Ecommerce consultant vs agency vs fractional: who owns the P&L?

The hire question is rarely “do we need help?” Most fashion and beauty brands already have help. A media partner. A theme shop. A freelancer on /products/ templates. A fractional operator in the weekly meeting. The useful question is who owns the P&L when contribution is thin, paid is up, and the site still converts like fashion often does.

Job titles will not answer that. “Consultant” can mean a deck, a CRO audit, or someone who builds the system that changes the number. “Agency” can mean media, a retainer, or a theme rebuild. “Fractional” can mean a senior operator two days a week, or a title with no build capacity.

This piece compares the common models by what they own, not by how they describe themselves. It will not tell you which one to hire. That call depends on whether your constraint is mix, creative, unit economics, operations, or the join between them.

Who owns the P&L

Ownership is not a slide. It is the person or partner who can change a commercial number and be held to whether it moved.

On a fashion or beauty P&L the loud tiles are easy to name: revenue, ROAS, conversion rate, AOV. The number that usually decides whether the business is healthier is contribution after product cost, shipping, payment, returns, markdown, ads, and the variable cost of running the store. A partner can look successful on a platform tile and still leave contribution worse.

Ask four questions of any model you are considering:

Which number do they accept as theirs?
What can they change without another supplier?
What do they hand back as “not our scope”?
How will you measure whether the work paid for itself?

If the answers are “awareness”, “the theme”, “the ads account”, or “alignment”, you do not yet have a P&L owner. You have a specialist. Specialists are useful. They are not interchangeable with someone who will sit on contribution.

A beauty brand replenishing a serum and a fashion brand living on /collections/new-in drops are not the same P&L. Replenishment, size, shade, returns and markdown change who should own which line. Treat “ecommerce help” as one category and you will hire the wrong shape.

What an ecommerce consultant typically owns

“Consultant” is the loosest word in the hire vocabulary. In practice it usually means one of three jobs.

Advice. A senior person reviews the business, writes a plan, and leaves. Useful when the team can implement and the gap is judgment. Weak when the plan dies in the gap between /checkout and the ads account.

Diagnosis. Someone investigates a stuck number. Conversion, payback, retention. Then they hand back a prioritised list. Useful when you already have builders. Incomplete if nobody then tests, measures and implements.

Delivery. Someone frames one commercial objective, investigates, designs a route, and builds or integrates the working system. An experiment, an automation, a measurement join, a piece of software. Then they measure whether it worked.

Only the third shape owns a P&L line in any durable sense. The first two can still be the right hire if your constraint is decision quality and you have people who will ship.

When you hear “ecommerce consultant”, ask what they will still be in the room for after the workshop. If the answer is a PDF, they do not own the number. If the answer is a retainer with no named outcome, they own presence. If the answer is a defined objective, a build, and a measured effect, you are closer to a partner who can sit on contribution.

TwoKai sits in that last shape. Start from the commercial opportunity. Then engineer the route around the stack you already have. Not as a media agency, theme shop, or platform reseller.

What a media agency typically owns

A media agency typically owns spend, the platform accounts, creative testing cadence, and the ROAS or CPA tiles those platforms report. The good ones also own incrementality questions. They will tell you when a 4× ROAS is recycled demand.

They do not, by default, own:

whether /products/vitamin-c-serum or /products/linen-shirt is the right first SKU to buy traffic against
size and shade friction on the PDP
/cart and /checkout drop-off
return rates that eat contribution after a “good” ROAS
whether the first order ever becomes a second order
the join between Meta, Google, the ESP, and the store’s order data

That is not a criticism. It is scope. Problems start when the brand treats platform ROAS as the P&L, then hires more media when contribution is the constraint.

Fashion and beauty make this worse. A drop can make ROAS look excellent for three weeks and hide a markdown problem in week five. A shade-matched foundation can look expensive to acquire and still be the healthiest unit because it replenishes. A media partner who only sees the ads account can spend more efficiently inside a bad unit. That is still a bad unit.

If your constraint is creative quality, auction pressure, or wasted spend on obvious junk, a strong media agency is often the right owner. If your constraint is unit economics, site conversion, or retention after the click, more media will not own the number you care about.

Theme shops and platform implementers

A theme shop or platform implementer typically owns the storefront: templates, speed, app installs, merchandising modules, and the look of /collections/ and /products/. On Shopify that work is familiar and often necessary. The same job exists on other stacks.

They do not own contribution. A faster theme can improve conversion and still lose money if you bought the wrong traffic, priced into markdown, or attached eight apps that tax every order. A beautiful PDP does not fix a first-SKU problem.

Treat this model as build capacity for the storefront, not as a growth owner. Hire it when the site is the constraint: broken templates, slow pages, untested checkout, missing size or shade tools. Do not hire it as a substitute for someone who will sit on payback.

The common failure mode is a rebuild sold as a growth programme. Rebuilds can be the right route. They are expensive when the P&L problem was mix, offer, or measurement. Ask what commercial number the rebuild is for, and how you will measure it.

What a fractional operator typically owns

A fractional head of ecommerce, CMO or COO typically owns weekly decisions: what to prioritise, what to stop, which supplier to brief, which report the leadership team should believe. The value is judgment and coordination, not a deliverable catalogue.

That can be the highest-leverage hire in a founder-led brand, if they can change the work, not only narrate it. The failure mode is a senior title in the meeting and the same gap underneath. Nobody builds. Nobody tests. Nobody joins the numbers.

Ask what they will do with their days. If the calendar is reporting, agency management and “alignment”, you have bought a coordinator. If the calendar includes investigating a stuck number, briefing a test, and staying until the measured effect is in, you have bought an operator.

Fractional does not automatically own the P&L. Ownership still depends on mandate and on whether anyone will implement. Two days a week can own a number if the scope is one objective and the builders exist. It cannot own five workstreams and a rebuild.

Specialist vs generalist in fashion and beauty

Fashion and beauty are not a vibe. They change the P&L.

Fashion often lives with lower conversion, higher returns, size risk, seasonal drops, and markdown. The second order is a different problem from the first. A partner who only knows subscription SaaS or a generic “DTC playbook” will misread a 1–2% conversion rate as a site emergency.

Beauty often lives with replenishment, shade, kits, gift-with-purchase, and a first SKU that either creates a repeater or a one-time discounter. A partner who only knows fashion drops will push acquisition as if every order were a one-off.

Specialists who understand those mechanics are worth more than a generalist with a longer service list. The specialist still has to own a number. A fashion-fluent theme shop that does not touch contribution is still a theme shop. A beauty-fluent media agency that will not discuss payback is still a media agency.

TwoKai works in fashion and beauty ecommerce by design. We do not pretend one firm should own media, brand, theme and P&L at once. When a project needs that depth, work with the partners you already have, or bring in a specialist. Keep one commercial objective in control.

How to choose: match the gap, not the title

Choose the model that matches the constraint.

Creative and auction waste → a media agency that will test and tell you when the tile is lying.
Storefront friction → a dedicated implementer, not a strategy workshop.
Decision quality and supplier chaos → a fractional operator with a mandate, not another retainer.
A stuck commercial number that sits between systems → a consultant who will investigate, build and measure, not a deck.

Most stuck brands have more than one gap. That does not mean you hire all four. It means you prioritise the constraint that is eating contribution now. Paid can look like the problem when /checkout is the leak. Conversion can look like the problem when the first SKU never repeats. Retention can look like an email problem when the only people who come back are the people you trained to wait for 30% off.

A practical sequence. Name the number you care about (contribution, payback, second-order rate). Not the loudest dashboard. Name what you can already change in-house. Name what a current supplier already owns. Hire for the remainder. If you cannot name the remainder, you are not ready to hire a shape. You are ready for a short investigation.

Do not hire a full-service promise because it is simpler to sign. Retainers hide the ownership question. Someone still has to own the P&L line. If that person is you, say so, and hire specialists who will ship their part.

What's missing

This article can tell you who typically owns which number, and how the common models differ. It cannot tell you whether your constraint is traffic mix, creative, unit economics, operations, or the system that joins them. That call is an investigation on your orders, your stack and your team. Not a job-title comparison.

If you want that investigation led as one commercial piece of work, Explore an opportunity with TwoKai.