How to Upskill the Marketer You Already Have (Without Sending Them to a Course)

Here is the state of AI skills on the average marketing team in 2026, in three numbers from three different surveys. The Australian industry body ADMA found that 75 percent of marketers use AI at least weekly and 29 percent have had formal training in it. The Marketing AI Institute’s 2025 study put the untrained share at 68 percent. And the US CMO Survey, the long-running Duke study, has training and development at 3.8 percent of marketing spend, down from 5.8 percent before the pandemic, in the same year Gartner reports CMOs putting 15.3 percent of their budgets into AI.

So the tools are bought, the people are using them, almost nobody has been taught, and the training line is shrinking. The usual response from a CMO who notices this is to book a course. This post is about why that is the wrong unit, and what to do instead with the marketers you already have.

The course is the wrong unit

A course teaches two things: how to prompt, and what the tool’s features are. Both are real. Both are also the parts of the skill that change every quarter, which is why the marketer who took the workshop in March is back to copying and pasting by June.

What a course cannot teach is the thing that actually caps a marketer. Every marketer is strong at about half of what a piece needs. The writer has the argument and not the visual. The demand gen manager has the channel and the test and writes copy that sounds like the last ad. The ops person has the lifecycle and the data and is asked to “just write the nurture”. The gap is not that they cannot operate a tool. It is that they cannot evaluate work in a discipline they have never produced in, so even when a tool drafts the image or the copy, they do not know whether it is good, and they default to waiting for the specialist who does.

That skill, judging work in the half of the job you do not have, grows only one way: by repeatedly briefing, reviewing and approving real pieces in that discipline, against a written standard, with a specialist available by exception. It is an apprenticeship, not a lecture, and in 2026 the apprenticeship can happen inside the marketer’s own week because the drafting is done by agents.

The MiniMBA survey of its own alumni this year made a related point from the other side. Only about ten percent of the 216 trained marketers it asked said keeping up with AI was their biggest challenge; strategic thinking was the gap they named. Take the sample with the caveat that it is small and self-selected, but the direction is useful. Marketers who have been through a structured programme do not feel short of tool knowledge. They feel short of judgment.

What actually grows

Three skills, in the order they develop.

Briefing. Stating intent, audience, the claims you will and will not make, the constraints and what good looks like, clearly enough that a draft comes back usable the first time. Most marketers brief badly because they have only ever briefed people, and people fill gaps with questions. A brief that produces a usable draft from an agent is a better brief than most teams have ever written, and the skill transfers directly to briefing humans.

Evaluation. Looking at a drafted image, a sequence or a landing page in a discipline you do not have, and knowing whether it meets the standard. This is the one that expands what a marketer can do. A writer who can judge an image against the brand look, pick the right one of four options and say what is wrong with the other three has acquired most of what they needed from a designer, without becoming one.

Execution. Putting an approved piece into motion: scheduling the publish, setting up the follow-up, knowing what happens after send. Marketers who have always handed this to ops do not know what their campaign does after it leaves them. Learning it closes the loop between the idea and the result.

Notice prompting is not on the list. It is a sub-skill of briefing and the least durable of the three.

The twelve-week plan

This is written for a CMO with a team of three to eight. It assumes you have a written brand standard (voice, look, claims) and one approval gate. If you do not, those are weeks one and two; the hub post covers the setup in expand what your marketing team can do.

Expect a particular shape to the quarter. Weeks three and four produce complaints about quality, from the specialists and from the marketers themselves, because the first evaluations are poor and the first drafts show it. Week five is usually the turn, once the brand standard has absorbed the common mistakes and the marketers have seen twenty pieces each. By week twelve the question “can the writer ship their own images” has stopped being asked, because they have shipped forty of them.

By persona

The plan is the same for everyone and the discipline each person is acquiring is different. We profiled each persona’s strengths and gap in the creative ability series; here is what each one is learning to evaluate.

The writer learns to judge images and layouts. Their advantage is that they already know what the piece is trying to say, so the question “does this image carry the argument” is one they can answer from day one. What they learn is the brand look: which of four on-message options is also on-brand.

The demand gen manager learns to judge copy and creative as a set. Their advantage is the test design; they know what the variants need to isolate. What they learn is voice, so that the three variants sound like the brand and not like three different ad networks.

The product marketer learns to judge a whole asset set at once. A launch needs thirty pieces and they have only ever produced the narrative and the one-pager. What they learn is consistency across formats: the email, the post and the landing page saying the same thing in the right shape for each.

The ops person learns to judge copy. Their advantage is that they know what the sequence is for and when each message lands. What they learn is that “correct” and “good” are different, and what the difference looks like in a nurture.

The founder doing marketing learns to brief in thirty minutes and approve on a phone. Their advantage is conviction and domain knowledge. What they learn is to stop rewriting every draft and start correcting the standard that produced it.

The standard that makes it safe

The objection every CMO hears in week three is “if everyone makes their own images and writes their own copy, the brand will drift”. It is a fair objection, and the answer is not to keep the specialists as gatekeepers. It is to make the specialists the authors of the standard.

The brand standard is three documents and they are short. The voice: how we sound, what we never say, the claims we can substantiate. The look: colours, type, image style, what an on-brand image never contains. The rules: what gets approved by whom, what goes to a specialist review, what never sends without a person. Drafts are made inside the standard, so a marketer working outside their specialty cannot easily produce something off-brand, and the specialist’s expertise reaches every piece without their hands touching every piece. We covered the look side in visual brand consistency at scale and the copy side in off-brand content despite guidelines.

A one-page AI policy sits alongside it: what goes into the tools, what comes out, what context they must use and who approves. The template is a one-pager for a reason.

What the CMO is actually buying

Reframe the training line. The 3.8 percent is spent on sessions that teach the perishable part of the skill to people who are already using the tools. The alternative costs nothing extra in budget and a quarter of the CMO’s attention: a standard, a gate, and the instruction that each marketer carries whole pieces from now on.

What comes back is a team where each marketer’s expertise sets what they can do, rather than the specialists they can borrow. The writer who ships images. The strategist who ships finished content. The ops person whose nurture reads like the brand. And the handoff count, which is the number that shows it happened, falls from ten to two. The proof points are in the productivity metrics a CEO should read.

Where Marqeable fits

Marqeable is where the apprenticeship happens. A marketer writes the brief. Agents draft every piece from it, copy and images, in the team’s voice and brand look. The marketer reviews, adjusts and learns what good looks like in the discipline they never had. One person approves, and the same workspace sends the email and the text, publishes the post, runs the follow-up and traces revenue to the exact message. The skill grows inside the work, on real pieces, every week.

Frequently asked questions

How do you upskill a marketing team on AI?

Inside the work, not in a classroom. Give each marketer a piece to carry end to end, with agents drafting the parts outside their specialty (images for the writer, copy for the ops person, the full asset set for the product marketer), and have them brief, review and approve. The skill that grows is judgment about what good looks like in a discipline they could not previously produce in, and it grows in weeks because every piece is a repetition.

Why do AI training courses for marketers not work?

A course teaches prompting and tool features, which are the parts that change every quarter. It does not teach the marketer to evaluate work in a discipline they do not have, and it happens away from the brand rules, the review gate and the real pieces. Surveys in 2026 find most marketers already use AI weekly with no training at all; what they lack is a process that puts the tool inside the work.

What skills should a marketer learn in 2026?

Briefing (stating intent, audience, claims and constraints clearly enough that a draft comes back usable), evaluation (judging copy, images and sequences in disciplines beyond their own against a written standard), and execution (putting an approved piece into motion: publishing, messaging, follow-up). Prompting is the least durable of the skills because the tools keep changing.

How long does it take to upskill a marketer with AI?

About a quarter to see a measurable change, if the marketer ships real pieces every week rather than attending sessions. Expect quality complaints in the first three weeks, a turn around week five once the brand standard is doing its job, and by week twelve a marketer who carries whole pieces in a discipline they could not produce in before.

The bottom line

Three in four marketers use AI weekly, fewer than a third have been trained, and the training budget is shrinking. The course is the wrong unit anyway: it teaches the perishable part. The marketer you already have gets upskilled by carrying whole pieces, with agents drafting the half they never had, inside a written brand standard, with one approval gate and a specialist reviewing by exception. Twelve weeks, two pieces a week, and the handoff count tells you whether it worked. That is how a team does more of the work itself, and how it generates more leads and wins more of them as customers.

See each station of the factory: build campaigns from one brief, review and approve content, run follow-up with automations, answer visitors with website chat, and tie revenue to the exact message with attribution.

Marqeable runs your campaigns, answers every visitor, text, and email in seconds, and turns them into booked jobs and meetings - even at 9pm on a Saturday. We’re in private beta with a small early cohort. Get early access