Should you hire a CRO consultant, a CRO agency or an in-house CRO specialist? The right pick depends on your revenue, your traffic and, above all, how much of the work your own team can take on.
If your brand needs CRO help, all three can get you there, but they don't hand you the same thing.
Most brands we talk to land on this question after a few frustrating quarters: paid traffic keeps getting more expensive, the conversion rate hasn't moved in three quarters, and someone in the weekly meeting asks whether it's time to hire a CRO person, bring in a consultant or sign with an agency.
This guide walks through what each option covers, what it costs you beyond the invoice, and which one fits your revenue and your team.
Full disclosure: We're SplitBase, a CRO agency, so you can guess where we land for most lean teams doing $10M to $100M a year. But each option wins in the right setup, and we'll show you where a consultant or an in-house hire is the better call.
A CRO consultant is an independent conversion specialist who audits your store, finds where shoppers drop off, and recommends what to fix and test so your own team can make the changes.
When people talk about CRO consulting, that's usually the work they mean: an audit, a testing roadmap, and some reviews or coaching along the way, billed by the hour, by the project or on a monthly retainer.
A consultant tells you what to change, but most don't design, code or QA the test variants themselves. You get the blueprints, and someone on your side still has to build the house.
That works if your team has a designer and a Shopify developer with spare hours. If it doesn't, the roadmap ends up in a shared doc, slipping behind every launch and bug fix.
Not every consultant stops at advice, though. Some also design and code their own tests, but then one person is your strategist, designer, developer and QA all at once, so your program runs only as many tests as their hours allow.
And being able to do all of it isn't the same as doing all of it well. Someone strong on strategy may write copy or design pages that don't meet your brand's standard, and you end up redoing part of every deliverable.
An apparel brand came to us after working with a CRO consultant for years. The brand described him as someone with good ideas who didn't really have the range to step back and see the bigger picture, dig deep into the data, or write his own copy and do his own designs.
You'll see three flavors on the market:
A consultant is at their best when you already have the team and the resources, and your goal is to train that team to run CRO itself. They can also be the right call for a second opinion or a one-off diagnosis. But if your goal is simply to get CRO done, a consultant will often be the limiting factor.
A CRO agency is a team that runs the full optimization cycle for you: customer research, hypotheses, copy, design, development, QA, A/B testing and analysis. The difference from a consultant is who builds. A consultant hands you a plan. An agency ships the tests and reports the result in revenue terms.
A simple way to picture it: a consultant is a personal trainer who writes your program and sends you to the gym. An agency is the trainer, the gym and the spotter.
The difference shows up in staffing. A working CRO program needs five roles: a strategist, a researcher or analyst, a conversion-focused designer, a front-end developer and someone accountable for reporting. At most brands past $10M, with several funnels, launches and promos running at once, you'll likely need a project manager too, to keep all of it moving.
An agency brings all of them, a consultant brings one, and an in-house hire is one person who still needs the rest.
Generalist marketing agencies list CRO next to SEO and paid media, as one item on a long menu. A specialist conversion rate optimization agency does only this work, and it shows in the research that happens before the first test.
Design is part of it too. With a brand-aligned conversion optimization agency, the designers who build your tests and landing pages work inside your brand, so a winning variant still looks like you.
An in-house CRO hire is a full-time employee, often titled CRO manager or experimentation lead, who owns testing inside your company. The upside is context and continuity. The catch is that one person rarely covers research, design, development and analysis, so the hire only works when the rest of the team has capacity to build.
That context is worth a lot. A good in-house specialist knows the brand cold, sits in the merchandising and paid media meetings, and catches promo conflicts before a test launches.
You also get all of their time. A full-time hire puts 100% of their focus on your brand, your customers and your roadmap, which neither a consultant nor an agency can offer. The trade-off is that more hours from one person still don't add up to the output of a team, since one person can only research, design, build and analyze so much in a week.
CRO is still a young field. In more than a decade of running SplitBase, we've watched the average skill level stay low, and people with deep experience are scarce. So a solid hire who doesn't need training easily costs six figures a year.
After the offer letter come ramp time, tools and the hours your designers and developers have to lend out. And when your only CRO person leaves, most of what they learned leaves with them unless someone wrote it down.
We see this all the time. A brand hires someone to own CRO, and they quickly find out how constrained they are. Every test they want to run waits behind a busy development team and a busy design team, and the program stalls. At six figures, a hire without the resources, time and process to do the work is an expensive way to stand still. For the same money, an agency does the whole job, with less hassle than an employee, and answers for the results.
There's a quieter problem, too. Unless you hire a proven CRO leader, nobody else in the company knows more about CRO than they do. That makes them hard to train, and hard to second-guess. When they're wrong, who would notice? Wrong calls in CRO can cost a brand a lot.
AI makes the solo hire look more tempting. One person with AI tools can now build a landing page or set up an A/B test in an afternoon. But launching was never the hard part. Knowing what to test, and why, still takes research, analysis and design judgment, and none of that comes bundled with the tool.
With AI, we probably need fewer people than we used to, and we can do more. But it doesn't replace skill or taste. AI can produce a design, just probably not one as good as a designer using AI. It can write code, but not code as solid as a developer using AI would ship. Copy was one of the first jobs people expected AI to take over, and brands and agencies are still hiring copywriters, because AI copy often doesn't cut it, or reads as AI.
So even a skilled person with AI tools can't do everything, and a serious brand at scale that wants a high level of execution needs a team. That's often where a single consultant falls short, and the same goes for a fractional CRO lead. A team brings the range, plus something no tool has: the pattern recognition that comes from seeing what has worked across many brands in your vertical.
The three models differ on who does the work, how fast tests go live, and what happens to the knowledge when the relationship ends.
| CRO consultant | CRO agency | In-house CRO hire | |
|---|---|---|---|
| Who does the work | Advises and plans. Your team builds. | Researches, designs, builds, tests and reports. | Owns the program. Borrows designers and developers to build. |
| Customer research | Often included in an audit. Rarely ongoing. | Ongoing in a research-led agency. Thin in a generalist one. | Possible, but competes with every other task. |
| Speed to first live test | Depends on your dev queue. | Weeks after kickoff. | After recruiting. A first test needn't take months, but a full process takes longer. |
| Cost structure | Hourly, daily, per project or monthly advisory. | Monthly retainer or fixed project fee for a whole team. | Salary, benefits, recruiting, tools and build hours. |
| Brand context | Outside view. Learns what you share. | Builds it through onboarding and research. Not in your daily meetings. | Deepest. Can also mean blind spots. |
| What happens when it ends | You keep the plan, not a running program. | You keep what the contract says you own. Agree on it upfront. | Knowledge leaves with the person unless documented. |
| Best fit | Teams that can build, or brands too small to test much. | Lean ecommerce teams with enough traffic to test. | Larger brands with design and dev capacity. |
Look at where the gaps fall. With a consultant, they land on your dev queue. With an in-house hire, they land on designers and developers borrowed from other projects. Either way, somebody on your team ends up doing that work.
The price you're quoted is only part of what you'll spend. What you're really paying for with each:
Whichever model you pick, most tests won't win. SplitBase's guide to what a CRO program should include puts it at roughly two-thirds of well-designed tests, and a Wingify study (the company behind VWO) found that only about 1 in 7 A/B tests produce a winner. That's normal, not a sign the program is failing. You're paying for a learning rate, not a lift on every test.
Under about $10M a year, a CRO consultant or a one-off audit usually fits best, because traffic is too thin for a steady testing program. From $10M to $100M, an agency fits most lean DTC teams. Past $100M, many brands build an in-house experimentation lead and keep an agency or consultant for capacity and specialist work.
Why does revenue matter so much? Mostly because it tracks traffic. As a rough guide, SplitBase looks for at least 3,000 orders a month to run A/B tests comfortably. Below that, research, landing pages and high-confidence fixes return more than a heavy testing calendar. Running a full program on thin traffic is like reading a poll with twelve respondents: you get an answer, just not one worth betting the quarter on.
| Annual revenue | Typical team | Usual best fit | Why |
|---|---|---|---|
| Under $10M | Founder plus a small marketing team | Consultant or one-off audit | Too little traffic for many tests. Fix clear problems and learn from customers first. |
| $10M to $30M | Two to five people in ecommerce, no dedicated CRO, outsourced dev | CRO agency | Enough traffic to test. No spare hands to research and build tests. |
| $30M to $100M | Growth team, maybe an in-house designer or developer | Agency with an internal owner | Test velocity starts to matter. An internal owner keeps tests in sync with promos and launches. |
| $100M and up | Dedicated product, design and dev teams | In-house lead plus outside support | Enough volume for a full-time owner. Outside help adds capacity and a fresh view. |
Do you have a designer with free hours, a conversion copywriter, a developer who knows your Shopify theme and someone who owns analytics? If two of the three are missing, a consultant's plan will stall, and a solo hire will spend most days waiting on other teams.
Having the people isn't enough, either. If their goals and incentives aren't tied to the CRO program and the person leading it, CRO work slides to the bottom of their list every week. That risk applies to a consultant or an in-house hire, since an agency brings its own team.
Most $20M DTC brands get more from a CRO agency than from a first in-house hire. At that size, the ecommerce team is usually small, with no spare designer or developer, so a lone specialist ends up writing test plans nobody has time to build. An agency brings the whole team and can have tests live within weeks.
A new hire shouldn't need months to launch a first test. That said, getting them fully onboarded and trained can take longer than it would for an agency.
An agency starts research at kickoff with a process that's already running, and launches tests in the first weeks. If we were running a $20M brand with no designer, marketer or developer to spare, we know which we'd pick.
Speed isn't the only factor, though. At this size, four questions decide it:
Hiring in-house makes sense if the support around that hire is already in place: designers, developers and copywriters with real capacity, the testing and research tools they need, and incentives that keep those teams invested in CRO. It also makes sense if you plan to build a full experimentation team over the next few years, and you're willing to wait out the ramp.
For most brands this size, an existing ecommerce manager owns the agency relationship, the approvals and a recurring call. That's a few hours a month, not a new headcount.
Switch when the bottleneck moves. Go from consultant to agency when good recommendations keep waiting on design and development. Go from agency to in-house when testing volume justifies a full-time owner and you can staff the builders. Add outside help to an in-house team when velocity stalls or one person carries the whole program.
The signals worth acting on:
We often start working with a brand at that last switch. An in-house CRO hire or a marketing manager brings us in to clear the backlog so tests run on time. Every week your site goes without a live test, you're losing money.
Whichever way you move, research, the test archive, design files and winning code should be yours by contract. Otherwise the next team starts from zero and you pay twice for the same lessons.
Transparent Labs made that kind of switch. The premium supplement brand had run internal A/B tests before, but testing volume was low and the process wasn't formalized, according to the Transparent Labs case study. It brought in SplitBase to run a formal testing program, and the first A/B tests went live within two weeks.
Judge any CRO partner on four things: where their test ideas come from, who actually does the work, how they measure results in revenue, and what you keep when it ends. Ask them to walk you through a test that lost and what it changed. Vague answers about best practices signal a checklist, not a program.
We like the losing-test question because it's hard to fake. Anyone can show you a winner; a partner who can explain what a loss taught them is showing you how they think.
Bring this table to your sales calls. Each row is one of the four checks, phrased for the option in front of you, with the answer that should worry you.
| What to check | Ask a consultant | Ask an agency | Ask an in-house candidate | Red flag |
|---|---|---|---|---|
| Where test ideas come from | What research comes before your recommendations? | Which customer research runs before the first test? | How did you pick tests in your last role? | A test list before anyone has studied your customers. |
| Who does the work | Who builds the tests, and how many hours am I buying? | Who is on my account, by role? | Who designed and coded your past tests? | "We'll sort out the build later." |
| How results are measured | How will we judge a test: conversion rate or revenue? | Walk me through a test that lost and what it changed. | Show me a test you called and why. | Wins reported only as conversion-rate lifts. |
| What you keep | Do I keep the roadmap and research files? | Who owns the research, test archive, designs and code? | Where did you document past tests? | Ownership that isn't written into the contract. |
If you only remember one row, make it the first. The clearest red flag is the same for all three options: a test roadmap that shows up before anyone has studied your customers.
SplitBase is a CRO agency for ecommerce and DTC brands, built for lean teams at roughly $10M to $100M+ a year, mostly on Shopify. It runs the full program: research, design, development, A/B testing and landing pages. For most brands it's an alternative to a first in-house hire, and a complement when you already have one.
The work starts with research, using SplitBase's 3Ps method: Patterns (how visitors behave, what customers say, and where they get stuck), Perception (how your brand is seen, how it competes, and how it is perceived), and Proof (the A/B testing). Tests come from those findings, not a swipe file.
Dr. Squatch shows how the right setup can change as a brand grows. When the brand first approached SplitBase in 2019, on-site A/B testing was "virtually non-existent," per the Dr. Squatch case study. So they hired us to run all of their testing as an agency, alongside full-site optimization and landing pages.
The research turned up something hiding in plain sight: analytics and customer surveys both showed that many shoppers bought more than one bar of soap, yet the product pages had no quantity field. The test that followed made two bars the default and lifted revenue per user by up to 54% for soap purchasers, and the case study estimates the program's total at $800,000+ in additional monthly revenue.
Over time, Dr. Squatch built a data department, with an internal CRO lead and a CRO champion. Even then, they kept us on as an agency for extra support: more tests launched than the internal team could handle alone, an outside point of view, and perspective from working with other highly successful brands in similar and different verticals. Eventually, they did a lot of CRO in-house, and an agency was part of how they got there.
A CRO audit is the smaller first step if you want an outside diagnosis before committing, and the full ecommerce CRO program is the agency option for teams that need research and execution.
If you're on Shopify Plus, our Shopify CRO agency page goes into how we work on Shopify stores specifically.
SplitBase isn't the right choice for every brand, and we'd rather say so upfront. Its programs and its CRO audit are built for stores doing $10M+ a year, so smaller stores usually get more from an independent consultant first. It also doesn't manage paid media and doesn't work on B2B or lead-generation sites.
If you want to see what a program would look like on your site, you can request a free CRO proposal.
A CRO consultant audits an ecommerce site, finds where visitors drop off, and recommends changes and A/B tests to raise the conversion rate. Most consultants advise and plan rather than build, so your own designers and developers usually create and launch the tests.
CRO consulting is advisory conversion rate optimization work: audits, test roadmaps, test reviews, and coaching for an internal team. It's usually sold by the hour, day, project or month. It differs from a CRO agency program, which also includes design, development and test execution.
A CRO consultant usually quotes a lower price than an agency, because you're paying for one person's advice. The real cost includes your team's design and development time to build every test. An agency fee bundles those roles, so compare the cost per live test, not the invoice.
Hire a freelance CRO consultant if your team already has design and development capacity and needs direction. Hire a CRO agency if you need the research, design, development and testing done for you. Most lean DTC teams between $10M and $100M fit the agency profile.
Yes, and at larger brands it's common. The in-house specialist owns priorities and the learning archive, while the agency or consultant adds research, build capacity or an outside view.
A CRO agency should have its first tests live within the first few weeks, with research running alongside them. Reliable results take longer, because each test needs full business cycles to reach a valid read. SplitBase aims for a program to pay for itself by around 90 days.
CRO consulting is often the best fit for a brand under $10M, because traffic is usually too low for a steady A/B testing program. A consultant or one-off audit can find clear problems and guide fixes. A full agency program makes more sense once traffic supports regular testing.
A fractional CRO lead is a part-time head of experimentation who runs your testing program for a set number of days a month. They set priorities, review tests and coach your team. Your own designers and developers still build the variants, so it fits brands that have builders but no program owner.
A CRO specialist usually works inside one company or agency team and runs tests day to day. A CRO consultant is an outside advisor who audits and plans, often for several clients at once. Specialists tend to build and launch tests, while most consultants recommend them.
A CRO agency can replace a first in-house hire, but it doesn't have to replace an existing team. Many brands keep an internal owner for priorities and approvals while the agency handles research, design, development and testing. SplitBase's FAQ says it often works alongside internal teams or other agencies.
Go back to that flat conversion rate and the question it raised. If your team can design and build tests, a CRO consultant's plan will ship, and it's the leaner choice. If it can't, that plan waits in the queue, and an agency that owns the work between the idea and the live test gets you there faster.
Before you compare a single proposal, look at your own team and ask who's going to build it.