- A/B testing tools use five distinct pricing models: MTU-based, impression-based, visitor-based, flat-rate, and free/open-source
- Understanding the model matters more than the starting price — the cheapest entry point often becomes the most expensive at scale
- Flat-rate pricing (Varify: from €149/mo) is the only model where costs are fully decoupled from traffic and usage
- Annual contracts and hidden multipliers (seat fees, overage charges) make direct price comparison misleading without TCO analysis
A/B testing pricing is deliberately confusing. Tools use different metrics (MTU, impressions, visitors, pageviews), different billing cycles (monthly, annual-only, custom), and different feature gates (basic vs. premium tiers) — making direct comparison nearly impossible without understanding the underlying model. This guide decodes each pricing model so you can calculate real costs, not just compare sticker prices.
Varify.io uses flat-rate pricing — the simplest model to understand and budget for. But we'll cover all models fairly, including where alternatives might be cheaper for specific scenarios.
The 5 A/B testing pricing models
1. MTU-based (Monthly Tracked Users)
Used by VWO. You pay based on how many unique visitors interact with experiments each month. Sounds fair — but "tracked users" is opaque: does a visitor who returns 3 times count as 1 or 3? The definition varies and always favors the vendor. Costs scale linearly with traffic.
2. Impression-based
Used by Optimizely. You pay per experiment impression (each time a visitor sees a variant). This double-charges: a visitor in 3 concurrent experiments generates 3 impressions. The more you test, the more you pay — punishing high testing velocity.
3. Visitor-based
Used by Convert, Crazy Egg. Similar to MTU but counted differently. Typically based on tested visitors (only those in active experiments, not all site visitors). More predictable than MTU but still scales with traffic.
4. Flat-rate
Used by Varify.io. One price regardless of traffic, experiments, or team members. From €149/mo (Growth, 5 active experiments) or from €249/mo (Pro, unlimited). The only model where costs are fully predictable.
5. Free / open-source
Used by GrowthBook. Zero licensing cost, self-hosted. Costs shift from subscription to developer time: setup, maintenance, troubleshooting, updates. "Free" in price, not in total cost of ownership.
Real cost comparison across models
| Traffic level | Varify (flat) | VWO (MTU) | Convert (visitor) | GrowthBook (free) |
|---|---|---|---|---|
| 50K/mo | €149/mo | ~$300/mo | ~$99/mo | $0 + dev time |
| 200K/mo | €149/mo | ~$500/mo | ~$299/mo | $0 + dev time |
| 500K/mo | €149/mo | ~$800/mo | ~$399+/mo | $0 + dev time |
| 1M/mo | €149/mo | ~$1,200+/mo | ~$699+/mo | $0 + dev time |
Source: Claude Research, May 2026. VWO/Convert estimates based on published tiers and industry benchmarks.
At 50K visitors, Convert is cheapest commercially. At 200K+, Varify overtakes every paid alternative. The crossover happens quickly for growing companies.
Hidden cost multipliers to watch for
The starting price is never the full picture. Watch for these hidden multipliers:
- Annual-only billing: Some tools show a "monthly" price that requires annual commitment. Varify offers genuine monthly billing — cancel anytime — alongside discounted annual plans.
- Per-seat fees: Adding team members costs $30-100/month each at some platforms. Varify includes unlimited seats on all plans.
- Feature gating: Basic tiers often lack multivariate testing, advanced targeting, or API access. Varify includes all features on all plans (Growth vs. Pro differs only in analytics integrations and experiment limits).
- Overage charges: Exceeding your traffic tier triggers automatic billing at a premium rate. Varify has no overages because there's no traffic tier to exceed.
- Setup and onboarding fees: Enterprise tools often charge $5K-20K for onboarding. Varify: zero setup fees, self-service onboarding with expert support included.
One price. Everything included. No surprises.
From €149/mo flat-rate. See all plans at varify.io/en/plans.
Choosing the right pricing model for your team
Match the pricing model to your situation:
- Growing traffic, budget-conscious: Flat-rate (Varify). Costs stay constant while your traffic and testing program grow.
- Very low traffic, testing the waters: Visitor-based (Convert at $99/mo) or free (GrowthBook). Cheaper at the start, but plan for the crossover point.
- Enterprise with procurement budget: Custom pricing (Optimizely, Kameleoon). Budget exists; the question is value per dollar, not total cost.
- Developer-heavy team, zero budget: Free (GrowthBook). Trade subscription cost for developer time.
For most teams between 50K and 2M visitors, flat-rate pricing delivers the best combination of predictability, value, and growth-friendliness. Start with a free 30-day Varify trial to see it in practice.