Short answer
CRO ROI is the extra gross profit a conversion rate optimization program produces, minus what the program costs, divided by that cost. To estimate it before hiring, multiply your annual revenue by a realistic lift in revenue per visitor and by your margin, then compare the result with the full yearly cost, including software and your own team’s time.
Key takeaways
- Calculate return on gross profit, not revenue. A 5% revenue lift at a 45% gross margin adds gross profit equal to 2.25% of revenue: $112,500 on a $5M store, not $250,000.
- Most tests do not win. Optimizely’s analysis of its customers’ experiments found only 12% produce a statistically significant improvement on the primary metric.
- Find your break-even lift first. If the lift needed to cover the program’s cost looks out of reach on your traffic, the program is too big for your store today.
- Lift builds over months, so year one usually shows a lower return than the run rate going into year two.
What goes into a CRO ROI calculation?
Will a CRO program return more than it costs, and how long will that take? You can estimate both before you sign, using five inputs you already have plus two honest assumptions.
| Input | Where to find it | Example |
|---|---|---|
| Annual online revenue | Shopify analytics | $5,000,000 |
| Gross margin after variable costs | Your P&L: revenue minus product cost, shipping, payment fees and returns | 45% |
| Program cost per month | The agency’s quote | $8,000 |
| Software and tools per month | Quote or vendor pricing | $400 |
| Your team’s time | Hours per month times loaded hourly cost | Leave out at first, add later |
| Assumed lift in revenue per visitor by month 12 | An assumption you choose (see below) | 4%, 8% or 12% |
| How quickly lift builds | An assumption you choose | Straight line from month 1 to month 12 |
Tip
Use revenue per visitor, not conversion rate, as the lift you model. A test that raises conversion rate by pushing discounts can lower order value and leave revenue flat. Revenue per visitor captures both.
For the software line, check real prices. Convert’s pricing page, for example, lists $399 a month on monthly billing for its Growth plan. For the retainer, one agency’s published 2026 pricing guide puts mid-tier retainers at $5,000 to $15,000 a month.
Book a call
A free 30-minute call about your store. Shopify Plus Partner, rated 5.0 from 8 reviews on Shopify’s Partner Directory.
Step 1: What lift do you need just to break even?
Start here, because it tells you whether the rest of the exercise is worth doing.
Break-even lift = yearly program cost ÷ (annual revenue × gross margin)
At $8,400 a month all-in ($8,000 retainer plus $400 software), the yearly cost is $100,800. With a 45% margin, the average lift in revenue you need over the year looks like this:
| Annual revenue | Profit from each 1% lift | Average lift needed to break even in year one |
|---|---|---|
| $1M | $4,500 | 22.4% |
| $3M | $13,500 | 7.5% |
| $5M | $22,500 | 4.5% |
| $10M | $45,000 | 2.2% |
The same program is a stretch at $1M and comfortable at $10M. The fixed cost of a testing program spreads across your revenue, so its return improves sharply as the store grows. If your store is near $1M, a smaller scope (an audit, a fix-first project, or fewer tests a month) usually beats a full retainer.
Your conversion rate also matters. If yours is well below the averages in ecommerce conversion rate benchmarks, there is more room for lift. Shopify’s own conversion rate guide cites a global average of 2.66% across more than 400 brands, from Dynamic Yield data, with wide gaps between industries. You can work out your own rate with the ecommerce conversion rate calculator guide.
Step 2: How do you choose a realistic lift?
This is where most projections go wrong. Agencies quote their best client stories; you need a planning number.
Two facts keep the estimate honest:
- Most tests lose or tie. Optimizely’s analysis of experiments run on its platform found that “Only 12% of experiments produce a statistically significant improvement on the primary metric.” The same analysis says the median company runs 34 experiments a year.
- Some lift does not need a test. Baymard Institute estimates that the average large ecommerce site can gain a 35.26% increase in conversion rate through better checkout design, counting only checkout issues it has documented as solvable. That is a ceiling for large sites, not a forecast, but it shows why fixing known checkout problems often pays faster than testing new ideas.
A simple way to build your number: estimate how many tests will ship in a year, apply a win rate, and give each winner a modest effect. If a program ships four tests a month (48 a year) and about one in eight wins, that is six winners. If each winner adds 1% to sitewide revenue per visitor, six winners compound to about 6.2%. Some winners touch a single page and add much less; a checkout fix can add more. Treat 1% per winner as a planning figure, then ask the agency to defend a different one.
Step 3: Run three scenarios for year one and year two
Worth knowing
Lift does not arrive on day one. Research, tracking fixes and the first tests take weeks. A straight-line build from zero to your month-12 lift means your average lift across year one is about half the month-12 figure.
Here is the $5M store from the table above, at $100,800 a year all-in and a 45% margin:
| Scenario | Lift by month 12 | Extra revenue in year one | Extra gross profit in year one | Year-one ROI | ROI at the month-12 run rate |
|---|---|---|---|---|---|
| Cautious | 4% | $100,000 | $45,000 | −55% | −11% |
| Middle | 8% | $200,000 | $90,000 | −11% | +79% |
| Strong | 12% | $300,000 | $135,000 | +34% | +168% |
These figures are an illustration built from the assumptions above, not a forecast or a client result.
How to read it:
- Year-one ROI counts only what the program earns in its first twelve months.
- Run-rate ROI asks what happens if you keep paying the same fee and the month-12 lift holds for the next year. Lift can fade as products, prices and traffic change, so treat it as an upper bound unless you keep testing.
In the middle case, the program loses a little money in year one and earns well in year two. That is a common shape, and it is why a three-month trial judged on revenue rarely tells you much. If even your strong case looks thin, reduce scope or wait until revenue grows.
Two adjustments to make before you decide:
- Add your team’s time. Ten hours a month of approvals and content at a loaded cost of $75 an hour is $9,000 a year. Add it to the cost line.
- Use your real margin. A store with 60% margins gets far more profit from the same lift than a store with 30%. If you are unsure of your number, start with Shopify profit margin optimization.
How do you check ROI once the program is running?
An estimate is only useful if you can check it later. Ask for these from the start:
| What to track | Why |
|---|---|
| A log of every test: hypothesis, page, result, estimated effect | Lets you count winners and see whether the win rate matches your plan |
| Revenue per visitor and gross profit per visitor, monthly | Shows the lift that matters, not just conversion rate |
| A note of outside changes: promotions, price changes, new channels | Stops a sale or a viral post being counted as CRO lift |
| Holdback or before-and-after checks on the biggest changes | Confirms that lift from a winning test survives after launch |
| Total program cost, including tools and your hours | Keeps the cost side of ROI honest |
Watch out
Be skeptical of reports that add up every winning test’s lift and present the total as growth. Effects measured during a test can shrink after launch, and tests on different pages can overlap. A smaller, verified number is more useful than a large, theoretical one.
How Build Grow Scale approaches ROI on CRO work
Build Grow Scale’s process page states the aim plainly: have the work pay for itself before the first invoice goes out. The order of work is set up for that. The first two weeks go to fixing what is provably broken, with no test needed. Review then runs from the checkout back toward the homepage, because that is where money changes hands. Tracking is audited and repaired before decisions rely on it, and by roughly week six to eight the data is clean enough for week-by-week testing. You can see what is included in the conversion rate optimization agency service and run the break-even math above against it.
FAQ
What is a good ROI for a CRO program?
There is no single benchmark, so judge it against your own break-even point. A program should at least cover its full cost, including software and your team’s time, from extra gross profit within a period you can afford to wait (for many stores, about a year to eighteen months; set your own). If your middle scenario cannot reach that, ask for a smaller scope or a fix-first project instead.
Should I calculate CRO ROI on revenue or on profit?
On gross profit. A revenue lift is not all yours to keep: product cost, shipping, payment fees and returns take their share first. A 10% lift in revenue for a store with a 45% margin adds 4.5% of revenue in profit dollars. Calculating on revenue makes almost any program look profitable and leads to overspending.
How long does it take for a CRO program to pay back?
It depends on store size, margin and how fast lift builds. In the worked example above, a $5M store paying $8,400 a month reaches a positive run rate in the middle scenario during year one but does not recover its full year-one cost until year two. Fast fixes to broken checkout steps can shorten that a lot.
Can a store with low traffic get a positive CRO ROI?
Yes, but usually not from a testing retainer. With low traffic, each test takes a long time to reach a clear answer, so fees go into waiting. A fixed-fee audit plus direct fixes to known problems, such as surprise shipping costs or a long checkout, costs less and can pay back faster. Add testing once traffic supports it.
What should I ask an agency about ROI before signing?
Ask how many tests they expect to ship each month, what share usually win, how they measure revenue per visitor rather than conversion rate alone, and how they separate their lift from promotions and seasonality. Then ask them to fill in the break-even table above for your store. A good agency will show its assumptions instead of quoting a best-case result.
Book a call
A free 30-minute call about your store. Shopify Plus Partner, rated 5.0 from 8 reviews on Shopify’s Partner Directory.
Sources
- Optimizely, “Top 10 takeaways from running 173,000 experiments,” https://www.optimizely.com/insights/top-10-takeaways-from-running-127000-experiments/ (fetched 2026-10-23)
- Baymard Institute, Cart Abandonment Rate statistics, https://baymard.com/lists/cart-abandonment-rate (fetched 2026-10-23)
- Shopify, “Ecommerce Conversion Rate” guide, https://www.shopify.com/blog/ecommerce-conversion-rate (fetched 2026-10-23)
- Convert, pricing page, https://www.convert.com/pricing/ (fetched 2026-10-23)
- An agency’s published 2026 CRO retainer pricing guide, https://clicksgeek.com/cro-agency-retainer-costs/ (fetched 2026-10-23)