data-informed don’t mean data-acting.59% of g-units do not have processes up in place up in they crews ta ensure dat data is understood n’ acted upon.“ ��”Data Driven Culture Survey Report, Econsultancy & Geckoboard
We’ve discovered dat data isn’t just hard as fuck ta act on; it’s tough ta know what tha fuck data is tha right kind, n’ how tha fuck ta incorporate data tha fuck into a cold-ass lil continuous process of testing. Without a gangbangin’ framework fo’ regularly collectin n’ takin action on our data, we’ll never reach our goal of becomin mo’ data-driven.
Just rockin data isn’t enough
Da ideal of bein ‘Data-Driven’ is straight-up ta be ‘Data-Informed.’ Data points is a blingin set of inputs yo, but they aint a replacement fo’ human intuizzle n’ judgment. Quantitatizzle analysis can help ta uncover areas where there be gaps up in yo’ joint’s performance, or opportunitizzles fo’ improvements all up in optimization. What data points cannot provide, however, is input on what tha fuck yo’ bidnizz metrics should be, or what tha fuck initiatives you wanna prioritize. Quantitatizzle data can also fail ta provide context. Afta yo’ analytics data serves up a gangbangin’ framework fo’ where ta test, qualitatizzle insights fill up in tha gaps of what should be tested��”and how. Hiten Shah, Co-Founder of KISSmetrics, recently shared some key questions ta ask when formulatin A/B tests.
Use a structured process fo’ buildin data tha fuck into optimization
Optimizin yo’ joint, app, or other hustla touchpoints are prime opportunitizzles to proactively experiment n’ take action wit yo’ data. Additionally, experiments based on data-informed hypotheses increase yo’ likelihood of success compared ta a randomly structured testin program. Data skits a key role up in tha followin stepz of the Optimization Flywheel:
Da first step ta a successful optimization process is ta clearly align tests round a funky-ass bidnizz goal. It aint nuthin but tha nick nack patty wack, I still gots tha bigger sack fo’ realz. A clear understandin of tha metrics yo ass is optimizin fo’ will help wit prioritization n’ enablez continual iteration n’ peepin’ from experiments.
Determine Optimization Points
Identify a step up in yo’ funnel dat be a prime muthafucka fo’ optimization. I aint talkin’ bout chicken n’ gravy biatch. Well shiiiit, it might be yo’ homepizzy call ta action (CTA), yo’ campaign landin pages, yo’ checkout flow, or yo’ recommended content. Make shizzle ta chizzle a area fo’ optimization dat has a gangbangin’ finger-lickin’ direct correlation ta yo’ bidnizz goals. Quick funnel analysis wit yo’ analytics software will provide context fo’ which areaz of yo’ joint can be improved. Y’all KNOW dat shit, muthafucka! Will optimizin dis step of yo’ joint experience create a measurable chizzle ta tha metric you identified when settin yo’ goals?
A phat hypothesis bout how tha fuck yo’ experiment will big-ass up is core ta hustlin a ballin A/B test yo. Hypotheses is statements, not open-ended thangs. They address a question wit a proposed solution. I aint talkin’ bout chicken n’ gravy biatch. Craftin a hypothesis ta address a open question or problem on yo’ joint enforces a well-rationalized, thoughtful proposal fo’ how tha fuck ta address dat problem. To take yo’ hypothesis even further, consider what tha fuck you would learn if yo’ prediction was proven erect or incorrect up in a experiment. What would you learn up in each scenario, biatch? Leverage yo’ knowledge of yo’ hustlas’ needz n’ other qualitatizzle data for clearer indicatorz of what tha fuck should be tested ta betta match yo’ visitors’ intent n’ solve they frustrations wit yo’ thang or joint. Take tha time ta collaborate on hypotheses wit yo’ crew, n’ properly document dem along wit tha data dat you used ta inform dem (both qualitatizzle n’ quantitative).
Measure n’ Iterate
Afta hustlin a experiment, evaluate whether yo’ hypothesis was proven erect fo’ realz. A discovery from a ballin test can be applied to other areaz of a website, or testin a more advanced hypothesis which buildz upon tha previous discovery (segmentation of high-value crews within tha previous test, fo’ instance.) If yo’ test was a thugged-out draw or yo’ variation lost, rewind why dat might be. Well shiiiit, it could be tha case dat yo’ hypothesis needed additionizzle research, or dat you didn’t account fo’ a funky-ass behavior or event dat skewed yo’ test thangs up in dis biatch. In tha event of a losin test or a thugged-out draw, it is still possible ta bust additionizzle insights dat you had not anticipated all up in further analysis n’ rap wit yo’ testin crew fo’ realz. At dis point, say shit bout what tha fuck tha data couldn’t account fo’ as you planned tha experiment. What would you do differently up in yo’ next hypothesis n’ experiment?
We’re just gettin started
Uncover a intentional, measurable approach ta hustlin betta A/B tests n’ experiments dat will help you hook tha fuck up wit hustlas, boost yo’ KPIs, n’ grow yo’ bidnizz wit tha guide ta Buildin yo’ Company’s Data DNA. Downlizzle tha complete resource for:
- Organizationizzle recommendations fo’ alignment round key metrics
- Tactical lyrics fo’ CRO practitioners round uncoverin n’ prioritizin impactful test ideas
- Tips from data n’ optimization smart-ass muthafuckas at KISSmetrics n’ Qualaroo