Websites & CRO
Systematic CRO that compounds, not one-off redesigns.
A single page redesign is a hypothesis. A structured CRO programme is a system for continuously validating and compounding those hypotheses. Most Indian companies treat conversion optimisation as a one-time project. The ones that grow fastest treat it as an ongoing function: test, measure, iterate, repeat. We run that function for you.
Why most conversion optimisation efforts do not compound.
CRO is done wrong more often than it is done right. The failure modes are predictable: companies run single tests, draw conclusions from insufficient data, and make changes that cannot be separated from other site changes happening simultaneously.
Tests are run without statistical rigour.
Running a test for one week on a page with 200 visits and declaring a winner is not CRO, it is opinion with an A/B test wrapper. Statistical significance requires a minimum sample size based on your current conversion rate and the minimum detectable effect. Most companies stop tests too early, declare winners that are false positives, and implement changes that either do nothing or hurt conversion.
There is no prioritisation framework.
Testing the button colour on a page that gets 50 visits a month before testing the hero copy on the homepage that gets 5,000 visits a month is a misallocation of the testing capacity. Without a prioritisation framework that weights by traffic, conversion impact, and effort, the highest-leverage tests are never run.
Conversion improvements are not isolated.
If you are running CRO tests and your paid media team is changing campaign targeting and your content team is publishing new case studies at the same time, you cannot attribute a conversion rate change to any specific variable. CRO requires a controlled testing environment and a change management process that prevents contamination of results.
Buyer behaviour data is not being used.
GA4 tells you what happened. Session recordings and heatmaps tell you why. Scroll maps show you where buyers stop reading. Click maps show you what they are trying to click that is not clickable. Without this qualitative layer, every test hypothesis is a guess about buyer behaviour rather than an inference from observed behaviour.
CRO is treated as a one-time project.
A single round of CRO testing runs its course in 60 to 90 days. The compounding effect of CRO comes from successive rounds: each test informs the next hypothesis, each winner becomes the new baseline to beat. Companies that run CRO as a continuous function see 2x to 3x the total conversion lift of companies that treat it as a one-off project.
How we run CRO as a systematic programme.
We run CRO as a structured quarterly function with a defined research, hypothesis, test, and measure cycle. Every test is justified by data. Every winner is documented. Every iteration builds on the last.
Research and baselining
- GA4 audit: confirm conversion events are tracked correctly and data is reliable
- Page-level conversion rate baseline: establish the starting point for every page in scope
- Heatmap and session recording review: identify specific buyer behaviour patterns
- Scroll depth analysis: identify where buyers are abandoning the page
- Form analytics: field-level completion rates and drop-off points
- User interview synthesis or review mining: understand buyer objections in their own language
- Minimum detectable effect calculation: set the sample size required for statistical significance
Hypothesis backlog and prioritisation
- Hypothesis generation: draft 15 to 20 test ideas from research findings
- PIE scoring: rank each hypothesis by Potential impact, Implementation ease, and Evidence strength
- Test calendar: schedule tests by available traffic, page priority, and inter-test dependencies
- Control and variant definition: document exactly what changes in each test and what stays constant
- Success metric definition: primary metric (conversion rate), secondary metrics (engagement), and guardrail metrics
- Contamination audit: confirm no other site changes will run in parallel with active tests
Testing and measurement
- A/B test build and QA in your testing platform (VWO, Optimizely, or Google Optimize equivalent)
- Traffic split confirmation and monitoring for test validity
- Weekly check-ins: monitor for anomalies, sample ratio mismatch, and early stopping conditions
- Statistical significance check: tests run until significance is reached, not until a deadline
- Winner documentation: record the result, the hypothesis that was validated or invalidated, and the learning
- Loser analysis: understand why a variant underperformed, often more informative than a winner
Iteration and reporting
- Winner implementation and new baseline update
- Hypothesis backlog refresh based on test learnings
- Quarterly conversion rate report: cumulative lift vs. baseline
- Compounding projection: model the impact of continuing the programme for six and twelve months
- Quarterly strategy review: reprioritise pages and hypotheses based on business goals
What is included in a CRO engagement.
Research and analytics
- GA4 audit and conversion event verification
- Heatmap, scroll map, and session recording analysis
- Form analytics and field-level drop-off review
- Baseline conversion rate report
- Buyer behaviour research synthesis
Testing programme
- Hypothesis backlog with PIE scores
- Monthly test calendar
- Test build, QA, and deployment in your platform
- Statistical significance monitoring
- Winner and loser documentation
Reporting and insights
- Weekly test status updates
- Monthly conversion rate dashboard
- Quarterly performance report with cumulative lift
- Compounding projection model
- Recommendations for next quarter priorities
Strategy and management
- Monthly strategy call with your team
- Change management process to prevent test contamination
- Integration with paid media and SEO roadmaps
- Test learnings library: documented insights from every test run
- Platform and tool recommendations based on your traffic volume
This is right for you if:
- You have at least 3,000 monthly visits to the pages in scope: below this, tests cannot reach statistical significance in a reasonable time frame
- You are spending on paid media and want to compound those returns by improving the conversion rate of the destination pages
- You have completed a website redesign or landing page project and want to continuously improve on the baseline
- Your sales team tracks lead quality and can tell you which website-sourced leads are converting to customers: this data is essential for optimising toward revenue, not just form fills
Not the right fit if:
- You have under 2,000 monthly website visits: the traffic volume is too low to run statistically valid tests in a quarterly window and the engagement would not deliver measurable results
- You are unwilling to commit to at least a 90-day engagement: CRO does not produce meaningful results in 30 days, and single-test projects are not a CRO programme
Frequently asked questions.
What testing platform do you use?
We work in VWO and Optimizely for clients with higher traffic volumes, and in Google Tag Manager with custom event-based testing for clients who want to avoid additional platform costs. The right tool depends on your monthly traffic, the complexity of the tests you need to run, and your technical setup.
How many tests can you run per month?
For a site with 5,000 or more monthly visits, we typically run two to three tests per month. Below that, we run one test at a time to allow sufficient traffic per variant. Running more tests than traffic supports is the most common CRO mistake and produces false positives that hurt conversion when implemented.
How do you measure the impact of CRO separately from other traffic changes?
We track conversion rate, not conversion volume. If traffic goes up because of a new campaign, conversion rate is unaffected by that change. We also segment by traffic source so a sudden influx of lower-quality traffic does not artificially suppress the conversion rate we are optimising.
Do you handle the implementation of winning variants, or does our developer?
During the testing phase, we build and deploy variants within the testing platform. When a test reaches significance and a winner is confirmed, the final implementation into the production codebase is done either by your developer using our specification document, or by us if you are on the website management retainer.
Ready to turn your website into a compounding conversion asset?
Book a 30-minute CRO scoping call. We will review your current traffic, identify which pages have the most conversion potential, and outline what a 90-day programme would look like.
Book a call