Curation Over Volume: The Anti-Portfolio Playbook

The Dilution Trap of the 50-Logo Grid

Open the “Work” or “Portfolio” page of almost any web studio, and you will find the same visual pattern: a sprawling grid of dozens of clickable project cards.

• A rebrand for a national retailer.
• A local bakery’s online store.
• A failed startup’s landing page.
• A quick redesign for an accounting firm.
• A side project built during a weekend.

│                                                                             │
│               THE PORTFOLIO DILUTION TRAP (AVERAGING EFFECT)                │
│                                                                             │
│  STUDIO BELIEVES:                                                           │
│  'Look at our vast experience! 50 projects proves we can do anything!'      │
│                                                                             │
│  CLIENT SUBCONSCIOUS DEDUCTION:                                             │
│  'They built a site for a big company (Grade: A), but they also took        │
│   a $3,000 gig for a local dentist (Grade: D) and a pet app (Grade: C).     │
│   Average Studio Caliber = (A + D + C) / 3 = Grade B- (Mediocre).'          │
│                                                                             │
│  THE PSYCHOLOGICAL LAW: In luxury, you are judged by your WEAKEST piece.    │
└─────────────────────────────────────────────────────────────────────────────┘

The studio believes it is demonstrating breadth, versatility and experience.

In reality it is triggering the averaging effect: readers do not add positive signals together, they form a rough average of the whole set. That is why a restaurant that also does garden waste collection reads as a slightly less serious restaurant.

By leaving small, rushed or compromised projects on your site, you drag down the perceived value of your best work. The projects you are proudest of end up being read as the average of everything else.

---

1. The Art Gallery vs. The Supermarket

Think about how the two kinds of business display what they have.

A discount supermarket packs tens of thousands of items onto metal shelves, competing on inventory volume and thin margins.

An art gallery hangs three paintings in a quiet room with polished concrete floors and museum lighting. It does not feel obliged to prove the artist painted fifty other canvases that year. The restraint of showing three works creates an aura of value.

┌──────────────────────────────────────┬───────────────────────────────────────┐
│                       THE TWO PRESENTATION PARADIGMS                         │
├──────────────────────────────────────┼───────────────────────────────────────┤
│ THE COMMODITY SUPERMARKET            │ THE SOVEREIGN ART GALLERY             │
├──────────────────────────────────────┼───────────────────────────────────────┤
│ 50+ project tiles in a dense grid    │ 3 to 5 curated, monumental works      │
│ Superficial 2-paragraph summaries    │ Forensic case studies                 │
│ 'Look how many clients we have'      │ 'Look at the rigour of our work'      │
│ Competing on volume and speed        │ Competing on craft                    │
└──────────────────────────────────────┴───────────────────────────────────────┘

This is not only true of galleries. It is true of the studios and consultancies you have decided to hire from. Most of them show forty projects. A few show three, properly explained. The ones that show three are, almost without exception, the ones that get the interesting work.

When you adopt the Anti-Portfolio Playbook, you stop presenting yourself as a vendor asking to be validated, and start presenting yourself as a studio with a point of view.

---

2. The Five Parts of a Case Study Worth Reading

If you only display three or four projects, each one has to carry its weight. Delete the two-line marketing summaries and build each project around five parts.

Part 1: The Commercial Problem

Describe the exact bottleneck the client had before you arrived. Avoid polite generalities. State the stakes plainly. The client’s own sentence does that better than your adjectives, and it reads like this (illustrations of the kind of claim, not results we have measured):

• A legacy site causing 5.4-second mobile load times.
• A 42% drop in organic traffic after a bad platform migration.
• An annual hosting bill of $180,000 for a marketing site.

Part 2: The Diagnosis

Show how you found the cause that the client’s own team and previous agencies missed. This is where you demonstrate judgement rather than labour:

• “The previous agency recommended adding more caching. Our audit found main-thread locking caused by hydrating the whole page.”

Part 3: The Intervention

Document the actual engineering decisions:

• Before-and-after structural diagrams.
• The typographic and spacing decisions, and why.
• The move to static hosting, and what that changed.

Part 4: The Evidence

Give numbers a reader can check:

• Core Web Vitals before and after. Google’s Web Vitals guidance puts 2.5 seconds as the good threshold for Largest Contentful Paint, so this is the number a reader can check for themselves — an illustrative before and after (LCP 4.8s → 0.5s).
• Field performance data from the Chrome UX Report, which collects its data from real browsers in the field rather than from a lab test.
• Real screenshots across screen sizes, not mockups.

Part 5: The Result

Close with business outcomes measured 90 to 180 days after launch, not the week the site went live. A result worth publishing reads like this (again, illustrations of the shape of the claim, not figures we have measured):

• A 142% lift in qualified enquiries.
• $3.2M in attributed new contracts.
• $450,000 a year removed from software licences.

When a case study reads like this, the sales pitch becomes almost unnecessary.

---

3. The Purge: How to Prune Your Archive

Founders ask: “If we delete forty of our forty-five projects, won’t clients think we are inexperienced?”

The answer is no.

Prospective clients do not have the time or the inclination to review forty-five projects. They look at two, maybe three.

The Pruning Protocol:

1. Grade every project currently on your site from A to F, based on visual craft, technical quality and whether the work still represents you.
2. Delete everything below an A-. If the project only exists because a client forced bad compromises, it is actively working against you.
3. Choose four showcases that show the outer limit of what you can do.
4. Rewrite them with depth. Spend a couple of focused days documenting those four properly, with real evidence.

The moment you prune your archive, your perceived standard goes up. You are no longer judged by your compromises; you are judged by your best work.

---

Conclusion: Curation Is Confidence

Studios shout about volume because they lack confidence that any single project can stand on its own.

Good studios show three works and fall silent, because the work says it better than they can.

Delete the mediocre. Document the good ones properly.

A short, well-argued portfolio makes a small studio look like the only one worth calling — and that is worth more than fifty logos.

STRUCTURED PROTOCOL // FAQS

Frequently Addressed Technical Inquiries

What is the 'Anti-Portfolio Playbook' in design and web work? [+]

The Anti-Portfolio Playbook is a positioning strategy where a studio deliberately deletes most of its past work and presents only three to five projects, each documented in depth. Instead of thumbnail grids with two-line summaries, each project becomes a short account of the problem, the diagnosis, what was changed, and what happened afterwards.

Why does showing a wide variety of past projects hurt your positioning? [+]

Because of a well-known averaging effect: readers do not add up the good and bad together, they form a rough average of everything they see. Put a small, rushed project next to your best work and it drags the perceived standard of the whole set down. What you show is what you claim to be.

What should a strong case study contain? [+]

Five things: the specific commercial problem the client had, the technical diagnosis, the concrete changes made, the evidence it worked, and the business result measured after the site had been live for months. A case study with no numbers and no reasoning is just a screenshot with a caption.

Sources

1. [1] Web Vitals | web.dev Verifies the Core Web Vitals benchmark the evidence section tells a reader to measure against: Google's documentation states that LCP should occur within 2.5 seconds, INP at 200 milliseconds or less and CLS at 0.1 or less, assessed at the 75th percentile of page loads.
2. [2] Overview of CrUX | Chrome UX Report | Chrome for Developers Verifies the field-data point the evidence section makes: the Chrome UX Report is Google's dataset for the Web Vitals programme, and its data is collected from real browsers around the world rather than from a lab test.