Field notes.
Eleven conclusions I am carrying forward. Each one opens onto the experience that caused it — because a principle without its origin is just a good-sounding sentence.
Marketing is a system.
What looked like separate tasks turned out to be one customer journey wearing different job titles.
Day 2, learning BistroChat Manager and then reviewing the Chicco, Casa Mori and Il Toro websites. The reservation layer, the WhatsApp messages, the online menu and the follow-ups were all the same conversation at different moments.
That marketing was the visible output — the posts, the photographs, the campaigns — and everything else was administration.
The output is the smallest part. A strong post attached to an unclear website and an awkward booking step still loses the customer, and the failure will get blamed on the post.
It changes where you look for a problem. Weak results are often a broken handover between touchpoints rather than weak creative.
Before proposing content, I check the whole path a customer takes to act on it, and fix the weakest step first even when it is not the interesting one.
Every brand needsa different answer.
Context beats copying tactics — including copying from the restaurant next door with the same owner.
Week 3, editing for several venues, and Week 5, building an events kit for Humo, Chicco and Il Toro in the same document.
That once I found the approach that worked best, the sensible thing was to reuse it. Consistency looked like quality control.
Consistency means a shared standard, not identical execution. The same edit style across four brands makes four concepts sound like one, which removes the reason to choose between them.
A group's venues compete for the same customer's occasion. If the marketing blurs them, the group is paying to make its own choices harder.
I write down audience, occasion and tone per brand before producing anything, and treat any reuse as a decision I have to justify.
Preparationcreates flexibility.
Preparation is not the opposite of spontaneity. It is what makes spontaneity affordable.
Week 2, on the Casa Mori menu shoot, where a dish only stays presentable for a short window and the camera has to already be right when it arrives.
That preparation was mostly about avoiding mistakes — a safety measure for people who are not quick on their feet.
Preparation buys options. Testing settings on an empty plate converts the most valuable minutes of a shoot from setup into filming, which is where the unplanned good shots come from.
In hospitality the operational window is the real constraint. A kitchen cannot pause service while marketing gets ready.
I set up against the constraint rather than against the plan: what is the moment I cannot repeat, and am I ready before it arrives?
Creativity needsan objective.
I became less interested in whether something looked good and more interested in what it was supposed to change.
Week 4, reviewing Chicco's TikTok. Human-led content engaged best while the most polished dish showcases performed weakest.
That better production would produce better results, so the way to improve an account was to raise the quality of the visuals.
Quality only compounds when the content has a job. A beautiful video with no hook, no person and no reason to save it is a well-made thing nobody needed.
Production time is the scarcest resource a small marketing team has. Spending it on polish rather than purpose is expensive in a way that does not show up anywhere.
I decide what a piece is supposed to move — a save, a booking, a first-time viewer understanding the concept — before deciding how it should look.
The fastest solutionis not alwaysthe best solution.
Efficiency is valuable. Depth is valuable. The skill is knowing which one the task in front of you deserves.
Week 6, when I realised I had assumed what was required on parts of a project instead of checking, and the work had to be redone.
That speed was straightforwardly a strength, and that asking mid-task would suggest I had not understood the brief.
Speed compounds direction. Applied to a misread brief it produces more wrong work, faster — which is precisely what happened.
Rework is the most expensive thing a small team absorbs, and it is nearly always cheaper to prevent at the start than to fix at the end.
I confirm my understanding in writing before starting, and I sort tasks by whether the decision is reversible before deciding how fast to move.
Thinking helps.Overthinking doesn't.
My supervisor identified that I can overanalyse. I want to get better at noticing when more thinking has stopped adding anything.
Week 7, setting reward costs and tier thresholds for the loyalty proposal with almost no transaction data to justify them.
That the responsible thing was to keep refining until the numbers could be defended properly.
There was no amount of analysis that would produce that data. The right move was a first version explicitly designed to be adjusted once real behaviour existed, and to say so in the proposal.
Businesses launch with incomplete information constantly. A decision framed as a test is more useful than a decision delayed until certainty that is not coming.
I ask what information would actually change my decision. If nothing available would change it, the analysis is finished and I should be acting.
Confidence isa professional skill.
My supervisor encouraged me to trust my judgement and take on more challenging work. That is the honest edge of my development.
Across Weeks 5 to 7, in the recommendations I made on positioning, loyalty and social strategy — and in how much I hedged them.
That certainty had to be earned before an opinion could be stated, and that stating one early was overstepping as an intern.
A recommendation with the reasoning shown is more useful than a balanced summary with no position in it. The team can disagree with a stated view; they cannot do anything with a hedge.
The value of an outside perspective is the perspective. Withholding it to stay safe removes the only thing a new person can uniquely offer.
I state a recommendation and the reasoning behind it, name my confidence level honestly, and volunteer for the work where the answer is not already obvious to me.
Executionchanges the idea.
A concept still has to survive contact with everything that comes after it.
Week 6, filming the TikTok storyboard reenactments. Concepts that read well as frames fell apart on pacing and angle once they were actually shot.
That the hard part of creative work was having the idea, and that execution was the process of carrying it out faithfully.
Execution is where an idea is tested. Building the rough version is the cheapest way to find out whether a concept was ever real.
Time spent developing an idea that cannot be produced is time nobody gets back — and the failure only becomes visible at the most expensive moment.
I build the rough version early, and design ideas around what can realistically be produced during a normal week of service.
School taught the framework.Work taught the constraints.
The theory was not wrong. It was simply missing everything that surrounds a decision in practice.
Week 5 and Week 6, on the events kit and the Events Toolkit website — where a good structure still had to clear feedback, branding, operational reality and someone else's deadline.
That marketing decisions were made by choosing the strongest option from a set of alternatives.
The strongest option is only a candidate. What ships is the strongest option that fits the time, resources, approvals, brand rules and the way the venue actually operates on a Friday night.
A recommendation that ignores operational feasibility is not a recommendation, it is an opinion. Any promotion still has to work for the kitchen running it.
I ask early who has to approve this, who has to produce it, and what it costs the venue operationally — and treat those as design inputs rather than obstacles.
The customer doesn'tsee departments.
Search, website, WhatsApp, the room, the menu, the service, the loyalty card — the customer experiences one restaurant.
Day 1 and Day 2 back to back: the physical journey through four venues, then the digital layer sitting behind them in BistroChat, the websites and the reservation flow.
That these were separate specialisms — content, web, CRM, operations — that happened to serve the same business.
They are one perception, assembled by a person who has no idea which team produced which part. An inconsistency between the Instagram and the room is not a coordination problem, it is a broken promise.
This is why positioning has to hold across every touchpoint: the weakest one sets the customer's belief about the brand, not the best one.
I check any new piece against the touchpoints on either side of it — what did the customer just see, and what happens immediately after this?
An idea only creates valuewhen it survives execution.
The difference between a concept and a contribution is everything that happens in between.
Week 7's loyalty proposal, which only became real once I built the mock membership screens, and Week 6's Events Toolkit, which only became useful once it existed as a draft people could react to.
That a well-argued proposal was the deliverable.
A description makes people imagine; a rough version makes them respond. Designing the actual screens forced decisions the concept had allowed me to leave open — what status means, how a reward is redeemed, how a casual and a premium venue share one account.
Feedback on something concrete is specific and actionable. Feedback on an abstraction is agreement, which teaches nobody anything.
I build the roughest possible working version early, and treat it as a thinking tool rather than a presentation.