Ask a class what they are doing for the final project and roughly two thirds will describe the same thing: a full marketing plan for a well known brand. Nike, Vodafone, a bank. It is a reasonable choice and it is almost useless as evidence, for one reason. Nobody can tell whether you were right.
A hiring manager reading a strategy for a company you have no access to cannot check any of it. What they can check is whether you did something real, measured it, and were honest about the result. That is the whole brief.
What the project is actually for
Employers use a portfolio piece to answer three questions that a certificate cannot: can this person define a problem, can they execute something end to end, and do they know whether it worked. A hypothetical plan answers none of them. A small real thing with a disappointing result answers all three.
It is worth knowing that the market side of this is shifting in your favour. Egypt's Labour Law No. 14 of 2025, published in the Official Gazette on 3 May 2025, requires employers to allocate 0.25% of payroll to staff training, alongside a minimum 3% annual salary increase. A firm that is now obliged to spend on developing people has a reason to hire someone who can already demonstrate a learning loop, and a project that shows you measuring your own work is exactly that demonstration.
Six projects that prove something
- A real campaign with a real budget, however small. Two thousand pounds spent on a genuine offer, with the targeting decisions, the creative iterations and the actual cost per result written up. Small budgets are not a weakness here. They force the decisions that large ones hide.
- A local business audit with a before and after. Pick a business that will let you touch something. Fix one measurable thing, the Google Business Profile, the tracking, the landing page, and record the number either side.
- A landing page test. Two versions, one variable, enough traffic to be worth reading. Include the version where you were wrong, because that is the part that reads as real.
- A content series with retention data. Eight pieces on one narrow topic, with the analytics on which ones held attention and your reading of why.
- A measurement rebuild. Take a business whose tracking is broken, which is most of them, and fix the events, the definitions and the reporting. Unglamorous and instantly credible, because everyone who has done the job knows how common it is.
- A competitor teardown with primary evidence. Not a summary of their website. Actual ad library screenshots over time, the offers they tested, what they stopped running and what that implies.
The structure that makes any of them legible
Whatever you pick, present it the same way: the question you were trying to answer, what you did, what happened, what you would do differently. Four sections. Most students invert this and lead with the activity, which buries the only part a reader cares about.
Include the numbers even when they are bad. A project reporting a cost per lead that was worse than the client's existing channel, with a clear analysis of why, is stronger than one reporting an unverifiable success. Anyone experienced enough to hire you has run campaigns that lost money and will trust you more for showing one.
Mistakes that cost marks and interviews
The commonest is scope. A project covering strategy, content, paid, SEO and email at surface depth demonstrates less than one covering a single channel properly. Depth is the signal.
The second is unverifiable claims. Percentages with no baseline, engagement numbers with no denominator, a rise you cannot attribute. If you cannot show where a number came from, remove it, because the reader who notices will discount everything else on the page.
The third is finishing on the deck. The project that gets you hired usually exists somewhere public, as a live page, a repository, a written case study, rather than as a PDF attached to an email.
Where to look for a subject
The best projects come from businesses in reach: a family business, a friend's shop, a charity, your current employer's neglected channel. Access beats prestige every time, and a small business that lets you touch the account is worth more than a famous brand that will never know you existed.
What you learn along the way maps onto the modules directly, and the full module breakdown is a reasonable place to pick the area you want the project to sit in. If measurement is the weak spot, the five numbers that actually matter is the shortest route to being able to report your own results. For paid work specifically, what a media buying course should teach sets the bar for what your write-up should contain, and our honest answer on the diploma puts the project in the context of what the qualification does and does not do for you.
If you have no access to any business at all, Khaled Badr's walkthrough of building a marketing portfolio with no experience yet sets out three projects that need no client, which is the honest fallback and better than inventing a brief.
The honest limit
A final project will not, on its own, get you a job. It gets you past the first filter and gives an interview something concrete to discuss, which is a smaller claim than the one usually made for it.
It is also worth saying that a strong project takes considerably longer than the time a diploma allocates to it. If you want it to work as a portfolio piece rather than as an assessment, plan to keep working on it after the course ends, when there is no deadline and no marker, which is the point at which most people stop.