Improve CRO, ICPs, and Streamline A/B Testing
Solving the of Conversion, Retention, and Growth Problems
Companies tend to treat low conversion rates as a performance issue to be tested away. We hope that a few more rounds of guesses will bring success.
While we cycle our Product, Design, and Engineering teams through cycles of slight changes and varied concepts, we’re losing.
- Visitors abandon the funnel or purchase.
- They bounce off our landing pages.
- They leave the checkout before completion.
Companies try to solve these issues by throwing marketing, SEO, PPC, and influencer money at the problem. The typical results: more traffic to a site that still doesn’t convert, more trial users for products and services, but they don’t want to stay or pay.
We find out why.
It’s hard to solve problems when you don’t fully understand the problems or your Experience Ecosystem. Delta CX uses qualitative research, human-to human conversations and observations, to learn your market’s tasks, behaviors, decisions, collaborators, preferences, and realities.
We can answer the hows and whys, and deliver actionable suggestions on what to do (and not do). For an example, see our case study about a hosting company struggling with funnel abandonment and low conversion rates.
ICPs: Shift from Demographics to Behaviors
Do all 50-year-old women behave the same? Need the same things? Have the same levels of savvy, knowledge, or experience? Of course not. Yet companies still segment and group people by demographics as if it’s 1950s America, conformity is key, and you can count on every housewife having the same amount of money, 2.7 children, and identical shopping behaviors.
“Personas” got muddy, complicated, and non-actionable, so we tried renaming them… customer profiles, segments, and target audiences… but that was a poor solution for a different problem. Call these groups anything you’d like, and redefine their parameters.
Reality is diverse.
You can’t say that you understand our market or audiences if you don’t understand them. You can’t have empathy for people you avoid understanding.
Assumptions aren’t empathy. Role-playing or imagining what people might do isn’t empathy. False empathy can push us down the wrong paths. Teams that “really knew their customers” are surprised later when utilization is low, nobody cares about the new feature, or everybody who said they’d pay for it… won’t.
Delta CX uses properly planned and recruited qualitative research to help companies identify or refine their ideal customer profiles, segments, and personas. Shift away from assumptions and demographics.
Streamlining
A/B testing is normally done after products or services are live since you want to compare how A, our existing offering, performs against B, a changed or new version. We can remove false confidence by shepherding B through Delta CX’s Atomic Product-Market Fit process (or any good variation of The Scientific Method), before we build, release, or A/B test. But for many teams, B is their next guess dressed up as their new “hypothesis.”
Teams love A/B testing because it feels scientific, and the data makes the decision. But A/B testing won’t be the right tool for every job. It’s for optimization, not design, problem solving, or innovation. You can’t optimize your way out of strategic problems.
A/B testing has some downfalls, including:
- A/B testing requires you to create products or services that are available to the public. You won’t learn if something works or not until very late in the product or service development process. By then, you have invested time, money, and resources in live code, physical products, or personal services.
- A/B testing doesn’t tell us why something is better or worse. Something we’re measuring happened more or less often. Why? We claim that we A/B test to learn, but what did we learn? We don’t know more about what users need, so we don’t know what a significantly better design would look like. It’s FailureSquared™: failure to learn from failure.
- A/B testing can’t tell you what to design or build. We learned that something happened more or less frequently, but what should we change? What should remain unchanged because it’s working? A/B testing isn’t a design or problem-solving technique.
Strategic and streamlined A/B testing.
The best way to have more successful A/B tests is to have higher quality standards for B. Your customers have higher standards and expectations. Don’t disappoint them.
Evaluative research techniques can tell us long before an A/B test if B is likely to succeed. We learn all of this while B is still a concept or prototype. Failures are behind the scenes rather than being public releases competitors, investors, and the public sees… while paying customers are treated like laboratory rats.
Evaluative research tells you weeks or months before an A/B test whether B is worth building at all. That’s time, money, and dev cycles back in your pocket.
It’s also a more efficient path for innovation. You can’t A/B test your way to innovation. If you want fresh, new ideas that become competitive advantages, helping you leap over the competition, cycles of small experiments are unlikely to get us there.