VWO

VWO logo
Runs A/B and multivariate tests on websites to track visitor behaviour and optimise conversion rates.

VWO API

Sign in
API key. API key; API access on: EnterpriseSource

MCP server

What it is
Endpoint is per region (global https://mcp.wingify.ai/mcp, EU mcp-eu.wingify.ai, Asia mcp-as.wingify.ai); developers.vwo.com redirects to developers.wingify.com; auth is a token in a request header, OAuth 2.0 applies to the MCP only

Checked on 2026-10-10 in the developer documentation.

How to use VWO

Overview of the VWO platform

VWO Testing

Client-side A / B, split-URL and multivariate experiments all run here.

VWO Insights

Heatmaps, clickmaps, recordings, funnels, form analytics and on-site surveys supply behavioural research from the same interface.

VWO Personalise

Rules-based content targeting and audience-level personalisation campaigns.

VWO Full Stack

Server-side testing, feature flags and mobile SDKs for product-level experimentation.

VWO Deploy

Roll out winning variations to every visitor instantly without waiting for a development release.

VWO Plan

Idea backlog, prioritisation and workflow management to organise your experimentation programme.

Setting up VWO on your site

Install the script

Copy the JavaScript snippet into the <head> tag or publish it through your tag manager. Verify installation with VWO’s checker so data flows before proceeding.

Define goals and funnels

Create global goals for key conversions such as purchases or lead submissions. Build funnel steps to track drop-off points and reuse these metrics across all experiments.

Validate the setup

Use preview mode to confirm the snippet loads without layout shifts. Run a short dummy test and check that data appears in reports before launching real experiments.

Creating and running experiments

Read the full guide

Variation design with the visual editor

Open the target page, adjust text, images or layout and save each version. For advanced changes jump into the code editor to inject custom HTML, CSS or JavaScript.

Audience targeting and traffic split

Choose who sees the test by device, location, referral or custom attribute. Allocate traffic evenly or bias towards control for low-risk roll-outs.

Goal configuration

Attach primary and secondary goals that match the hypothesis, for example form completion, clicks on CTA or time on page.

Launch the test

Review all settings, schedule the start if required and activate. VWO immediately divides visitors and begins recording results.

Monitoring and analysing results

Real-time reporting

Track conversion rates, uplift percentages and probability-to-beat-control as data accumulates. Use the duration estimator to ensure a statistically sound run length before drawing conclusions.

Behaviour insights

Switch to heatmaps and recordings filtered by variation to observe interaction differences. Collect qualitative evidence that explains the quantitative outcome.

Segmentation

Break down results by device, geography or visitor type. Confirm the winning variant performs consistently across key segments.

Deploying winners and next steps

Roll-out with development or VWO Deploy

Hand winning changes to developers for permanent implementation, or use VWO Deploy to serve the variation to one hundred per cent of visitors immediately.

Iterate and personalise

If results are inconclusive, refine the hypothesis and test again. Leverage VWO Personalise to deliver segment-specific experiences based on insights gained.

Personalisation and advanced features

VWO Personalise

Create customised content or offers for defined audiences without splitting traffic.

VWO Full Stack

Run experiments inside application logic or mobile apps and toggle features with flags for gradual release.

Surveys and polls

Collect visitor feedback at critical points and connect responses to session recordings for richer context.

Best practices for success with VWO

Formulate a clear hypothesis for every test and record it in VWO Plan.

Run experiments for a complete business cycle to account for variability.

Segment results before declaring a winner to catch hidden swings in performance.

Archive inactive tests so the snippet remains lightweight.

Use insights to drive continuous optimisation: test, learn, implement and repeat.

Following these steps will help you harness VWO as an end-to-end optimisation system, uniting data, testing and deployment in a single cohesive workflow.

How to automate VWO

MCP server

Wingify runs an official MCP server with OAuth 2.0 sign-in, so you do not manage API tokens by hand. Choose the server URL for your account region. It connects AI tools such as ChatGPT and Claude to your live campaign data.

An idea: ask your assistant which campaigns from the last three months have no winner, then decide which to stop or extend.

API and webhooks

There is a REST API for managing accounts, campaigns and variations, documented at developers.wingify.com. A separate MCP repository exists for feature experimentation.

An idea: when a test reaches significance, create a task in your project tool with the winning variant and the next step, so results do not sit in a dashboard.

No-code automation

The vendor's help article says the MCP server can connect to automation platforms such as n8n or Zapier, and gives automated weekly reports sent to Slack as an example. The same route works with n8n.

An idea: a weekly summary of running experiments, posted in Slack on Monday morning.