How I think
about marketing
now.
Two frameworks. Neither was handed to me — each stage below is attached to the week that produced it, because a loop I cannot source is a loop I did not learn.
The marketing loop
In Week 1 my version of this had three steps and started at “create”. Everything before and after it turned out to be where the decisions live.
Understand
Who is the customer, and what occasion are they choosing for?
Day 1, visiting Chicco, Il Toro, Kulto and Cenzo in a single day. Four restaurants in one group, four different people walking through the door for four different reasons.
Define
What are we actually trying to change?
Week 4, when the most polished videos on Chicco's TikTok were performing worst. Without a defined job, content just gets prettier.
Position
What should this brand communicate, and what should it never say?
Day 4, writing captions for several venues in one sitting and realising they all sounded like the same restaurant.
Create
What should we make, and how should it be made?
Day 3 and Week 2, discovering that bright light makes a dish read casual and warm light makes the same dish read premium. A lamp is a positioning decision.
Distribute
Where will the customer meet it, and who is standing there?
Week 2 in Meta Ads Manager, filtering customer profiles into audiences by restaurant and occasion, and Day 4 scheduling in Metricool.
Experience
What happens when they act on it?
Day 2, seeing BistroChat behind the public pages. Discovery, booking and follow-up are one path, and the weakest step decides the outcome.
Retain
Why would they come back — and to which venue?
Week 2's Avocado call and the Week 7 loyalty mock-up, where rewards had to differ per brand to avoid discounting a premium venue's positioning.
Measure
Did it do the job we defined?
Week 4, recommending that Chicco's TikTok be judged on profile visits, saves, comments and follows rather than views.
Improve
What does the result change about the next attempt?
Week 6, revising the TikTok strategy and Events Toolkit on feedback. The first version's job is to reveal what needs to change.
Marketing is a loop.
Not a post.
My decision
threshold
My supervisor noted that I can overanalyse, which occasionally slowed decisions. This is my correction written as a process — a development goal, not a habit I have already mastered.
One reason I overanalyse is treating every decision as if it needs complete certainty. Most do not. The distinction I am learning to make first is not how important a decision feels, but whether it can be undone.
Move, then review.
Drafts, layouts, a first version of a content series, an editing choice. The cost of being wrong is a revision. Deliberating here is expensive and buys nothing.
Slow down deliberately.
Anything that sets a pattern others will reuse, changes what a brand is understood to be worth, or commits the business financially — a loyalty structure being the clearest example.
What information is actually necessary?
Not everything knowable — only what the decision depends on. Listing it is usually shorter than expected.
What would change my decision?
If no available information would change the answer, the analysis is finished whether it feels finished or not.
Do I have enough to act?
Week 7 taught me this the hard way: there was no transaction data coming, so waiting for it was not caution, it was avoidance.
If it is reversible, move.
Most work is reversible — a draft, a layout, a first version of a series. Reversible decisions should be made fast and reviewed, not debated.
Review the result and adjust.
Was the reasoning right, or only the outcome? Both are worth knowing, and only one of them repeats.
Think enough to make a good decision. Not so long that no decision gets made.
The rules I now
work by.
Each one is a standing decision rule, paired with the test that would prove or break it in the next role.
Name the task reversible or high-consequence before deciding how much thinking it deserves.
Classify a week of tasks in advance, then review which classifications were wrong and what the misjudgement cost.
Nothing is produced for a brand until its audience, occasion and tone are written down.
Hand unlabelled captions to a colleague and ask them to match each to a venue; anything below full accuracy means the positioning is not sharp enough.
Show structure at twenty per cent complete, when redirection costs minutes rather than days.
Share an outline or wireframe early on the next three deliverables and log how much rework each one avoids.
Define what would count as success before the work runs, not after the numbers arrive.
Attach one primary action — booking, save, enquiry — to each campaign and review it at a fixed date instead of on instinct.