- On a phone
- 90%
- Larger text
- 5%
- Slow 3G
- 5%
- Gives up after
- 70–120 s
Most start with search (50%).
Black Friday is in 35 days. Rehearse now, fix what breaks, then re-run before the rush. How →
Store rehearsal for Shopify, WooCommerce, Shopware, Magento & more
Your store, in the hands of 90 different customers.
Prova sends 90 synthetic shoppers from six age groups into your store, each with their own phone, connection speed and text size. Can they find a product, add it to the cart and reach checkout? You get a report with every tap as evidence.
One-time payment · 3 Fix & Verify checks and 1 full re-run included · no order is ever placed
90synthetic shoppers
6age groups, 18 to 65+
~10 minper rehearsal
0orders placed
Meet your shoppers
Fifteen people from every age group. Not one average user.
Your team tests the store on fast phones with perfect eyesight. Your customers do not. Each Prova shopper gets their own device, connection, text size, habits and patience, based on typical behaviour for their age.
- On a phone
- 82%
- Larger text
- 8%
- Slow 3G
- 5%
- Gives up after
- 80–140 s
Most start with search (45%).
- On a phone
- 72%
- Larger text
- 15%
- Slow 3G
- 5%
- Gives up after
- 100–170 s
Most start with the menu (40%).
- On a phone
- 63%
- Larger text
- 30%
- Slow 3G
- 8%
- Gives up after
- 120–200 s
Most start with the menu (45%).
- On a phone
- 55%
- Larger text
- 45%
- Slow 3G
- 10%
- Gives up after
- 140–240 s
Most start with the menu (50%).
- On a phone
- 48%
- Larger text
- 65%
- Slow 3G
- 12%
- Gives up after
- 160–280 s
Most start with the menu (50%).
Shares are the model's weights for each age group. The same seed gives the same 90 people every time, so a re-run compares like with like.
Watch them shop
Every shopper, every step, live.
While the rehearsal runs, you see each shopper move from the home page to search, product, cart and checkout. Anyone who gives up is marked, with the reason.
From a real rehearsal of a Shopify store before launch, 16 October 2026.
How it works
- Add one line to your themePaste a short verification tag into your theme. It proves the store is yours.
- Pay €35 onceSecure checkout with Stripe. No subscription.
- Get your report90 shoppers rehearse your store. We email the report as soon as it is ready.
- Fix, then verifyRe-check a single finding in your dashboard, or bring back the same 90 shoppers for a full before/after. Both included.
Before Black Friday · 27 November
Your busiest day is the worst day to find a broken cart.
On Black Friday, more people arrive on phones and slow connections at the same time, many for the first time. A tap that lands on the wrong page or a checkout button they cannot find costs you on that day most.
- Rehearse this week90 shoppers from six age groups try your store, on phones, 4G and slow 3G.
- Fix what they hitEach finding comes with the screen and a fix, or 34Devs fixes it for a fixed price.
- Re-run before the 27thThe same 90 shoppers return, so you see the fix worked before your customers arrive.
A real finding
5 shoppers tapped a product and landed on the FAQ.
On a store about to launch, a card in the search results had an invisible clickable layer that covered the product above it. Nobody on the team had noticed: on their own phones, it looked fine. Prova caught it because 25 different people searched.
One line of CSS later, the same 90 shoppers came back.
5→0wrong taps, before → after
What Prova catches
The problems you cannot see on your own phone.
- Taps that go to the wrong pageAn invisible layer covers a product, so a tap opens something else.
- Tap targets that are too smallButtons and links with a tap area under 24 px (WCAG minimum), and under 44 px as a suggestion.
- Pages that load too slowlyPages that take longer than a shopper is willing to wait, on 4G or 3G.
- Search that finds nothingNo search icon on mobile, or no results for common words.
- Menus that hide productsCollections that cannot be reached from the mobile menu.
- Add to cart that failsMissing button, no feedback, or an error in the wrong language.
- Checkout that is hard to reachNo visible checkout button in the cart drawer or cart page.
- Pages wider than the screenSideways scrolling on phones, usually from an image or table.
- German shop essentialsFor stores selling in Germany: is the Impressum, the withdrawal button, the unit price or the shipping note missing from the page?
- Accessibility barriersAutomated WCAG 2.2 test on home, product, cart and checkout, and adding to cart with the keyboard only. For German stores with a BFSG note.
- Core Web VitalsLCP, CLS and an INP estimate per page type, measured on phones with 4G.
- Checkout frictionAccount required? How many form fields? Which payment methods, and for Germany: PayPal and invoice?
- The first phone screenProduct or shop button, search, shipping info, price and "Add to cart" before scrolling.
Marked issues are always reported as high priority, even if only one shopper hit them.
New in the report
From findings to the fix.
- Your first three jobsPage 2: ranked by measured effect and effort, with what to change.
- Page by pageEach problem marked on a real screen, with "what happened → what to change → expected effect" (a simulation, not a promise).
- A video per serious findingA 5–10 second clip of the moment the shopper got stuck, with a QR code in the PDF.
- Ready-made fixesFor clear defects, CSS with your store's own selector to paste; otherwise how to check it.
- Revenue estimateOptional: enter monthly visitors and average order value, and the report turns the measured loss into euros, with the calculation shown.
- Interactive reportA private link with the report, all videos and a session player for every shopper.
The report
Six pages you can hand to your developer.
Every finding comes with the screen of a shopper who hit it, the spot they tapped and a fix. In English, German or Turkish, by email.
01 SummaryThe verdict and the key numbers at a glance.
02 Your first three jobsRanked by measured effect and effort, each with what happened, what to change and the expected effect.
03 Page by pageHome, collection, product, cart and checkout, with the problem marked on a real screen.
04 Home and product page checksIs the product, search, price and "Add to cart" on the first phone screen? With fixes for your platform.
05 German shop checksImpressum, withdrawal button, unit price and more, for stores selling in Germany.
06 Speed and heatmapsPage load as shoppers felt it, and where they tapped.
Why not just…
How Prova compares.
| Prova | Testing it yourself | Session recordings | Paid user testing | |
|---|---|---|---|---|
| Needs visitors first | No | No | Yes | No |
| Six age groups, many devices | Yes, 90 shoppers | Your own phone | Whoever visits | A handful of testers |
| Slow connections and larger text | Built in | Rarely | Only if visitors have them | Depends on testers |
| Tells you what to fix | Ranked, with fixes | — | You watch and decide | Notes per tester |
| Same people after your fix | Yes, re-run included | — | No | Rarely |
| Time to result | About 10 minutes | An afternoon | Days to weeks | Days |
Prova does not replace analytics or user research with people. It finds the friction before your customers do.
47 seconds
Prova in under a minute.
What happens when 90 shoppers rehearse a store, from the first tap to the report.
One rehearsal
€35
one-time · no subscription
- 90 synthetic shoppers from six age groups test your store
- Prioritised issue report: PDF with evidence and fixes, plus an interactive web report with videos
- 3 targeted Fix & Verify checks: after a fix, re-check a single finding instead of the whole store
- 1 full re-run with all 90 shoppers and a before/after comparison
- Share findings with your developer: a read-only link per finding, valid 30 days, revocable
- Customer dashboard in English and German: every finding, its status and your checks
- German shop checks for stores selling in Germany · report in English, German or Turkish
The 3 Fix & Verify checks are valid for 90 days from delivery of your first report. A check counts only when it delivers a valid result; technical failures on our side do not use one up. To protect your store from too many test visits in a short time, a check right after a full rehearsal or re-run waits for a protection period of about 35 minutes, so we do not give an exact start time.
Want us to fix the findings too? Pick them on your order page and 34Devs sends you a fixed-price quote.
Prova is built by 34Devs, a Shopify agency in Korschenbroich, Germany. We built it to rehearse the stores we launch for our own clients, before their customers arrive.
Questions
Is it safe for my store?
Yes. Shoppers browse, add one product to the cart and open checkout, and that is where they stop: no order is placed and no payment form is filled in. Checkout is only read.
Will it show up in my analytics?
Mostly not. Shoppers block the events your pixels send: Meta, Google Analytics and Ads, TikTok, Pinterest, Snapchat, Microsoft, Klaviyo, Shopify Analytics and session recorders like Clarity or Hotjar, so your ad campaigns do not learn from fake add-to-carts. Server-side tracking (for example a Conversions API plugin) runs on your server and cannot be blocked from the browser. No customer email is entered, so no abandoned-checkout emails are sent.
Why do I add a tag to my theme?
It proves the store is yours. Only someone who can edit your theme can add it, so nobody can send Prova to a store they do not own. Keep it while you use your re-run and Fix & Verify checks; you can remove it afterwards.
What is Fix & Verify?
After you fix a finding, mark it as fixed in your dashboard and start a check. Prova re-tests only that finding, with up to 24 of the shoppers who hit it or a direct check of the affected page, and shows whether it is fixed. You have 3 checks, valid for 90 days from delivery of your first report; a check counts only when it delivers a valid result. Some findings can only be confirmed with the full re-run; the dashboard tells you which. Right after a full rehearsal or re-run, a check waits for a protection period of about 35 minutes, so we do not give an exact start time.
Are these real people?
No. They are synthetic shoppers in a real browser. Their devices, connection speeds, patience and habits are modelled on typical behaviour by age, not measured from your visitors. Prova finds friction; it does not predict your conversion rate.
Which stores work?
Shopify, WooCommerce, Shopware 5 and 6, Magento 2, PrestaShop and JTL-Shop, and most other shops including custom-built ones: Prova reads the products from your sitemap and the standard product data (schema.org) on your product pages, which most shops have for Google. We check this before you pay; if your products cannot be read, nothing is charged. Shopify stores can also be password-protected (enter your storefront password when you order).
What are the German shop checks?
If your customers are in Germany, we also check whether your home page and a product page show what German shops usually need: Impressum, privacy policy, terms, cancellation policy, the withdrawal button, shipping information, the tax note and, where it applies, the unit price. We only check visibility; it is not legal advice.
How long does it take?
A rehearsal takes about 10 minutes once it starts. Runs are queued one at a time, and we send the report by email when it is ready.
What if something goes wrong?
If a rehearsal cannot complete for a reason on our side, you get a full refund.
Find out before your customers do.
90 shoppers · about 10 minutes · €35 once, re-run included.
Rehearse my store