Pavlo

Making website changes that make you millions | CRO for 8-fig Shopify brands

[Pavlo @pavcro]

We deliver full CRO services for DTC brands on Shopify — research, testing, UX, development, copywriting, and more.

I’ve gathered all the work I’ve shared so far, including A/B tests, into this thread.

I’ll keep adding new insights as we go.

Check it out 👇

[Pavlo @pavcro]

@jameshyujinlee @JamesBoostBerry 🙏🤝

[Pavlo @pavcro]

@zakburgers And this is already ripped product 🤣

[Pavlo @pavcro]

In the trenches

[Pavlo @pavcro]

Everyone thinks CRO means ADDING features to increase conversions.

We did the opposite and made MORE money.

Client's cart drawer had upsells and pre-checkout popup added over time.

We removed them and simplified the flow.

Results:

→ CVR: +30.71%
→ RPV: +25.97%
→ PPV: +25.75%
→ AOV: -3.63%

More conversions but slightly lower AOV.

The main result: 25%+ more profit per visitor.

Removing friction increased conversions more than upsells increased AOV.

Sometimes the best optimization is removing things, not adding them.

[Pavlo @pavcro]

How we prioritize tests:

Impact: How much revenue could this move?
Confidence: How strong is the data backing?
Ease: How hard to implement?

Score each 1-10 and add together.

Highest score runs first.

Simple system and no random testing.

[Pavlo @pavcro]

@mouss_us actually one of the highest levers to pull

[Pavlo @pavcro]

Everyone thinks CRO means ADDING features to increase conversions.

We did the opposite and made MORE money.

Client's cart drawer had upsells and pre-checkout popup added over time.

We removed them and simplified the flow.

Results:

→ CVR: +30.71%
→ RPV: +25.97%
→ PPV: +25.75%
→ AOV: -3.63%

More conversions but slightly lower AOV.

The main result: 25%+ more profit per visitor.

Removing friction increased conversions more than upsells increased AOV.

Sometimes the best optimization is removing things, not adding them.

[Pavlo @pavcro]

@johnhickey1970 Jealous I’m not american to get these jokes lol

[Pavlo @pavcro]

@johnhickey1970 Super smart, need to try at one point

[Pavlo @pavcro]

@zachlduncan Visually is good if you don't want to code sections too much. We use Intelligems for our needs

[Pavlo @pavcro]

@BambinoShopify We usually manually go through reviews like you described but also feed them to AI to objectively get most used language and USPs.

Then just use it for our CRO ideas. One of the best research methods

[Pavlo @pavcro]

@0xROAS Advertorial gifs became so easy to get

[Pavlo @pavcro]

@paskotomasz @CarlWeische I've heard Quiz Kit is good

[Pavlo @pavcro]

@thedennis Research helps a lot. What are your main research areas?

[Pavlo @pavcro]

Most brands optimize mobile for device

We optimized for WHO uses the device

Client had standard mobile optimization:

→ Responsive design
→ Fast loading
→ Touch-friendly buttons

Mobile CVR: 2% (85% of traffic)
Desktop CVR: 3% (11% of traffic)

The insight: Client targeted people aged 60+.

Older customers have worse eyesight - they zoom differently on phones.

Standard "mobile responsive" didn't account for this.

Solution:

Increase font size by 2px on mobile.

Result:

Mobile CVR increased from 2% to 2.1% (+5%).

Mobile optimization isn't just technical responsiveness - It's understanding who your actual customers are.

Different age groups use phones differently.

Optimize for your audience, not just the device.

[Pavlo @pavcro]

We offer 3-month money-back guarantee - no questions asked.

Our clients can refund if:

→ Results are bad
→ They don't like working with us
→ They simply changed their mind

We can offer it because:

→ High test win rate
→ We are confident in ideas we propose
→ Deep technical knowledge of Shopify ecosystem and apps
→ Always overdeliver is part of our culture
→ No hanging messages in Slack
→ 1.5 years average client relationship

I don't like contract lock-ins.

I like working based on trust, results, and relationships.

We give everything to each client.

We don't scale fast to keep quality high.

If we're not delivering value, you shouldn't pay.

That's the guarantee.

[Pavlo @pavcro]

@yojimmykim Clarity always wins. That’s why we’ve seen the estimated delivery date on the PDP work, even when it’s long (7–14 days). Some visitors just need the right expectations set

[Pavlo @pavcro]

@IstvanicMarin Great explanation why you shouldn't have CVR as your primary metric!

[Pavlo @pavcro]

Announcement bar CTA test:

Week 1-3: Looked like loser
Week 4: Won

Collection page filter test:

Week 1-2: Looked like confident winner
Week 3-4: Became neutral

Pattern:

→ Early data misleads
→ Volume reveals truth

Our rule: 1,000+ conversions per variant minimum.

The difference:

→ 95% confidence tells you statistical certainty
→ 1,000+ conversions tells you if you have enough data to trust it

Small samples create false signals.
Large samples reveal real patterns.

Wait for real data and don't let it lie to you.