In short
- Monthly revenue equals sessions multiplied by conversion rate and average order value, so you can model CRO by changing one input at a time.
- Model conservative, base and optimistic scenarios instead of relying on a single promised lift.
- Judge CRO on contribution margin after product, shipping, payment, GST and RTO costs, not on gross revenue.
- A sound CRO business case should at least break even in the conservative scenario over a year.
- Check your sample size before planning A/B tests, because low-traffic stores may need months to prove a small lift.
To calculate the revenue impact of CRO, multiply your monthly sessions by your conversion rate and average order value to get today's revenue, then model the same sum with a realistic uplift and convert the difference into contribution margin, not just sales. Compare that monthly margin gain with what the CRO work costs, and you have a payback period before you spend anything. The rest of this guide walks through the maths with a worked example in rupees, including the COD and RTO adjustments most calculators miss.
The core formula
Every ecommerce revenue model starts with three numbers you already have in Shopify Analytics:
- Sessions per month.
- Conversion rate: orders divided by sessions.
- Average order value (AOV): revenue divided by orders.
Monthly revenue = sessions x conversion rate x AOV. CRO work usually targets conversion rate, though cart and checkout changes such as bundles or a free shipping threshold also move AOV. Because the formula multiplies, a modest improvement in both compounds.
Step 1: Establish your baseline
Use the last 90 days, not your best month. Festive peaks such as Diwali or a big sale inflate conversion rate and will make every projection look better than reality. If your business is seasonal, compare like-for-like periods instead.
Here is the example store we'll use throughout. The numbers are illustrative, not a benchmark.
| Input | Example value |
|---|---|
| Monthly sessions | 1,00,000 |
| Conversion rate | 1.5% |
| Orders per month | 1,500 |
| AOV | ₹1,800 |
| Monthly revenue | ₹27,00,000 |
Step 2: Model scenarios, not a single promise
Nobody can promise a specific CRO lift in advance, and you should be wary of anyone who does. Results depend on how many real problems the store has, how much traffic you have and how well fixes are executed. Instead, choose three scenarios and treat them as inputs you control. For this example we'll model relative conversion rate improvements of 5%, 10% and 20%.
| Scenario | New conversion rate | Orders | Revenue | Extra revenue per month |
|---|---|---|---|---|
| Baseline | 1.50% | 1,500 | ₹27,00,000 | - |
| Conservative (+5%) | 1.575% | 1,575 | ₹28,35,000 | ₹1,35,000 |
| Base (+10%) | 1.65% | 1,650 | ₹29,70,000 | ₹2,70,000 |
| Optimistic (+20%) | 1.80% | 1,800 | ₹32,40,000 | ₹5,40,000 |
Note that a 10% relative lift means going from 1.5% to 1.65%, not from 1.5% to 11.5%. Mixing up relative and absolute changes is the most common mistake in CRO business cases.
Step 3: Convert revenue into contribution margin
Extra revenue is not extra profit. For each extra order you still pay for the product, packaging, shipping, payment gateway fees and, for Indian stores, the cost of orders that are never delivered. What matters is contribution margin: what is left from each order after all variable costs.
Adjust for COD and RTO
If part of your volume is cash on delivery, some of those orders will return to origin. Those orders generate shipping costs in both directions and no revenue. Adjust extra orders before calculating margin.
Continuing the example, assume 40% of orders are COD and 20% of COD orders return to origin. That means 8% of all orders are never delivered (40% x 20%). In the base scenario, 150 extra orders become 138 delivered orders.
| Line | Per delivered order |
|---|---|
| AOV (net of GST, for simplicity) | ₹1,800 |
| Product cost (landed) | -₹540 |
| Shipping and packaging | -₹110 |
| Payment and COD fees | -₹45 |
| Share of RTO losses | -₹25 |
| Contribution per delivered order | ₹1,080 |
Base scenario: 138 delivered orders x ₹1,080 = ₹1,49,040 of extra contribution per month, or roughly ₹17.9 lakh a year. Swap in your own costs; the structure matters more than these example figures. If you report GST-inclusive prices, remember to remove GST from revenue before this step, as it isn't yours to keep.
Don't forget ad spend
If your traffic is paid, CRO doesn't increase ad spend for the same sessions, so the extra contribution falls straight through. This is why CRO often improves blended ROAS: the same budget buys more orders. It also means you can bid more aggressively for the same profitability, which is a second-order gain worth noting but hard to predict, so leave it out of the base case.
Step 4: Compare with the cost of CRO
Now compare the monthly contribution gain with the cost of the work, whether that is an agency retainer, a freelancer, testing tools or your own team's time. Using illustrative costs:
| Scenario | Extra contribution per month | Example CRO cost per month | Net gain per month |
|---|---|---|---|
| Conservative | ₹74,520 | ₹1,00,000 | -₹25,480 |
| Base | ₹1,49,040 | ₹1,00,000 | ₹49,040 |
| Optimistic | ₹2,98,080 | ₹1,00,000 | ₹1,98,080 |
Two points make this more favourable than it first looks. First, a successful change keeps paying after the project ends, so a one-off audit and implementation costs once and earns monthly. Second, gains compound: a better product page and a better checkout multiply together. Still, if only the optimistic scenario makes sense, the case is weak. A sound business case should at least break even in the conservative scenario over 12 months.
Step 5: Check whether you can measure it
Before committing to an A/B testing programme, check whether your traffic can detect the lift you're hoping for. A widely used approximation for the sample size per variant (at 95% confidence and 80% power) is:
n ≈ 16 x p x (1 - p) / d², where p is your baseline conversion rate and d is the absolute difference you want to detect.
For our store, p = 0.015 and a 10% relative lift means d = 0.0015. That gives roughly 1,05,000 sessions per variant, or about 2,10,000 sessions for a simple two-way test. At 1,00,000 sessions a month, one test would take around two months. You can check your own figures with Evan Miller's free sample size calculator.
What this means in practice:
- Lower-traffic stores should focus on fixing clear problems found in an audit (broken tracking, slow pages, missing delivery information) and measure before and after, rather than running many small tests.
- Higher-traffic stores can test more, and should test bigger, bolder changes that could plausibly produce detectable lifts.
- Everyone should test on the pages with the most traffic first, typically the top product pages and the cart.
Use a calculator to speed this up
Once you understand the maths, a calculator saves time and makes it easy to share scenarios with a co-founder or finance team. TNM is launching a free Shopify CRO calculator (coming soon) that takes your sessions, conversion rate, AOV and margin and shows conservative, base and optimistic outcomes. Until then, a spreadsheet with the tables above works perfectly well.
Common mistakes to avoid
- Using a peak month as the baseline. It inflates every projection.
- Confusing relative and absolute lifts. 1.5% to 1.65% is a 10% relative lift and a 0.15 percentage point absolute lift.
- Stopping at revenue. Margin, RTO and GST change the answer significantly.
- Ignoring mobile and desktop differences. If most traffic is mobile and mobile converts lower, model them separately.
- Assuming every test wins. Many tests are flat or negative. Your scenarios should reflect the programme as a whole, not the best individual test.
Build your CRO business case with real numbers
A good CRO partner should be able to walk you through this model using your own data before you sign anything. If you'd like help building the case, finding where your store leaks revenue and prioritising the fixes with the highest expected return, talk to our Shopify CRO team. If you want to see the kinds of quick wins that often feed the conservative scenario, read our five practical CRO tactics for Shopify.
Frequently asked questions
It varies widely with how many real problems the store has, its traffic and how well fixes are built, so treat any promised number with caution. Instead of relying on benchmarks, model a conservative, base and optimistic scenario with your own data, and make sure the investment still makes sense in the conservative case.
Profit, specifically contribution margin. Extra orders still carry product, packaging, shipping and payment costs, and for Indian stores some COD orders will return to origin. Remove GST from revenue, deduct variable costs per delivered order and compare the result with what the CRO work costs. Revenue-only ROI overstates the return, sometimes dramatically.
It depends on your conversion rate and the lift you want to detect. A common approximation is 16 x p x (1 - p) divided by the squared absolute difference, per variant. At a 1.5% conversion rate, detecting a 10% relative lift needs roughly 1,05,000 sessions per variant, which is why smaller stores should prioritise clear fixes over constant testing.
It depends on your traffic, margins and the cost of the work. Divide the total investment by the expected monthly contribution gain in each scenario to get a payback period. Remember that a successful change keeps earning after the project ends, so one-off audits and implementations often pay back faster than ongoing retainers.
Yes. The Night Marketer is launching a free Shopify CRO calculator at thenightmarketer.com/tools/shopify-cro-calculator. It takes your sessions, conversion rate, AOV and margin and shows conservative, base and optimistic revenue and profit outcomes. Until it launches, the formulas and tables in this article can be rebuilt in any spreadsheet.
