Hi, I’m Kayla. I run a small candle shop on Shopify, and a tiny notes app on the side. Pricing used to make my stomach flip. Charge too little, I feel silly. Charge too much, folks bounce. So I ran split tests. Real ones, with live buyers, not pretend charts.
Let me explain what I did, what went wrong, and what I’d do again tomorrow.
The tools I actually used
- Shopify + Intelligems (the exact app I used – it sets prices by visitor, smooth and quiet)
- Stripe for payments on my app
- PostHog flags for routing app visitors to different price pages (their quick testing guide helped me set things up fast)
- Google Sheets for simple math (yes, still)
Quick aside: I discovered that the same “you get what you pay for” rule applies far beyond pricing software. I used to hop on product-demo calls with the free webcam baked into my laptop—until I read this eye-opening piece on why free webcams just don't cut it. It lays out how grainy video quietly erodes trust and explains the ROI of upgrading your gear—super helpful if you pitch live or record course content like I do.
For a broader look at how other small shops run statistically sound price tests, I often browse Optimization World for fresh tactics and sanity-checks.
One of my favorite reads there is this full play-by-play of another merchant’s price experiment, which you can find here.
I’ve tried fancy dashboards. But I keep coming back to clean tables and clear goals. Cash per visitor. Profit per order. That kind of thing.
Candle shop test: $18 vs $22 for a 12 oz jar
I love this one because it scared me. I sell a 12 oz soy candle. COGS is about $7. Shipping runs near $4. We tested two prices for two weeks: half of folks saw $18, half saw $22.
- Traffic: 16,214 sessions total (split even)
- Time: 14 days, no promos, no big holidays
Results:
- $18 price: 3.9% bought
- $22 price: 3.1% bought
At first glance, $18 looks better. More buyers, more happy pings on my phone. But here’s the thing: profit per order matters more than applause.
- Profit per order at $18 = $18 – $7 – $4 = $7
- Profit per order at $22 = $22 – $7 – $4 = $11
Profit per visitor:
- $18: 0.039 x $7 = $0.27
- $22: 0.031 x $11 = $0.34
So the $22 price won. Fewer orders, but more money in the jar. And returns? Same. Reviews? Same. Only my nerves changed.
Side note: hot months do weird things. Candles melt in trucks. I don’t test during heat waves now. Learned that the hard way.
App test: $8 vs $10 per month (with a 14-day trial)
My notes app is tiny. It’s for teachers and busy parents—people like me—who need quick lists and calm screens.
Set up:
- New signups split 50/50 with PostHog flags
- Stripe plans at $8 and $10
- Me watching Sheets at midnight like a hawk
Results after 30 days:
- $8 plan: 6.4% of trials became paying
- $10 plan: 5.8% of trials became paying
- 30-day churn: 7.2% ($8) vs 6.9% ($10) — close enough
Money math per signup:
- $8: 0.064 x $8 = $0.51 MRR per signup
- $10: 0.058 x $10 = $0.58 MRR per signup
$10 won. A little less “Yes,” but more cash per person. Support tickets didn’t spike. My favorite email was, “Honestly, ten bucks is fair.” Thank you, Sam.
Tiny twist: I tried adding an annual plan too.
- 20% off yearly ($96): 18% picked it
- 30% off yearly ($84): 26% picked it
Cash looked bigger at 30% off. But refunds stung more when folks bailed early. I went back to 20% off. Calm beats chaos.
Course presale test: $149 vs $179
I made a short course on simple email copy. Not fancy. Tight scripts, real examples, and my messy drafts. We tested two presale prices with the same page, same bonuses, same friendly tone.
- 2,000 leads split even from my list
- 5 days, no timers, just a clear close date
Results:
- $149: 3.2% bought (32 orders, $4,768 gross)
- $179: 2.9% bought (29 orders, $5,191 gross)
- Refunds: 6% vs 5% — almost the same
$179 made more and didn’t dent my inbox. A few folks asked for payment plans. So I added 2-pay at $95 and kept the $179. Win-win.
If you’re curious how dialing in a sales page itself (not just the tag on the invoice) can lift revenue, I learned a ton from this ClickFunnels split-test teardown.
What felt great (and what bugged me)
What I liked:
- Clear wins on profit, not just clicks
- Easy set up with Intelligems on Shopify
- Simple flags in PostHog for the app
- Seeing “value” emails instead of “too pricey” rants
What bugged me:
- Price flicker on cache. One visitor saw two prices due to a weird refresh. I pinned price per session after that.
- Running tests during busy weeks. One newsletter skewed a whole day. Now I pause email blasts or mark the spike in my sheet.
- My own fear. I had to stop peeking every hour. Let the sample grow. Go eat lunch.
Guardrails I use now
- Don’t test during big sales weekend or heat waves (candles).
- Freeze promos. No added gifts or pop-ups mid test.
- Set a floor. I won’t drop below break-even just to watch graphs dance.
- Track profit per visitor, not only conversion rate.
- Keep tests 14 days if traffic is normal. Longer if slow.
- Be kind to past buyers. I don’t show a higher price on a return visit for 30 days.
Tiny lessons that stuck
- People won’t read your margin sheet, but they can feel your value. Raise price and raise care too.
- One clean sentence on the page beats three fancy badges.
- Shipping costs are part of price. Say it. I added, “We eat part of shipping so your candle makes it safe.” Complaints went down.
- Optimization isn’t just for stores or SaaS—analysts even crunch data to build better sports rosters, as shown in this hockey lineup case study.
Before we wrap, here’s a fun parallel: a healthy commercial exchange is really just a relationship where both sides know what they’re giving and getting. That same give-and-take shows up in modern dating scenes, too. For instance, locals who want clear expectations and mutually beneficial arrangements often look into a sugar daddy Lenexa guide—there you’ll find straightforward advice on setting boundaries, understanding value on both sides, and keeping communication honest, lessons that map neatly onto fair pricing strategies in business.
You know what? The fear fades. The math helps. And the messages get nicer when your product stays honest.
My take, as a human who sells stuff
Split testing price works. It kept my candle shop alive when wax and jars got pricey. It gave my app some breathing room. I didn’t need magic. I needed a fair test and a quiet mind.
Would I tell a friend to do it? Yep. Start small. Test one product. Pick two clear prices. Run it long enough. Then choose the one that pays you fairly and still feels good to say out loud.
If you ever feel stuck, write the price on a sticky note and leave it on your screen for a day. If you still feel okay seeing it at dinner time, test it. If your stomach flips, listen to that too. Data guides. Your gut still matters.