A/B Testing and Conversion Optimization
Every team has opinions about what will convert. I design and run A/B tests that settle the question with evidence, then turn the results into decisions your team can act on without guessing.
Conversion optimization is the discipline of finding out why visitors stop short of the action you want and testing changes that remove the obstacle. A/B testing is how those changes earn their place. I write the hypotheses, define the metric each test must move and specify the variants. I confirm that tracking fires correctly, then read the results at the sample size and duration they need. The output is a decision, not a chart.
I run A/B tests as part of persona-based web strategies in my consulting practice, and I have coordinated them inside an Agile campaign team. That experience taught me to start from the conversion funnel and the analytics behind it, not from a list of page elements to change. Tests are prioritized by the size of the problem they address and the evidence behind the hypothesis, so the program spends its traffic where it can learn the most.
Each test arrives as a brief your designers and developers can build from: the hypothesis, the variant specification, the targeting rules, the primary and guardrail metrics and the QA checklist. I run acceptance testing before anything goes live. Once a test concludes I record the result in a decision log, so the program builds shared knowledge rather than a pile of disconnected experiments.
What a testing program includes
Conversion Audit
A walkthrough of the key conversion paths with analytics data on where visitors drop off, friction points by page and device, and a short list of candidate problems worth testing.
Hypothesis Backlog
A prioritized list of test ideas, each written as a hypothesis with the evidence behind it, the metric it should move and a score for expected impact and build effort.
Test Plans
A plan for each experiment covering the primary metric, guardrail metrics, audience and targeting, traffic split, minimum sample size and the duration needed before a result can be trusted.
Variant Briefs
Specifications for each variant with copy, layout changes and annotated mockups, written so design and development can build the challenger without a second round of questions.
QA and Launch Checklist
Pre-launch acceptance testing of each variant across browsers and devices, confirmation that goals and events fire in analytics, and a sign-off record before traffic is split.
Results Readouts
A written readout for each test stating whether the result is conclusive, what the data supports, what to ship and what to test next, logged in a running decision record.
How I run a testing program
Map the funnel
I define the conversion actions that matter, trace the paths visitors take to reach them and confirm the analytics can measure each step before any test is proposed.
Build the backlog
Drop-off data, session behavior and stakeholder input become hypotheses. I score each on evidence, expected impact and effort, then agree on the first round of tests with your team.
Brief, build and QA
Each test gets a plan and a variant brief. Your team builds it, I run acceptance testing and verify tracking, and nothing goes live until the checklist is signed off.
Read and decide
When a test reaches its planned sample and duration I call the result, write the readout, record the decision and feed what the test showed into the next round.
Where this comes from
A/B testing, analytics and KPIs sit alongside landing page and conversion optimization at the core of my practice. In my consulting work I set KPIs and run A/B tests to lift engagement, discoverability and lead generation. At Bio-Techne I coordinated landing pages, tracking and A/B tests in an Agile team. At UnitedHealth Group I defined KPIs and reported performance through Google and Adobe Analytics. Requirement specifications and user acceptance testing were part of how website work shipped in that role.
See my experienceBrittany was an outstanding employee and will be missed in our organization. She has a great work ethic, is passionate and knowledgeable about Internet Marketing, and applies thoughtful and perceptive analysis regarding her work. She has a wonderful attitude about work and was always cheerful and upbeat.
Founder and Senior Business Leader, Pyron Technologies
Managed Brittany directly
Frequently asked
How much traffic do we need to run A/B tests?
Enough to reach a reliable sample within a few weeks, which depends on your current conversion rate and the size of change you need to detect. I calculate the required sample for each test before it launches. If traffic is too low for a split test, I will say so and propose sequential tests or qualitative methods instead.
Which testing tool do you use?
I work with the testing capability already in your stack and the analytics you already report from, including Google Analytics and Adobe Analytics. The method matters more than the tool. Hypothesis quality, correct tracking and disciplined reading of results decide whether a program works, and those carry across platforms.
What happens when a test is inconclusive?
An inconclusive result is still a decision: keep the control, record what was learned and move on. Many tests end this way, and a program that only reports winners is hiding something. I write up why the variants did not separate, whether the hypothesis deserves a different treatment and what the next test should be.
Is this the same as landing page optimization?
No, though the two overlap. Landing page optimization builds and refines individual campaign pages. A/B testing and conversion optimization is the broader program of forming hypotheses across the site, testing them against a control and deciding what to ship from the evidence. The two often go together, since landing pages are a natural place to start testing.
Related services
Landing Page Optimization
A landing page has a single job. I plan, build and improve campaign pages so the visitor who arrives from an ad, an email or a search result finds the offer they expected and a clear next step.
Learn moreAnalytics, KPIs and Dashboards
Most marketing teams have more data than they can use and fewer answers than they need. I define the KPIs tied to your business goals, make sure the tracking behind them is sound, and build dashboards that people open because they help with a decision.
Learn morePersonas and Customer Journeys
Personas and journey maps are only useful when they change what gets built and published. I research who your customers are and how they decide, then document it in a form that content, design and campaign teams will use.
Learn moreLet's find out what converts
Tell me where visitors drop off and what you have tried so far. I will tell you whether a testing program makes sense for your traffic and what the first round of tests would look like.