A prospective client does not need to see every project I have touched. They need enough relevant evidence to decide whether I can help with their problem. I organise a service portfolio around that decision.
A large project may contain several worthwhile examples. Separate the contributions so that a reader can understand what each one demonstrates, then choose the examples that support the services you actually offer.
Write the questions first
For a website service, a reader might ask whether I can handle a complex content structure, translate an identity into a usable site or support editors after launch. Those questions suggest the examples to choose. They are more useful than arranging projects only by which screenshot looks newest.
I make a simple table of services and evidence. Against each service I put the strongest relevant project, the contribution I can explain and any permission needed before publication. Empty rows show where the portfolio is making a promise without much supporting material.
Give each project a different job
Two examples can share a platform while demonstrating different judgement. One might explain a difficult content organisation problem; another might show how a visual identity was carried across a site. I state that distinction in the introduction rather than expecting the reader to infer it from images.
I remove repetition where it does not add confidence. Ten similar landing pages may say less than three well-explained projects. The aim is not to hide the breadth of the work, but to give each featured item a clear reason to be there.
Make my contribution visible
I describe what I personally designed or built and credit the work around it. This matters especially when a project has multiple agencies, an internal team or a long history. A logo in a client list cannot explain those boundaries on its own.
I also separate a project from a service. A small technical contribution should not appear as evidence that I created an entire brand. If the contribution is worth showing, I give it an accurate title and explain the specific problem it addressed.
Link the evidence to the enquiry
A service page can show a small selection with a sentence explaining why each project is relevant. A reader interested in editorial tools should not have to search through unrelated visual experiments to discover that I have built them.
The next step should fit the example. After a case study about organising a large content site, I invite a conversation about that sort of problem. I avoid adding an unrelated sales pitch simply because the page needs a button at the bottom.
Review the collection as a whole
I periodically check whether the featured work still represents what I want to offer. I look for outdated descriptions, broken public references and examples whose current state no longer matches the screenshots. Historical work can remain useful when the date and contribution are clear.
I also check the balance. A portfolio should show enough variety to establish range without making the offer impossible to describe. When every project has a clear purpose, the collection becomes easier to maintain and much easier for a client to use.