- On September 16, 2026, VWO and AB Tasty launched as one brand, Wingify, which describes experimentation, personalization, feature management, insights and commerce on one data foundation, with one AI.
- Agencies tell us the same thing in calls: the more a testing platform bundles and automates, the thinner the line between tool and service gets. That is what we hear. It is not a finding about Wingify’s plans.
- When the client pays the license, every euro of license comes out of the same budget as testing and consulting. The tool decides how much is left for the agency.
- Five questions help you check any vendor, Varify.io included: who holds the contract, what moves with traffic, who talks to your clients, what the AI assumes about you, and what leaving costs.
In discovery calls with CRO agencies, we keep hearing one worry in different words: as testing platforms add AI and bundle more into one suite, they move closer to what agencies sell.
We cannot verify that for Wingify. We know of no public statement from Wingify about how it sees agencies, and this article does not claim one. What we can do is describe what changed, show where the budget logic bites, and give you five questions you can put to any vendor before you recommend its tool to a client.
Some background first. If you have not read it yet, our article on the VWO and AB Tasty merger covers the facts and timeline.
What changed with Wingify, in one paragraph
On September 16, 2026, VWO and AB Tasty launched as a single brand, Wingify. According to the press release, experimentation, personalization, behavioral insights, commerce optimization, feature management and customer engagement run on one data foundation and one AI.
For a team with a single brand and one budget, that can be convenient. For an agency that recommends tools to many clients, a bundle changes three things: what the client pays for, which features the client already owns, and which of your tasks the AI layer claims to cover.
Who pays for the license decides who keeps the budget
Many agencies work in a setup where the client pays the tool license directly. Then the CRO budget has to cover two things: the license and the agency’s work. Both come out of the same pot.
An illustration with made-up figures: a client sets aside €6,000 per month for conversion optimization. Neither license is a real price of Varify.io or Wingify.
| License A | License B | |
|---|---|---|
| License per month | €1,800 | €60 |
| Left for testing and consulting | €4,200 | €5,940 |
| Share of the budget that is tool | 30% | 1% |
The agency’s work does not get cheaper when the license gets more expensive. It gets squeezed. That is why license models matter to agencies more than most feature lists do: a predictable license is budget you can still spend on strategy, hypotheses and analysis.
Where tool and service start to overlap
A testing program has recurring steps: find hypotheses, build variations, run the test, read the result, decide. Agencies are paid for the judgment inside those steps.
Whenever a vendor’s AI moves from describing data to recommending what to test and building the variation, it touches that work. That is not bad by itself. Faster variation building helps everyone. The question is who the speed is built for:
- For the expert: the AI drafts, the expert reviews and decides, the client gets more tests per month.
- Instead of the expert: the AI runs the loop, and the human mostly pays the license.
You can see which of the two a vendor means in how it talks about people. Does it describe the AI as a tool for your team, or as a replacement for the team’s time? We wrote about why we think the first model wins in Expert Support vs. Automation in CRO.
Five questions to ask any testing vendor before you recommend it
These work for Wingify, for Varify.io and for everyone else. Ask them in writing, so the answers survive a change of account manager.
| # | Question | Why it matters to an agency |
|---|---|---|
| 1 | Who holds the contract: the client, the agency, or either? | Decides who controls billing, access and the exit. Ideally both models can run side by side, per client. |
| 2 | What moves the price: traffic, seats, domains, modules? | Traffic-driven pricing makes a client’s best campaign month your worst budget month. |
| 3 | Does the vendor sell services, and who contacts your clients? | If the vendor has a services arm, ask how partner leads are handled and get it in the partner terms. |
| 4 | Does the AI assume a human expert, or aim to replace one? | It shapes your role in two years: reviewer and strategist, or license reseller. |
| 5 | What does leaving cost: data export, code portability, notice period? | An agency that cannot move a client in an afternoon is locked in, whatever the contract says. |
Question 5 is the one most teams skip. If you are already weighing a move, our guide on switching from VWO or Wingify to Varify shows how little of a test is tied to the old tool.
How Varify.io answers the same five questions
- Contract. A client can hold a single-domain account and invite you as a user, and you still earn a lifetime commission on the referral. Or you run one agency account and bill each client yourself. Both models can run side by side, differently per client.
- Price. Varify.io charges a flat price per domain, independent of traffic. In the agency model, one credit covers one active domain for one month, and a domain you set inactive consumes nothing. Details on the agencies page and the pricing page.
- Services. Varify.io works with in-house CRO teams and agencies alike, and never competes with its customers’ service business. We used to be an agency ourselves, so we build for the agency workflow. Certified partners are listed in our expert network.
- AI. The AI CRO Audit proposes test ideas, Prompt-Based Experimentation builds the variation from a description, and your team reviews and decides. AI generates the menu, your team picks the dish.
- Exit. Code variations are plain JavaScript and CSS. Results live in your own analytics, for example GA4 or BigQuery, or in Varify Reporting if you choose Varify Realtime Tracking, with CSV export either way.
Every point above is something you can hold us to. If an answer is not in a contract or on a public page, treat it as not promised.
Agency or freelancer? Let us look at your portfolio.
Book a free demo call. We go through your clients together and suggest the license model that fits.
Our conclusion: pick the tool that leaves your work to you
The question is not whether Wingify is a good product. It is whether a platform’s direction, price logic and AI fit the way you earn your money. You can check that with five questions and a day of reading terms.
If you want the product comparison next, read Varify.io vs. Wingify. For a tool shortlist built around agency criteria, see our guide to A/B testing tools for agencies.