How I approach conversion optimization: find where revenue leaks, then close it.

Conversion optimization (CRO) is the disciplined work of turning more of a business’s existing traffic into customers, without paying for more clicks. It starts by diagnosing where and why visitors stop (message mismatch, decision uncertainty, missing trust, friction, weak relevance, or a measurement gap) and then tests the fix instead of assuming it worked.

I led enterprise conversion optimization programs for more than 10 years, and before that ran my own CRO consultancy for ecommerce sites and landing-page programs. Across thousands of conversion audits and website reviews, the method has stayed the same: bring the analytics, testing and behavior data together, find where intent breaks down, and fix it with evidence.

Most traffic leaks out of the funnel before it converts – CRO seals the leaks A conversion funnel. Visitors enter at the top; at each stage – landing, product, cart, checkout – a share leaks away, shown as drips escaping the sides. Only a small fraction converts at the bottom. CRO finds and seals the leaks so more of the traffic you already paid for converts. CONVERSION RATE OPTIMIZATION More customers from the traffic a site already earns. Visitors – 100% Engaged Cart / intent Convert Confusing landing → bounce Friction in the journey Checkout drop-off Same traffic already paid for The button is the last 5%. The real work is finding where revenue leaks – and sealing it with evidence.

What I look for first

Five signs the problem is conversion, not traffic.

  1. Traffic is stable or growing, but the conversion rate hasn’t moved in the same direction.
  2. Visitors reach the page that matters, but analytics show them leaving at a specific, repeatable step.
  3. Sales or support fields the same “which one is right for me” question from prospects who already found the site.
  4. A paid channel’s cost per click looks fine, but its cost per customer keeps climbing.
  5. The team can point to a page or step where people hesitate, but no one has tested a fix, only discussed one.

When three or more of these show up, I treat the constraint as what happens after the click, not how many people click.

CRO is not right when qualified traffic is still extremely limited. It improves the percentage of visitors who convert; it cannot manufacture visitors who were never there. If the constraint is demand, not conversion, the right next investment is acquisition or SEO, not CRO.

Case studies

Where this was the actual fix.

Case studies where conversion optimization was the main lever: the problem first, then what changed and what it showed.

Case study

Entrepreneurs, Authors, and Professional Speakers

Trust & Credibility + Lead Generation

Conversion OptimizationPersonalization
Read the case study

Working across a range of individual sellers, authors promoting a book launch, speakers filling a seminar, coaches enrolling a course, the constraint was the same each time: none of them had the built-in recognition of an established publisher or media brand. A landing page or email sequence that leaned only on the seller's own claims about themselves read as self-serving to a skeptical visitor. The direct-response systems built for this segment were engineered around that constraint from the start: personal-brand sites, registration and opt-in flows, and email sequences that made room for testimonials, endorsements, media mentions, and other third-party proof points to carry the trust argument the seller's own copy couldn't carry alone. The mechanism is the underdog play in miniature: when you can't win on name recognition, you win on borrowed credibility, structured into the funnel rather than bolted on as an afterthought.

Case study

Alternative Short-Term Lending Experience

Complex Decision-Making + Trust & Credibility

Conversion OptimizationPersonalization
Read the case study

Payday-loan alternatives face a specific trust problem: the category itself is distrusted, and a paid-search visitor arriving from a competitor's ad has no reason to take "safer and more affordable" on faith. Telling prospects the terms were better wasn't going to close that gap. Showing them was. The interactive comparison tool let a prospect enter the details of their existing loan or payment obligation and see it set directly against the company's proposed terms, geolocated so the comparison reflected the rates and terms actually available to that visitor. That turns a decision that normally requires trusting an unfamiliar lender's claims into a decision the prospect can verify with their own numbers, the same guided-decision mechanism that works for mortgage lending and jewelry purchases, applied to a category where distrust is the default starting point.

Case study

Children's Specialty Sock Ecommerce

Trust & Credibility

Ecommerce OptimizationConversion Optimization
Read the case study

For a specialty product like distinctive, playful children's socks, the real competition isn't another sock brand. It's the generic multipack sitting one click away at a lower price. Winning that comparison isn't a pricing argument; it's a trust argument about why the specialty version is worth choosing at all. The work centered on making that case visually and emotionally rather than through claims: lifestyle photography showing real children wearing the socks, framed for the two buyer types who actually shop this category: parents outfitting their own kids and gift buyers shopping for someone else's. The shopping experience itself was built to carry that emotional register through from browse to checkout, so the specialty product read as an obviously better choice rather than an upsell the shopper had to be talked into.

Case study

Personalized Health and Vitamin Planning SaaS

Product Discovery + Education

PersonalizationConversion Optimization
Read the case study

A vitamin and nutrition catalog presented flat is a Product Discovery problem before it's anything else. There's no way for a visitor to know which of dozens of possible recommendations actually applies to them without some form of guidance, and no way to trust a generic recommendation once they get one. The quiz-based journey solved both halves at once: starting from only basic information (age, gender), it walked the visitor through a guided path that narrowed toward a personalized set of vitamin and nutrition insights, rather than asking them to self-navigate a full catalog. That personalized result did double duty as the education layer too, a tailored insight is inherently more explainable than a generic one, which is what made the next step (connecting to a nutrition professional, or enrolling in the newsletter) a natural continuation of the journey rather than a separate ask.

The problem

Traffic is fine. Too few of those visitors take the next step.

Most of the budget goes to getting people to the site: ads, SEO, email, events. Far less attention goes to what happens once they arrive. So the pattern repeats: visitors land, hesitate, and leave. The business pays full price for traffic and collects a fraction of what it is worth.

The leaks are rarely where opinion says they are. A confusing landing page sends people straight back. Friction in the journey wears them down. A checkout asks for too much, too soon, and the closer that step is to revenue, the less room there is for ambiguity. Each one quietly bleeds revenue, and without measurement nobody can say which is costing the most.

What it actually is

Improving the percentage of visitors who do what the business wants.

Conversion optimization (CRO) is the systematic practice of getting more existing visitors to take the action that matters (buy, book, sign up, call) without buying more traffic.

It is not moving buttons around. The button is the last 5%. The real work is understanding why visitors hesitate, what they need to feel confident, and removing the friction in their way - informed by data, validated by tests. You change one thing, show both versions to real visitors, and let the result decide what wins.

How I do it

It runs through the loop - Discover, Prioritize, Experiment, Learn.

Conversion optimization is not a one-time fix. It is the core loop of the Experience Optimization Framework. I Discover where intent breaks down, Prioritize the leaks by likely impact and effort, Experiment to prove a fix on the live audience, then Learn from every result (win, loss, or flat) and feed it into the next round. Each cycle builds on the last, so performance compounds instead of plateauing.

This is a discipline I've practiced for 26 years, from building and selling ecommerce sites, then moving upstream into optimization itself, learning alongside conversion pioneers like Bryan Eisenberg. Across 100+ ecommerce sites and 7,000+ experiments across my career, the lesson holds: opinion is a hypothesis, and only evidence gets to decide. a test is only valuable when it answers a real business question.

The button is the last 5% of CRO; the real work is the 95% below the surface An iceberg. Above the water, button and surface tweaks - about 5 percent. Below the water, the real mass: understanding why visitors hesitate, building confidence, removing friction, and proving changes with real visitors - about 95 percent. CRO ISN'T MOVING BUTTONS The button is the last 5% WATERLINE Buy now Button & surface tweaks color, wording, placement - the visible 5% THE REAL 95% Why visitors hesitate What builds confidence Removing friction Proving it with real visitors Moving the button is the finishing touch - not the work.

See how the system runs →

See the diagnosis method in more depth →

What I have worked on

Every place a visitor decides.

From the first landing page to the final step of checkout: diagnosed, prioritized, and improved as a continuous program, with every change reviewed before it goes live.

Work I have done
Conversion Audits Roadmaps Landing Pages Checkout Funnels Behavior Analysis AI Opportunity Discovery

The technology

The platforms I work in.

Conversion optimization runs on the analytics, testing, and behavior tools a company already owns. My job is to wire them into one system, extend them with AI where it helps, and make what the company pays for work harder.

Quantum MetricHotjarContentsquareOptimizely Web ExperimentationAdobe Target

What people said

In the words of people I worked with

For over six months we'd struggled to increase our conversions and cost per lead. One call with Alex... we incorporated A/B testing and, with his guidance, our conversions doubled and our cost per lead was cut in half.
Michael C.
Alex is awesome. He has an amazing perspective and the ability to focus on the elements that have the biggest return on investment. His expert review of my site and subsequent advice has been so beneficial. Alex Designs just continues to add value!
Roger G.

What's actually different

CRO vs. Landing Page Optimization. CRO improves conversion across an entire site or a whole customer journey: any page, any traffic source, any step where a visitor decides whether to continue. Landing page optimization improves one specific page tied to one specific traffic source - the page an ad, email, or search result sends someone to - against the exact promise that source made. A CRO program might touch a dozen pages across months. Landing page work is scoped to the page-to-promise gap for a single campaign or channel. If nobody knows which ad or link sent the visitor, it is a CRO question. If the source is known and the page doesn't match what that source promised, it is a landing page question.

Experimentation vs. CRO. CRO is the diagnostic and strategic work of deciding what to fix and why. Experimentation is the validation method that confirms whether a specific fix actually worked, run as an A/B test, a multivariate test, or a smaller validation method sized to available traffic. Every CRO program uses experimentation to prove its hypotheses; not every experiment is CRO: a pricing team can run an experiment that has nothing to do with conversion diagnosis. If the question is "what's broken and why," that's CRO. If the question is "did the fix we already chose actually work," that's experimentation.

Related reading

Go deeper on conversion.

Common questions

Common questions about CRO.

How is CRO different from a site redesign?

A redesign changes everything at once and hopes; CRO changes one thing, shows both versions to real visitors, and lets the result decide. The button is the last 5% - the real work is understanding why visitors hesitate and removing that friction.

How much traffic does a test need?

Enough to reach statistical significance on a test in a reasonable window. If traffic is too low for that, I say so and start with qualitative diagnosis instead.

How long until CRO shows results?

A prioritized roadmap and the first tests can usually ship within the first month; the larger gains build over the following cycles as each result feeds the next.

What tools do you work with?

Whatever the team already owns: Optimizely, Adobe Target, Quantum Metric, Hotjar, Contentsquare and others, connected into one way of working rather than replaced.

See how this played out in real work.

The case studies show the problem, the decisions, the teams involved and what changed.