title: "For the Head of Strategy"
slug: for-strategy-leaders
awareness: solution-aware
internal: false
description: "Written for the Head of Strategy or Creative Director who veto-gates the decision. Calibration methodology, framework alignment, named objections via Feel-Felt-Found, and a one-page Head of Strategy Brief PDF."
published: "2026-04-24"
audience: elena
register: strategist-intellectual
For the Head of Strategy
An intelligence layer that amplifies judgement -- calibrated to your data, not the internet.
Runs a pre-flight check on the brief before the creative kickoff. Scores the audience-sequencing call before the media spend commits. Does not replace your craft -- makes it legible in the rooms where craft currently gets lost to reassembly work.
Another brief where the AI hallucinated the audience
You are reading the fourth brief this week where the generated audience description is plausible, specific, and wrong. The category entry points are not what the client actually competes on. Your job has become the edit layer for a draft that should have been right before it reached you.
Your juniors are shipping faster than they can think
The draft leaves their desk in an hour. The reasoning underneath it would take them a day to defend. When the CMO pushes back on the audience logic on the status call, the junior cannot hold the room -- because she did not build the logic. The twin should have forced her to.
You are the bottleneck on every strategic call
Seven accounts, seven Monday status meetings, one of you. The calibrated judgement that moves a retainer from mediocre to compounding is a single-threaded resource, and your calendar is the constraint. The patchwork of dashboards and Slack channels holds the agency upright on Tuesday and collapses again on Monday.
The calibration methodology
Every senior strategist we have spoken to during research asked some
version of the same question inside the first ten minutes: how does
your system handle the thing that makes my agency mine? The answer
runs in a specific order. We teach the twin your agency's decisions
first, your clients' data second, and the public marketing-science
frameworks third. The order matters. A twin taught the frameworks
first and your decisions later reproduces the generic-AI failure mode
at a higher price point -- a confident-sounding recommendation built
on a foundation that has never met your clients, your category, or
the specific way your strategists write briefs.
The decisions-first order is load-bearing. Your agency's institutional
memory lives in the rejected drafts, the rewritten briefs, the
objections from the CMO on Tuesday's status call that turned into a
creative pivot on Thursday. Those decisions -- especially the ones
that overrode the obvious move -- are the highest-density calibration
signal available to any intelligence layer running inside your walls.
Generic AI never sees them. A twin calibrated to your agency is built
to learn from them first and treat the public frameworks as the
second-layer floor, not the first-layer source.
Calibration is the mechanism. Not scale, not speed. A twin calibrated
to one agency's decisions outperforms a larger model running cold on
the question that actually matters -- "what should we do on this
account this week?" -- every time. The depth of that answer sits on
the dedicated /mechanism/calibration page
for readers who want the operational walkthrough, including the
week-one to week-twelve calibration schedule, the data sources the
twin learns from, and the refusal lines that govern what the twin
will not touch.
The one-page Head of Strategy Brief (PDF) is the shortest honest
version of this page. It carries the methodology diagram, three
explicit "what this does NOT do" non-claims, and three questions
for you to bring to your next Monday strategy review. The component
below carries a direct download link; the brief is designed to be
forwarded to the CMO or Creative Director who needs a one-page
artefact, not a read-through.
The calibration methodology
The calibration loop runs inside one agency's walls. The orange path is where compounding happens; the white path is the read-only reference layer.
Worked example: a creative-fatigue call on a live account
A performance agency's strategist is running a Q2 creative-fatigue audit on a DTC skincare account. Six ad variants, four weeks of spend, CTR is trending down but not collapsing. The conventional move is to pause the bottom two variants and scale the top two; the less obvious call is that the top variant is still winning on CTR but losing on distinctiveness (the creative looks interchangeable with three competitor brands in the same feed). The twin flags the distinctiveness drift, surfaces the last three campaigns where this agency made the same call and regretted it, and scores the scale-the-winner move against the category's mental-availability history. The strategist still makes the call. She makes it with the twin's annotation open on her second monitor instead of from memory.
The operational depth of this loop -- what the twin learns from, what it refuses to touch, how the weekly calibration ritual runs -- sits on the dedicated spoke page.
We are not claiming to hold new science. The public marketing-science
canon is the floor the twin stands on, and we stay explicit about
distance from it: we read these authors, we apply the thinking, we
claim no endorsement or affiliation. This is the inoculating sentence
Elena needs before she accepts any framework-adjacent claim -- the
absence of it is a Stage 3 tell for audiences that have watched too
many vendors strip-mine academic reputation.
What the twin actually does with the frameworks, at the vocabulary
layer Elena will scrutinise first:
GET / TO / BY -- Julian Cole's strategic brief template. The
twin runs a pre-flight on the brief's GET/TO/BY stack before the
creative kickoff: who do we GET, TO do WHAT, BY HOW. The twin
catches the common failure mode where GET and BY are specific but
TO is fuzzy -- the shape most briefs ship with when the strategist
is tired.
CEP (Category Entry Points) -- Jenni Romaniuk, building on
Byron Sharp. The twin audits whether the campaign's creative
actually attaches distinctive assets to the CEPs that drive
category demand, not to the CEPs that flatter the client's
self-image. This is the most frequent structural failure mode in
the performance-marketing brief we have reviewed in research.
JTBD (Jobs To Be Done) -- Clayton Christensen, adopted into
planning frameworks. The twin cross-checks whether the audience
segment's job-to-be-done matches the creative's promise-stack, or
whether the creative is actually addressing a different job the
brand wishes the customer had.
SMP (Strategic Messaging Platform) -- the marketing-planning
artefact that rolls GET/TO/BY, CEPs, and JTBD into a single
messaging frame. The twin treats the SMP as the source-of-truth
document and flags every downstream creative asset that drifts
from it.
Mental availability -- Byron Sharp, How Brands Grow. The
twin's scoring function on creative variants weights mental-
availability gains (distinctiveness, CEP-linkage, category
fluency) above short-term CTR lifts. This is where the twin
diverges hardest from generic AI creative generators -- most of
them over-index on click-through and ignore the distinctiveness
cost.
For premium-category clients the scoring shifts. Ana Andjelic's work
on aspirational categories (The Business of Aspiration) is the
reason a twin calibrated for a luxury or premium-positioned client
weights cultural capital and symbolic-distinctiveness signals over
pure performance framing -- the generic performance-first calibration
undershoots in those categories, reliably.
Framework alignment
Byron Sharp
mental availability / CEP (Category Entry Points)
How Brands Grow (Parts 1 and 2)
The twin scores creative variants on mental-availability gains (distinctiveness + CEP-linkage + category fluency) above short-term CTR lifts. Where generic AI generators over-index on click-through, the twin treats the distinctiveness cost as a first-class constraint.
Jenni Romaniuk
distinctiveness audit / brand health
Better Brand Health Tracking / Building Distinctive Brand Assets
Romaniuk's operational reading of Sharp is the reason most brand trackers measure what they are not actually measuring. The twin runs a distinctiveness audit on every creative asset a client ships and flags drift against the category's established distinctive-asset inventory.
Julian Cole
GET / TO / BY / SMP (Strategic Messaging Platform)
Strategy Finishing School
Cole's GET/TO/BY template is the brief pre-flight. The twin catches the common failure mode where GET and BY are specific but TO is fuzzy -- the shape most briefs ship with when the strategist is tired. The SMP rolls GET/TO/BY + CEPs + JTBD into the source-of-truth document the twin treats as the single reference.
Ana Andjelic
aspirational category scoring / cultural capital
The Business of Aspiration
Andjelic's work on aspirational categories is the reason a twin calibrated for a luxury or premium-positioned client weights cultural-capital and symbolic-distinctiveness signals over pure performance framing. A generic performance-first calibration undershoots in those categories, reliably.
Named objections
Four named objections, reframed via Feel-Felt-Found
If the twin writes the brief, what is left of craft?
Feel
The work that made you a senior strategist -- synthesis across accounts, taste under pressure, the judgement that says "this campaign has a mis-shaped audience before we spend a dollar" -- is exactly the work every AI vendor now promises to automate. The fear is not that the tool is bad. The fear is that the tool is good enough to make craft invisible to the person signing the invoice.
Felt
Senior strategists in the 2024-2026 consolidation cycle have felt this version of the fear twice: once during the agency-holding-company centralisation wave, and again when the first generation of generic AI decks started circulating inside client organisations. The through-line in both cases is the same -- the work got commoditised not because it stopped mattering, but because the layer above it stopped being able to tell the difference between calibrated thinking and a plausible draft.
Found
Agencies running a calibrated twin for 30+ days report that strategy work moves up the ladder, not out the door. The twin handles the pre-flight check on the brief -- does the GET/TO/BY stack match the category entry points, does the mental availability argument survive contact with the distinctiveness audit, does the audience sequencing actually match the commercial model. The senior strategist still decides what the agency stands for, still runs the room with the CMO, still owns the retainer. What changes is that the hours between her and the judgement call shrink. The twin does not replace her. It makes her hour more expensive, because her hour is no longer spent reassembling context.
If juniors ship from the twin, when do they learn to think?
Feel
You remember how you actually got good at this. It was not the framework. It was the 200 hours of sitting with a brief that was almost right and figuring out why it was not. If your juniors ship the twin’s first draft unchallenged, they never do the reps. Four years from now, the agency has a hole where its bench should be, and there is no calibrated senior strategist available to hire because the market is drained.
Felt
Creative directors and heads of strategy at agencies that adopted generative AI for junior-level copy in 2023-2024 have watched exactly this pattern play out in slow motion. The juniors got faster at output and slower at judgement. The rework cycles did not disappear -- they just moved from the junior’s desk to the senior’s desk, and the senior stopped having time to teach because she was now the edit layer.
Found
A calibrated twin forces the junior to show her work in a way ChatGPT never does. Because every suggestion is annotated with which client signal, which calibration source, and which framework assumption it came from, the junior has to argue with the twin -- not just accept its output. The feedback loop the twin runs with her becomes its own apprenticeship. We do not claim this replaces mentorship. It does mean the twin refuses to let a junior ship something she cannot defend, which is the part of mentorship a senior strategist actually wants to protect.
Does every agency running this sound the same in 18 months?
Feel
The thing that makes your agency worth hiring is the voice. The angle. The in-house way you frame a problem that nobody else frames the same way. If the twin is trained on generic marketing science and then bolted onto every agency that pays for it, the distinctive edge flattens -- and you are back to competing on price and speed, which is the exact fight you did not get into this business to have.
Felt
Agency leaders who watched the last two waves of marketing-automation adoption (marketing automation in the early 2010s, then the first CDP wave) have lived this specific fear. Both waves ended with category-wide sameness at the operational layer and premium pricing only for the shops that held a line on craft at the strategic layer.
Found
The twin is calibrated per agency. One agency, one twin. It does not share inference, embeddings, or decisions across instances -- architecturally, not as policy. The reason this matters for your voice: the twin learns what your strategists reject, not just what they ship. Over 6-12 months the twin’s recommendations bend toward your house angle, not away from it. The Portability Covenant enforces this at the data layer (your data does not help anyone else, ever), but the cultural version of the promise is simpler: you cannot flatten an agency’s voice with a system that only ever sees that agency’s work.
I am not against this. I am just tired. Is that enough to block the deal?
Feel
You are not anti-AI. You have seen the live demos. You have read the decks. You know the calibrated version is different from the generic version. What you do not have is the appetite for another 90-day rollout where the promise front-loads and the operational cost back-loads onto your desk. The honest blocker is not disagreement. It is the quiet math of capacity you already do not have.
Felt
Heads of strategy in the 2025-2026 window have been pitched a version of "AI will make you faster" roughly once a fortnight for two years. The ambivalence is not laziness and it is not technophobia. It is calibrated exhaustion -- a rational response to a pattern where every previous tool promised to remove work and added a thin layer of it instead.
Found
The Parallax Test is designed around this exact posture. It is a 30-day, no-charge pre-flight on one live account. The work the twin does during the test is work your strategist would have done anyway -- brief pre-flight, creative-fatigue audit, audience-sequencing check -- not net-new work added to her plate. If the twin does not pull its weight inside the first 30 days, the test ends, nothing ships, nothing onboards. The ambivalence is the correct starting posture. The Parallax Test is how we earn the right to move you off it.
Further reading
The twin's framework floor is public work. The four books and
bodies-of-work below are the ones we refer new strategist hires to
when they want to get serious about the discipline itself. We claim
no endorsement or affiliation with the authors or their firms.
Further reading
JR
Better Brand Health Tracking
Jenni Romaniuk
The operational reading of Sharp. If you are going to argue about what your brand-health tracker is actually measuring, this is the floor.
BS
How Brands Grow (Parts 1 and 2)
Byron Sharp
Mental availability, distinctiveness, category entry points. The structural floor for every creative-fatigue and audience-sequencing call the twin runs.
AA
The Business of Aspiration
Ana Andjelic
The premium / aspirational-category reading -- for clients where pure performance framing undershoots and the brand architecture carries the weight.
JC
Strategy Finishing School
Julian Cole
Craft depth for strategists. The one body of contemporary work we refer senior hires to when they want to get serious about the discipline itself.
Notes from the builder
Notes from the builder
The honest version: the person writing this page is a builder-researcher, not a veteran agency operator. The marketing-science floor the twin stands on comes from the public work of Sharp, Romaniuk, Cole, and Andjelic -- read in order, applied deliberately, with no endorsement or affiliation claimed. The operational shape of the product was pulled out of six months of research with senior strategists at mid-market performance agencies. What is shipping here is a careful synthesis, not a war-story.
That posture is deliberate. A page on /founder that claimed nine figures of spend under management or a decade inside the big-five holding companies would be easier to believe at a glance and harder to trust at a second pass. A senior strategist reading the framework-alignment section above will catch a mis-used CEP or a mis-characterised mental- availability claim faster than any credential signal will repair it. So the credential signal stays honest, and the framework vocabulary gets the care.
What the twin produces on a live account is documented in the Living Brief -- the weekly artefact that captures the twin's pre-flight check, annotated recommendations, and the refusal lines it declined to cross. A representative sample of the Living Brief is available as a downloadable PDF from the /approach page; the Head of Strategy Brief (linked above from Section 3) is the shorter one-page companion sized for a Monday strategy review rather than a read-through.
If you would rather see the product run against a live account before you commit judgement, that is what the Stump Session and the Parallax Test are for. The builder- researcher frame is not a CV gap -- it is the signal that the person behind the product is spending the early time on calibration instead of on a press cycle.
Book the Stump Session
The Stump Session is a 45-minute call in which you bring the
hardest strategic question you have in front of you this week --
the account, the category, the creative fatigue call, the
audience-sequencing call you have not committed to yet -- and we
work it live against the twin's calibration loop. It is not a sales
demo. It is the room where you test whether the twin holds up under
the specific version of "hard" that lives inside your agency.
If the session is useful, the next step is the Parallax Test: a
30-day, no-charge pre-flight on one live account, structured so the
twin only does work your strategist would have done anyway -- brief
pre-flight, creative-fatigue audit, audience-sequencing check -- and
nothing net-new gets added to her plate. Nothing ships, nothing
onboards, nothing converts into a contract without a second explicit
decision at the end of the 30 days.
Your best strategist belongs in the Stump Session. If she is not in
the room, the session does not earn the right to ask you for the
Parallax Test.
If you are forwarding this page from a peer inside the agency, the
Head of Strategy Brief PDF linked above is the one-page companion
to read before the call so the 45 minutes are spent on the specific
call your agency needs to make this quarter, not on explaining what
the twin is.