Strategy guide

What a creative strategist does: the job, the week and the skills

What a creative strategist owns, a week in the job, a worked example from customer reviews to a brief, how to read results, the skills to hire for and what AI changes.

Guide · No. 22

Reading time
11 minutes
Updated
Topic
Strategy

The short version

  • A creative strategist decides what the next round of ads should say and why, then reads the results to plan the round after.
  • The job is research, briefs and learning. Design and media buying are separate jobs the strategist works closely with.
  • Good strategists work from customer words, rivals' long-running ads and the claims a product can prove, never from taste alone.
  • Hire for judgment you can see: a paid take-home brief built from real reviews tells you more than any certificate.
  • AI makes ads cheap to produce, which moves the bottleneck to choosing which ideas deserve a test.

The short answer

A creative strategist decides which ads a brand should make next, and why. They turn customer research, competitor patterns and past results into briefs, then read what the ads did to plan the round after. They sit between whoever runs the spend and whoever makes the ads, and what they produce is a steady flow of tested ideas rather than any single ad.

In paid social teams the title turns up under several names: performance creative strategist, creative lead, growth creative. The work underneath is the same. This guide covers what the role owns, a typical week, a worked example from customer reviews to a finished brief, how to read results, the skills to hire for and what AI changes. For the one-paragraph version, see the creative strategist glossary entry.

What the role owns, and what it shares

Area The creative strategist Shared with
Customer research Owns it: reviews, comments, support emails, survey answers Customer service, the founder
The angle backlog Owns the ranked list of ideas waiting for a test Anyone can add to it
Briefs Writes one per concept Designers, editors, creators, AI tools
What goes into each round Chooses the concepts and hooks The media buyer, who sets spend and structure
Production Reviews and approves, rarely makes Designers and editors
Results Reads them by concept and keeps the learnings log The media buyer or an analyst

Spend, bids and audience settings belong to the media buyer, and this guide leaves them there. The overlap that matters is the test plan: the strategist says which ideas need an answer, and the buyer says what it costs to get one.

In a small team one person does all of this, often the founder. Naming the role still helps, because it makes sure someone does the thinking part on purpose rather than whenever there's a spare hour.

A week in the job

Most strategists work in a weekly loop. The days move around, but the order mostly doesn't:

Day Main work Done by the end of the day
Monday Read last week's results by concept, update the log Three to five lessons written down
Tuesday Research: new reviews, comments on live ads, support tickets, rivals' long-running ads Two or three new angles in the backlog
Wednesday Write briefs, agree the round with the media buyer Briefs for this round's concepts
Thursday Review first drafts, fix claims, cut near-duplicates An approved set
Friday Names, launch prep, a look at early signals (not verdicts) The next round queued and paused

How many ads go into each round depends on spend, and how many ads per week covers that. A creative sprint is a good way to run the loop faster when a launch or a season needs more in less time.

Where good ideas come from

Strategists who last work from evidence first and opinions last. Four sources, roughly in order of how far to trust them:

  1. Customer words. Reviews, post-purchase surveys, support emails and the comments under your own ads. Customers describe their problem in words a brand would never write, and those words make the strongest hooks. Tag what people praise and what they complain about, then count the tags. See pain point.
  2. Rivals' ads that keep running. Meta's public Ad Library (opens in a new tab) shows active ads with the date each one started running, and its API reference (opens in a new tab) lists that delivery start date as a field (checked October 2026). An ad still live after months is more likely to be paying its way than one launched last week. Note the pattern, meaning the hook, the layout and the offer, and never copy the ad. BulkAds' competitor research sorts rivals' ads by how long they've run.
  3. What the product can prove. The claims on your pack, product page and certificates. A clever idea built on a claim you can't back up is not a usable idea. Keep approved claims in a brand kit so every brief and tool reads the same list.
  4. Your own test history. The learnings log. An angle that lost twice with different executions gets parked, and one that won gets new executions.

Trends and "what everyone's doing" come last. They help with formats and rarely with reasons to buy.

Worked example: from reviews to a brief

Hushwren is a fictional brand selling reusable silicone sleep earplugs: one pair with four sizes of ear tip, in a small case, for $15. The review counts and results below are made up for this example.

Research. The strategist reads Hushwren's 200 most recent reviews and tags each one by what the reviewer talks about:

Tag Reviews (illustrative) What reviewers describe
A partner's snoring 71 Sleeping through the night next to someone who snores
Side sleeping 44 Earplugs that don't press when they lie on one side
The first few nights 29 Feeling odd at first, then not noticing them at all
Price next to foam 18 Paying more than for a pack of disposable foam plugs

Insight. Happy buyers barely describe the earplugs. They describe the person next to them and their first full night. The "first few nights" tag is the most useful of the four, even though it's third by count, because it reads like the moment some buyers give up. An ad that answers it before purchase could sell to people the snoring ads only interest.

Angles. Each tag becomes a concept with a first hook to test:

Concept Built from First hook
The snorer on the next pillow Snoring reviews He still snores. I just don't hear it.
Side-sleeper fit Side-sleeping reviews Made for the side you actually sleep on.
Night one vs night four The first-few-nights tag Night one felt odd. By night four I forgot they were in.
One pair, not a drawer of foam The price tag One pair in a case, instead of a drawer of foam ones.

Claims check. Three lines from the first draft were cut or rewritten:

  • "Blocks 30 dB of noise." A decibel figure is a lab test result. The brief allows only the rating printed on Hushwren's own pack, next to the standard it was tested against, and nothing else.
  • "Ends sleepless nights." That reads as a promise to fix a sleep problem, which an earplug can't back up. Out.
  • "Can't sleep?" as an opener. As of October 2026, Meta's personal attributes policy (opens in a new tab) says ads must not assert or imply personal attributes, health conditions included, and lists "Do you have diabetes?" among its not-allowed examples. A hook that tells the viewer they have a sleep problem sits too close to that line, so every hook in the round speaks in the first person or about the product.

The brief for the concept with the most to prove:

Concept:     Night one vs night four
Insight:     the first few nights decide whether a buyer keeps them
Proof:       four tip sizes in the box, review lines quoted only with permission
Executions:  A) bedside table, open case, "night 1" and "night 4" on two sticky notes
             B) a wall calendar, nights one to seven, a tick on each
             C) phone shot of a pillow in morning light, case beside it, nobody in frame
Hooks:       "Night one felt odd. By night four I forgot they were in."
             "Give them four nights."
             "Four tip sizes, because ears aren't one size"
Must not:    claim to treat any sleep condition, use a decibel figure that isn't on the pack,
             tell the viewer they have a sleep problem
Judge on:    purchases by concept, and early returns if the store can report them

The prediction. Before launch, the strategist writes down two guesses: "the snorer" will win on clicks, because it came up most, and "night one vs night four" will win on purchases, because it answers the doubt that stops people buying. Two written guesses, one per measure, test the strategist's judgment as well as the ads.

For the brief itself, start from the creative brief template or the free creative brief generator.

How a strategist reads results

Ad-level winners are where reading starts, not where it ends. The questions that matter:

  • Which concept won, across all its executions? Add up purchases by concept. One great ad in a weak concept is a lucky image. Three good ads in one concept is a reason to buy.
  • Did it stop people, hold them, or sell? For video, hook rate and hold rate show where an ad loses people. Clicks say the promise worked. Purchases say the product kept it. A winner on clicks alone is a trap.
  • Did each ad get a fair chance? Spend is rarely spread evenly across the ads in a test, so an ad that barely spent is untested, not beaten.
  • Was the guess right? Compare the written prediction with the result. A strategist who is wrong in the same direction twice has learned something about their own taste.

All of it goes into the learnings log. An illustrative page for Hushwren:

Round Concept We expected What happened Lesson Next
9 Night one vs night four Most purchases Most purchases, fewest early returns Answering the doubt sells better than naming the problem Three new executions and a video version
9 The snorer on the next pillow Most clicks Most clicks, middling purchases The snorer draws people in but doesn't settle the doubt Retest with the four-nights line as the closer
9 Side-sleeper fit Third Weak on everything Fit may need to be seen, not read One more try as a close-up video, then park it
9 One pair, not a drawer of foam Low Small but steady Speaks to people already using foam plugs Keep it in the reserve

Both guesses held up in this round. When they don't, the log is where a strategist finds out which of their instincts to trust and which to check twice.

BulkAds' creative analytics groups results by angle, hook style, format and size, and its test ledger lists every Autopilot verdict in order. For ads launched through BulkAds that gives you the results half of a learnings log. The "lesson" column is still the strategist's job.

Skills to hire for

  • Finds the pattern in a pile of reviews and can show the count behind it
  • Writes hooks in customers' own words
  • Writes briefs a designer, editor or AI tool can follow without a call
  • Knows what feeds, Stories, Reels and carousels each need from a first frame
  • Reads a small results table without crowning winners on noise
  • Keeps claims honest and knows Meta's Advertising Standards (opens in a new tab)
  • Keeps a log, and opens it before every new round

The best test is a paid take-home task built from real material. Give a candidate 100 anonymised reviews and three long-running competitor ads, and ask for three concepts, one brief and nine hooks. Judge whether the concepts came from the reviews, whether they're genuinely different, whether the brief would survive without a call and whether every claim could be backed up.

Meta also runs a certification. As of October 2026, its certification page (opens in a new tab) lists the Meta Certified Creative Strategy Professional exam (300-101): 60 questions in 105 minutes and a passing score of 700 out of 1,000. Three sections carry 25.5% of the grade each (mobile-first fundamentals, research and insights, brief development and creative production strategy) and evaluation and measurement carries 23.5%. The certification is valid for 12 months from the date it's earned. It shows grounding in Meta's formats. A portfolio of briefs with their results shows judgment.

Founder, freelancer, agency or hire

Option Works best when The trade-off
The founder One product, early days Closest to customers, least time
A freelance strategist A steady account that needs outside eyes Needs real access to reviews and results to be useful
An agency You need production as well as strategy Briefs can drift towards the agency's house style
An in-house hire Enough rounds each month to fill a role One person's taste can become the brand's blind spot

Whichever you choose, the strategist needs the same three things: raw customer words, honest results and the authority to say no to ideas with no evidence behind them. The freelancers and agencies pages cover those set-ups, and the workflow for small teams covers the founder route.

What AI changes about the job

AI makes ads cheap and fast to produce, so the bottleneck moves to deciding which ideas deserve a test. A strategist who once got five concepts made a month can now brief ten and see first drafts the same day. That means more rounds and more learning, as long as the ideas are good.

It also adds three risks the strategist now owns:

  • Sameness. AI tools drift towards the same polished look, and near-identical ads teach little. Meta's December 2024 engineering post (opens in a new tab) describes a retrieval stage that picks a few thousand relevant ads from tens of millions of candidates, so variety has to be planned rather than hoped for. See concept vs iteration testing.
  • Invented claims. A model will happily write "clinically proven" if nobody stops it. Approved proof points have to live somewhere every tool reads from.
  • Unchecked output. Meta's own generative AI features (opens in a new tab) for text variations, image expansion and background generation come with a note, as of October 2026, that advertisers are responsible for previewing AI-generated creative before publishing. Someone has to look at every ad. Our guide to Meta's generative AI ad tools covers what each one does.

BulkAds is built for the production half of this: a brand kit read from your website that holds your voice and proof points, competitor patterns sorted by how long ads have run, batches of up to 12 ads per concept with a pre-flight check, and results grouped by angle. It doesn't interview your customers or decide what your brand stands for. That part stays with the strategist.

Mistakes new strategists make

  • Briefing scenes instead of reasons. "Beach, kitchen, office" is three executions of one concept, not three concepts.
  • Calling winners on two days of clicks. Wait for purchases.
  • Testing what the team likes. If no customer said it, it needs a better reason to be in the round.
  • Ten hooks on a weak idea. Fix the concept before polishing the opener.
  • Letting the log lapse. Without it, round twenty quietly repeats round six.

Questions people ask

What does a creative strategist do day to day?

They read last round's results by concept, research customers and rivals' long-running ads, write briefs for the next round, review drafts for claims and near-duplicates, and keep a log of what each round taught. Most work in a weekly loop.

Is a creative strategist the same as a media buyer?

No. The strategist decides what the ads say and which ideas need testing. The media buyer runs the spend and the account settings. They agree the test plan together, and in small teams one person often does both jobs.

Does a small brand need a creative strategist?

It needs the work done, not necessarily the job title. A founder who reads reviews every week, writes one brief per concept and keeps a learnings log is doing the role. A dedicated hire makes sense once there are enough rounds each month to fill it.

What skills does a creative strategist need?

Customer research, hooks written in customers' words, clear briefs, a working knowledge of Meta's formats, careful reading of small result tables and honest claims. A paid take-home task built from real reviews is the best way to check them.

Is there a certification for creative strategists?

Meta offers the Meta Certified Creative Strategy Professional exam. As of October 2026 its page lists 60 questions in 105 minutes, a passing score of 700 out of 1,000 and 12 months of validity. It's useful grounding, but a portfolio of briefs and results shows more about judgment.

Will AI replace creative strategists?

AI takes over much of the production, not the judgment. It can turn a brief into many ads quickly, but it can't decide which customer insight deserves a test or notice that a claim can't be proved. The role shifts towards research, briefs and review.

Keep going

Written by the BulkAds team. Published , last checked . Meta changes its tools often — if something here no longer matches Ads Manager, tell us at contact@bulkadsai.com.

BulkAds AI is an independent product and is not affiliated with, endorsed or sponsored by Meta Platforms, Inc. Facebook and Instagram are trademarks of Meta Platforms, Inc.

Your next fifty ads are one link away.

Paste your website, meet your brand kit and see your first batch in minutes. 7 days for $9.99 — cancel anytime.

yourbrand.comMake my ads

See pricingExplore featuresBrowse the library