For a long time I priced products the lazy way: take the cost, add a markup, done. When a competitor went cheaper, I matched them. It felt safe. It was also the slowest possible way to leave money on the table.
Here's the thing I wish I'd internalized earlier: price is the single biggest lever on profit. Raise a 1,000-yen product (600 cost) to 1,100 and the extra 100 is pure profit — gross profit jumps 25% with zero added traffic. Growing sessions by 25% is a quarter of brutal work for the same result. Yet most of us leave prices on autopilot.
This post is how I think about pricing now: the three ways to set a price, how much freedom you actually have by industry, a four-step way to raise profit, and how to measure whether a price change actually worked.
TL;DR
- Three ways to price: cost-based (floor), competitor-based (range), value-based (ceiling) — layer all three
- Pricing freedom varies hugely by industry — brand and uniqueness buy you room
- Profit-max in four steps: know true cost, articulate value, design tiers, test
- Discounting is a last resort — it cuts AOV and brand at the same time
- Judge a price change by RPS (revenue per session), not total revenue
1. The three ways to set a price
Every pricing method is one of three, and the mistake is using only one.
Cost-based adds your target profit on top of cost. Guarantees you don't lose money, ignores whether the price feels right to a buyer. It's your floor, not your answer.
Competitor-based matches the going rate. Safe until someone starts discounting and drags you into a war. Good for sensing the range.
Value-based works back from what the customer feels it's worth. This is where the profit is — but only if your branding and copy actually convey that value. The more unique your product, the more this pays.
In practice you stack them: cost sets the floor, competitors set the range, value aims for the ceiling. The quadrant above is the lens I use — products sitting in "high value, low price" have room to raise today.
2. How much pricing freedom your industry gives you
The strategies on the table change a lot by what you sell.
Cosmetics and supplements differentiate on brand and formula, so value-based premium pricing is realistic. Electronics get compared by model number and stay chained to competitor pricing. If you're in a low-freedom category, you don't price higher directly — you add value around the price: bundles, shipping framing, warranty, setup.
3. Four steps I use to raise profit
- Know your true cost — purchase price plus shipping, payment fees, packaging. That's the floor.
- Put your value into words — if you can't articulate why you're chosen, customers only compare on price.
- Design tiers — three options instead of one; the middle gets picked (decoy effect) and AOV climbs.
- Test — raise a few products, watch profit. Units can dip and profit can still grow, like the before/after above.
4. Why discounting is the move I avoid
A sale works tonight and hurts for months. It drops AOV (a lowered price is hard to raise), trains customers to wait, and the volume needed to recover a discount is usually more than you'll get. Before touching price, I fix how value is communicated — reviews, usage photos, clear warranty. When I must move with price, I use bundle or member pricing that protects AOV instead of a blanket cut.
5. Measure the change with RPS, not revenue
This is the part that took me too long. After a price change, "revenue went up" tells you almost nothing — revenue also moves with traffic. If you bumped ads the same month, the price effect is buried.
Two metrics isolate it:
- AOV = revenue ÷ orders — did people buy more per order?
- RPS = revenue ÷ sessions — did each visit produce more, accounting for the fact that price also moves conversion rate?
GA4 is session-first, so slicing "RPS by the channel where I changed price" is painful. This is actually the problem I'm building RevenueScope to solve — it lines up RPS and AOV by channel from a revenue-first view, so a price change gets verified in numbers instead of vibes.
How do you set prices today — cost-plus, competitor-matching, or do you work back from value? And do you check the effect on revenue per session, or just total sales?
Top comments (0)