Hide table of contents

TLDR: I looked at the branding of 89 EA organisations. I wanted to double check my intuition that it was all a little 'samey', spoiler: it was. So I made a game (in jest) to see how well people can actually tell them apart.

PLAY THE GAME >>> https://ea-bland-book.netlify.app/

Background

A few years ago I wrote a post arguing that it’s not effective to call everything effective. Sorry to say, I remain annoying about this topic, and sadly it seems there is still quite a lot to be annoying about.

I looked at the names, colours and website messaging of 89 EA organisations. There are lots of blues and teals, lots of Institutes, Centres, Initiatives and Projects, names that are essentially jargon acronyms and unclear websites.

Not all of these things are necessarily, individually bad. Blue is fine. Institutes can be good. Unclear websites can still look nice. But, the problem comes when you put everyone next to each other and realise how hard it is to remember which is which.

So, rather than only writing another post complaining about it, I made a game.

Play that before continuing: https://ea-bland-book.netlify.app/

 

Is this actually a problem?

Quick test:

  • Have you ever been confused with another organisation?
  • Has someone asked how exactly you are different from X organisation?
  • Or perhaps they understand roughly what you do, but can never remember what you’re called?

That’s a branding problem. Or if you reject the idea that branding matters (there's still a lot of that too), then it's a communications problem.

It isn’t just about wanting a prettier logo. If people can’t distinguish you from another organisation, they’re less likely to remember you, search for you, recommend you or know why they should donate to or work with you rather than someone else.

It’s normally fairly critical and important that people understand and remember your organisation if your plans rely on people doing something.

Something that looks nice vs something that works

I think part of the issue is what organisations believe they are buying when they commission a brand or website.

Usually they hire a designer. The designer designs something, and designers are really good at making something that looks nice. Everyone is broadly happy and the project is complete.

But looking nice was never really the brief. Sure, it matters somewhat from a trust and credibility POV, but it's not the job of branding and websites. 

The critical work here is deciding who the website is actually for, what those people need to understand, why this organisation is different, what evidence will convince them and what they should do next. Most designers wont aim to answer those questions.

If those questions haven’t been answered, even a very good designer is being asked to solve a strategic problem through design. They will likely produce a lovely-looking website, but it still won’t necessarily work.

I haven’t formally calculated this, but I would estimate that around 80% of the branding and website projects we complete at User-Friendly are for organisations that have already tried to do the work elsewhere through a free / probono service, hired a designer or built something internally (probably on AI). They received the thing they asked for, but over time it became clear that they ultimately weren’t getting the results they wanted and came to us to do it again.

This is obviously not a particularly efficient way to use limited organisational funding.

The audit behind The EA Bland Book

We saw a similar problem in the audit.

We assessed 47 organisations' websites, chosen at random, across seven critical values. The aim was to see how well branded the movement seems to be broadly speaking. We used the website as a proxy for quality of brand, clarity of their positioning and whether their key objective was clear.

Specifically, we looked at 7 metrics:

  • Name distinctiveness: Is the organisation’s name ownable and memorable, or generic and easily confused?
  • What-it-does clarity: Can someone quickly understand what the organisation actually does?
  • Audience clarity: Is it clear who the website (organisation) is for?
  • Differentiation: Is it clear how the organisation differs from similar organisations?
  • Credibility 'above the fold': Does the first screen provide evidence, proof or a reason to trust it?
  • Call-to-action clarity: Is the next useful action obvious and relevant?
  • Visual distinctiveness: Does the organisation look recognisable, or could its identity belong to almost any other EA organisation?

And here's the pass rates (read with one eye closed, it's not pretty):

The results:

  • Failed to pass all seven tests: 98%
  • Had at least one serious weakness: 72%
  • Did not show strong credibility above the fold: 81%
  • Did not clearly identify their audience: 62%
  • Did not strongly differentiate themselves: 40%
  • Did not have a strongly distinctive visual identity: 45%

Of the 47 websites we reviewed in more deeper detail, 29 scored well for design quality. However, 20 of those 29 still failed at least one basic test around whether the name, positioning or visual identity was distinctive.

Thirty showed little concrete evidence or credibility above the fold. Only one scored strongly across all seven areas we assessed.

There are plenty of fairly nice-looking websites. But we are less good at making organisations clear, different and memorable.

And now AI

There also seems to be a fairly big swing towards using AI for branding, websites and communications work.

Fundamentally, I don’t think this is necessarily bad. We use AI too, and it can be helpful if you have a strong brief, clear positioning, a well-understood audience and someone who knows what good looks like, AI can make parts of the process faster. But you can't just throw this stuff into AI and expect a good output. If you remove the human expertise and judgement, you mostly get rubbish.

In a movement that already suffers from lots of similar names, language, colours and websites, using a tool that naturally moves towards the average is probably not going to help...

Can we solve this?

At User-Friendly, we create brands and build websites, but that isn’t really where the work starts. It normally starts with the audience, positioning, messaging, communications strategy and what the organisation is actually trying to achieve.

We also bring the judgement and taste needed to decide whether an idea is genuinely good, rather than merely acceptable-looking.

That’s the difference between producing something that looks nice and producing something that actually works.

If you are planning a new website or rebrand, talk to us before jumping straight into design (or asking an AI to do it). We’d be happy to help.

https://www.userfriendly.org.uk/

18

1
2
1
3

Reactions

1
2
1
3

More posts like this

Comments8
Sorted by Click to highlight new comments since:

Admittedly knowing nothing about branding or design, the game didn't land for me. The taglines were easy -- knowing anything about the orgs, e.g. that the name includes "philanthropy", is enough to correctly guess which is which. And the others felt like they didn't make the case for themselves. Am I supposed to be able to tell the blacks of Nike, Adidas, and Puma apart?

A more convincing version of the game might have examples of both good and bad branding: show me that, actually, I can distinguish brands I know from just their shade of blue, but can't similarly tell EA orgs apart.

EA branding repeatedly converges on the same cluster of visual and verbal choices. Seeing the examples together makes that convergence unusually obvious.

No, I wouldn’t expect anyone to distinguish Nike, Adidas and Puma by “their blacks” alone; their brands are differentiated through other highly distinctive assets. My argument is that many EA organisations converge across several assets at once: colour, typography, imagery, tone and abstract language all appear the same or similar.

Your version of a game sounds great, I would love to see that too.

Thanks! To be clear, I think you've made your case about EA branding well in this post and others. But I think the game makes explicit or implicit claims that don't hold up, which makes it an ineffective education / persuasion tool in my case.

To massively overanalyze my playthrough:

  1. Redwood and MIRI have pretty distinctive colors. FAR AI, GovAI, and Apollo are all basically black. It seems like this is claiming that, if their branding were better in some way, I'd be able to tell those three apart from the shade of near-black.
  2. The text says "deep breath — every one of these lines could’ve come from any of the five." But I don't think that's true! Only the Simon Institute is involved in "supporting policymakers" (which I can tell by the fact that it has "governance" in its name). Only one of the taglines includes the word "philanthropy." Only Founders Pledge focuses on "entrepreneurs," etc.
  3. Kind of a repeat of 1, except I work in AI safety so I'm less familiar with these orgs to begin with, and have probably never seen at least two of these logos. A suggestion to make this type of question feel less "unfair" and more compelling: show us each of them in color beforehand, and then see if we can match them properly 10 seconds later.
  4. I'm not sure why I should be able to match color palettes to cause areas, even if we had incredibly good branding across the field.
  5. At this point I just submitted without moving any of the sliders to see the real answers. I don't really know what it means to pass each of these tests, or which organizations were included in the survey, so it's hard to take anything away from these numbers.

I think I'm reasonably close to your target audience for this, so hopefully this feedback is helpful!

There might be a bug in the game because the green square in round 3 wasn’t responsive, so I couldn’t complete it. Liking the concept though.

Yeah I couldn't connect the dots in round 2 on my phone; I also like the concept though!

Both fixed - thanks for flagging!

This is brilliant and very funny. Kudos to you for making it, and also for making it poignant. This post hits at the perfect level of 🌶️ spicy 🌶️ that I'd like to see more often on the EA Forum. I wonder whether there'd be a hilarious output if you got AI to average out the colour schemes of EA orgs by cause area (kinda like the face averaging thing that had a moment in the sun several years ago)

Curated and popular this week
Relevant opportunities