Selling CRO as an agency service

There's a good warning to hear first:

"Conversion optimization doesn't align with small business." Peep Laja, founder of the CRO agency Speero, in his lessons from running it

He gave two reasons. Small sites have too little traffic to test, and a small rise on small sales can't pay a big fee. He's right on both. So you sell smaller clients a different service, at a price their sales can pay for.

What CRO companies charge, and who they turn away

These are the prices and limits firms list on their own sites, as of November 2026:

Look at the shape of that list. The $10,000-a-month programs are built for big sites. The one-off jobs are the small front door, and Netalico says outright that its team can do the fixes afterward for a regular fee.

That's the model to copy: a small first job, then steady monthly work.

Why small clients can't buy the big version

An A/B test shows half a site's visitors the old page and half a new page, then counts which one sells more. It needs a lot of visitors.

Say a store gets 20 sales from every 1,000 visitors, and a new page would raise that to 24. Evan Miller's free calculator says you need about 20,000 visitors on each page to spot that change. That's 40,000 visitors for one test.

• At 5,000 visits a month, one test takes eight months.
• At 25,000 visits a month, it takes about seven weeks.

Nobody pays a monthly fee for eight months to learn about one headline. Conversion Sciences tells readers that an agency shouldn't even bring up A/B testing until a site gets 500 sales or leads a month. A lead is someone who gets in touch or signs up. Below that, it says, tests give false answers and waste money.

Work out the price before you pitch

Do this sum before you quote anyone:

Yearly sales × 10% × profit margin = what a good year of your work could be worth.

Then keep your yearly fee to about a third of that. The table below uses a 10% rise in sales and a 40% profit margin, which means 40 cents of profit on each dollar of sales. Those are planning numbers and they aren't promises. Swap in the client's real margin when you know it.

This sum protects you as much as the client. A client who pays $3,000 a month to earn $10,000 a year in extra profit will work that out by spring, and you'll be fired.

Laja's agency aimed to make each client five to seven times what it charged. He also advised turning down most companies under $10 million a year in online sales. That suits a firm built on testing. You can go smaller if you sort clients like this:

• Under $1 million a year: audit only, plus advice if they want it.
• $1 million a year and up: audit, then a monthly fix plan.
• 500 or more sales or leads a month: add A/B testing.

Build the offer in three steps

Step 1: The audit

A conversion rate optimization audit is a close check of a site to find what stops people buying, and what to fix first. Sell it at a fixed price, finished in about two weeks. Here's what goes in it:

1. Buy or book on the site yourself, on a phone. Write down anything that's broken or confusing.
2. Check the tracking. Make a test purchase and check that it shows up in the client's analytics exactly one time. If the numbers are wrong, every later decision will be wrong too.
3. Find the biggest leak. Count how many people reach each step: the page, the cart or form, the finish. Look for the biggest drop where money is at stake.
4. Read what customers say. Go through reviews, support emails and sales call notes. Ask whoever answers the phone what people always ask.
5. Watch three to five people try to buy. Stay quiet while they do it. This part hurts, and it's worth the most.
6. Rank five to eight fixes. For each one, write what's wrong, how you know, what to change and who does it.
7. Present it on a 30-minute call. Finish by offering the next step.

For price, $1,800 and $5,000 from the table above mark a sensible range. Taking the audit fee off the first month of a monthly plan gives the client a good reason to continue.

Now give it a name. Clutch, a directory of agencies, lists more than 11,000 companies under conversion optimization. Laja wrote that clients kept telling him every agency had the same pitch. So skip "CRO audit" and pick a kind of client you already know. "The Dental Booking Audit" sells itself to a dentist in a way "website optimization services" never will.

Step 2: The monthly fix plan

This is the main product for most small agencies. Keep it simple and fixed:

• Two changes live each month, taken from the audit list.
• One short note on what changed and what the numbers did.
• One 30-minute call.

Two a month is a fair pace. Invesp's smallest program runs two tests a month. A fixed list at a fixed price is also easy to sell, since the client knows what arrives each month. People call set packages like this productized agency services.

You won't have proof from an A/B test, so be straight about that. Compare four weeks before a change with four weeks after, and write down anything else that moved, like a sale or a jump in ad spend. Call the result a strong hint. Clients trust you more when you don't dress it up.

Then fill the gap with asking. Put one question on the client's busiest page, shown after someone has been there a while:

What's stopping you from booking today?

Offer answers people can tap, since typing on a phone is a pain. A site with too little traffic to test still has visitors who can tell you what's wrong, and their answers become next month's list.

That's also what we built Possible to do, so take this with the obvious pinch of salt. Possible is an AI built on real sales psychology. It asks each visitor the right question at the right moment, remembers the answers and guides them to a sale, sign-up or booking. It's free to start, so you can try it on one client's site. Here's more on Possible for agencies. Fix anything broken from the audit first, because a good question won't rescue a checkout that doesn't work.

If a client wants advice and has their own people to do the work, sell that instead. Nick Disabato's consultancy, Draft, lists a $2,500-a-month plan that's mostly two one-hour calls and Slack messages.

Step 3: Testing, for clients who've earned it

Once a client passes 500 sales or leads a month, you can add A/B tests. Laja's advice was to price by the number of tests you launch each month and never by the hour. Include the research and the reporting in that price.

Be ready for a smaller client's brother-in-law, who has read about a green button that doubled someone's sales. At 5,000 visits a month that test finishes in eight months, by which time he'll have a new theory.

When a client grows past what your team can handle, send them to a bigger firm. Speero's 100,000-visitor line tells you roughly where that is. You'll have done them a favor, and they'll remember it.

Who does the work

Laja listed what good CRO takes: testing, statistics, psychology, copywriting, design, analytics and website coding. A small agency usually has the copy, design and coding already. What's usually missing is the customer research, and one person who owns the result.

Pick that person and watch the numbers. In Laja's experience, one person could run three or four clients at most without the work slipping. So four clients at $1,100 a month is $4,400, and that has to cover this person plus design and coding time. If it doesn't, raise your lowest fee or stick to audits for now.

He also warned that people who read a couple of blog posts and wing it get fired. So practice first. Run your first audit on your own agency's site, because you'll find plenty and nobody can complain. If you'd rather not build the tools yourself, white label conversion tools can sit under your own brand.

Your first client already pays you

Laja found people who come to you far easier to sell to than strangers you contact, because a stranger has to be sold on CRO and on you. A current client has already said yes to you, so you only have one sale to make. If you run their ads, you have the perfect opening: they're paying for clicks that don't buy.

Adding this to existing work is a natural agency upsell, and a short email will do it:

Hi [name]. Last month [number] people came from your ads and [number] of them bought. I can't promise cheaper clicks, but I'd like to find out what stopped the rest.

It takes two weeks and costs [price]. You'll get a ranked list of fixes and a 30-minute call to go through them. If you want us to make the fixes, the fee comes off the first month.

Mistakes that get agencies fired

• Selling hours. Hours reward you for going slowly. Sell a fixed job at a fixed price.
• Taking a share of revenue as your only pay. Laja warned that this ends in fights, because once you win, the client starts arguing about what you earned. A flat fee with a bonus for hitting a goal is much safer. Some big firms do sell pay-for-results deals, since they can afford to wait for the money. You have payroll on Friday.
• Promising a number. Invesp and ConversionTeam back results with promises, and they've run thousands of tests to earn that. Promise the work, the dates and honest reporting.
• Testing without traffic. A test that can't finish gives you a false answer and a bored client.
• Going cheap to win small clients. If $400 a month can't cover the work, the client gets poor results and you get a bad reference. Sell them the audit and part as friends.

Your first two weeks

1. Day 1: List your current clients. Write their yearly sales and monthly sales or leads next to each, then sort them using the three groups above.
2. Day 2: Pick the kind of client you know best. Name your audit and write one page on what's in it.
3. Days 3 and 4: Run the audit on your own site. Time yourself, and use that time to check your price.
4. Day 5: Send the email to your two best-fit clients.
5. Week 2: Run the first paid audit, present it and offer the monthly plan.

After a few audits, you'll have your own list of what keeps going wrong for your kind of client. That list is where real skill in conversion rate optimization starts. Another agency can't copy it quickly.

Questions about selling CRO as an agency

Go deeper on conversion tactics.