AI and web design. Yes, AI is going to eat your website for lunch.

An AI image of a robot finding a needle in a haystack

The ideal website is one with perfect findability. It has no navigation, no search and no sitemap. Users get exactly what they want first time.

That's what AI will do.

Very soon an AI embedded on your site will give users just what they're looking for after a few text or voice prompts. 

I expect AI interfaces to get so good they will dominate how ordinary people find and engage with your content. Needle in a haystack? No problem.

Importantly, users are unlikely to treat AI-delivered information in the same way as search results. They will expect and assume that whatever it provides is correct and that everything they need is included.

For many, this new interface will be your website. No need to trawl through vast, confusing megamenus or landing pages with cryptic links anymore.

That's why AI is going to eat your website for lunch.

What's the biggest complaint on most websites?

"Why is it so hard to find anything on this f*****g website?!"

Findability. Even after decades of work and endless redesigns! most sites remain crushingly bad at this most basic aspect of delivery.

Sure, we web folk put our heart and souls into UX (usually in extremely straitened circumstances), but at best our solutions are barely tolerated and at worst they're actively hated.

Design doesn't get better - it just gets less bad.

That's why huge information-dense websites (typical in government) could really benefit from AI.

Let's look at an example.

How AI will eat your website

Suppose I wanted to find out how to visit an inmate in Portlaoise prison in Ireland.

I could to go to the Irish Prisons website - - and look for what I need using navigation or search. Or maybe not, based on results below!

(Note: I don't actually need to visit a prison inmate. I chose Irish Prisons as a good example of an information-dense website.)

A screengrab of search results from the irish prison service website

Alternatively, I could start my search in Google where it (sometimes) presents rich-results pulled directly from the website. That gives a sense of what's coming with AI.

A screengrab of search results from Google about Irish prison service website

But AI is already doing better.

Recently I set up a trial of Microsoft's ChatGPT-powered Copilot Studio which I pointed at the prisons website and then entered the same query as above. It returns quite a detailed description.

A screengrab of Microsoft's ChatGPT-powered Copilot Studio

And Google Bard does a similar job.

A screengrab of results from Google Bard about the Irish prison service

The AI search engine Perplexity returns similar results, but also includes 'related' items which appear both relevant and useful.

This really starts to show the power of AI for uncovering content and filling in the user journey. Done right, this will make it much easier for users to find what they need.

The poor state of content design on many government websites means they don't link to all the information a user needs. They pass the burden to the user to figure out what to do next.

Inevitably this results in an incomplete picture - leading to errors, complaints and wasted time for everyone.

A screengrab of results from Perplexity search engine

A screengrab of results from Perplexity search engine

The future is content (again)

The trajectory is clear.

AI can both find what users need and serve the content directly to them. An obvious effect is that many users will change their behaviour and may never look at your 'web pages' again.

Web teams will need to pivot to address this change.

UX won't be needed as much time for findability anymore. The emphasis can shift to producing high-quality, semantic content that can be served in a structured and coherent way.

Of course, it's still early days and AI has a lot of issues to overcome. Further, the current chat-dominated interfaces do not work well for people with literacy issues.

Additionally, I expect many existing findability patterns to remain. They'll always be a need for traditional navigation to support exploratory behaviour.

But jump ahead a few years to ChatGPT 5 or 6. Imagine their capability.

Remember, users don't give a damn about your website

This reminds us that a website is not an end in itself - it's a means to an end.

Yes, people want the information and services that are on your website. But the website itself? Forget it.

Your users don't give a damn about your website. Why would they? Most websites are painful.

If our users can get what they need any other way, we'll never see them again.

AI may be just what they're looking for.

Update: 25 March 2024

Some interesting reading on the extent to which Gen-AI will replace web designs from Nielsen Norman Group: Generative UI and Outcome-Oriented Design.

Labels: ,

Download a list of the most important skills for a small government web team.

The inside of an old general store with an anachronistic laptop on a box with a screen that says

Excel icon

Core skills for small government web teams (XLS 23KB)

This spreadsheet lists some of the most common skills needed by a small government web team, from QA to Content Design and more.

You might as well admit it.

You will never get all the resources you need for your web team - especially manpower. You'll just have to make do with that you've got.

That means the people on your team need to be able to support a very wide-range of activity - from ordinary day-to-day operations (QA, publishing, analytics, etc) to advanced, high falutin' digital awesomeness (UX, content strategy, etc).

Mom-n-Pop's general store

One way to think of web operations is like a small general store.

To keep a small store running, everyone needs to know how to do just about everything. In addition everyone also needs a deeper expertise in one or two areas for when the need arises.

For example, everyone needs to know how to operate the checkout - but one person must be adept at cash reconciliation and account management.

It's the same on a web team.

'Specialist Generalists' rule!

Small web teams thrive with Specialist Generalists.

Everyone on a small web team needs a good general competence in all core web disciplines (content, design, code, analytics, etc). But everyone also needs specialisms in a few key areas.

For example, everyone should know how to write in plain English, but at least one person must have advanced Content Design skills.

I've worked with web teams for 20+ years and the core skills needed to keep the show on the road are broadly stable. Download a sample list of core skills above or explore a more complete framework in the 'Web Management Masterclass'.

Warning! Don't romanticise small teams

Of course, we shouldn't romanticise the capability of small web teams. They're not "small". They're stunted. The reason is they're in penury.

Most colleagues outside of web teams have no idea what we web folks actually do ("...something about design?!?")

This blind spot is why web teams are critically underfunded. And that is why many websites fail (especially in government), leading to repeated and wasteful "redesigns".

The truth is that it is becoming harder and harder for even the best teams to cope.

Web has become so complex and specialised that we are well past 'peak geek'. No-one person can know it all any more.

Just think of how much more sophisticated GA4 is compared to GA Universal. It is a transformational change that many struggle with - even the specialists.

And it's only beginning.

Photo by Galt Museum and Archives on Unsplash.


Off-topic... If the first question of philosophy is 'why does anything exist?', the second must be 'why is it so boring?'

A picture of God being handed a P45-form with text that reads 'A nice buy but poor performer'

Don't get me wrong. The universe is certainly majestic.

Yet even the most simple video game offers a far more fantastical world than we seem to inhabit.

If God was a game designer, he'd be fired.

Why is that? Why is creation so pedestrian?1

If we assume creation was truly open in the beginning2, surely anything could have happened? Literally anything (logically possible). The degrees-of-freedom must have been almost infinite.

And yet we get just 3 spatial dimensions. 1 above flatland. Yawn.

Why not 4 dimensions? Or 10? Or 100?

It seems suspiciously ordinary.

Of course, in one way it does make sense.

We evolved within this arena and - as many philosophers old and new have related3 - natural selection imposes a structure on our senses and sets a limit to what we can experience.

The result is that we have limited degrees-of-freedom in how we can hold4 the world.

But the world itself is not so constrained.

Just 3 dimensions? You must be joking!

In the 'Game of Creation' we humans may be eternally confined to the Beginner setting of Level-1 - but if we can master it, we may just get a glimpse of what lies beyond.

(Read my post about 'Where to start with philosophy', including links to useful podcasts, videos and more.)

The footnotes...
1. Of course, creation is amazingly complex. But compared to the possibility space of what may have been, it is stupefyingly straightforward (at least in the naive sense of how we perceive it).
2. A big assumption I agree, but then again why wouldn't it have been open? Why was 3 dimensions the obvious default (excluding the possibilities of curled-up extra dimensions and the holographic model)?
3. From Immanuel Kant to Donald Hoffman to Anil Seth.
4. Shout out to my main man Hilary Lawson (IAI) and 'closure' theory (I'm a fan).

A good web team can do 5 things - and now it's time to choose.

An illustration of the 5 levels of web delivery as explained in this article against a background illustration of an antique ladder

Many senior management teams struggle with web. In fact, they often have no idea what their web teams actually do. As far as they are concerned, things sort-of just "happen".

This is a huge risk. It's one of the main reasons websites fail, especially in government.

Web teams are critically short of the funding they need because senior managers (quite rightly) won't invest in something they don't understand.

As web professionals, we have to take some of the blame for this. All our prissy techno-design-babble has gotten in the way of asking for the basics: manpower, skills, tools and money.

We have to divert conversations with senior managers away from the fun side of web (design! interaction!) - and onto the nuts-and-bolts needed to keep the show on the road.

What web teams do in 5 levels

To that end, I now describe what a web team can do based on 5 levels of delivery - from the ultra-basic ("this website is impossible to use, but at least it's legally compliant") to advanced, hi-falutin' digital awesomeness.

These 5 levels are:

  1. Compliance: Your web team can ensure the site is compliant with legal and regulatory rules, e.g. GDPR, cookies, accessibility, etc.
  2. Business-As-Usual (BAU): Your web team can satisfy basic day-to-day business needs, e.g. publishing press releases, creating images, adding PDFs, etc.
  3. Quality Assurance (QA): Your web team can deliver a minimum standard of quality for content/services, e.g. no broken links, no missing metadata, etc.
  4. User Experience (UX): Your web team can deliver a good standard of UX for core content/services, e.g. good findability, good content design, etc.
  5. Digital Goals: Your web team can successfully deliver on strategic goals, e.g. digital cost savings, etc.

As well as being easy to understand, these 5 levels act as a sort-of menu for senior managers to choose from.

Want your web team to deliver fabulous results for your strategic goals (level 5)? Great!

However, you can't simply jump straight to awesomeness. Web follows a strict ladder-of-progress.

The web ladder-of-progress

In that regard, levels 1-3 (Compliance, BAU, QA) are essentially non-negotiable. They are so fundamental that every web team must be resourced to deliver them.

Happily for senior managers, the level of investment needed is quite low. These activities do not require many specialist skills (though the tasks themselves are often very manual and need plenty of time).

The picture is quite different for levels 4 (UX, Content Design) and 5 (Goals). The step-up here is transformational and the resource needed for delivery is an order of magnitude greater.

That means that if your senior management team tells you to aim for stars, tell them to reach for their cheque books - they're going to have to spend a lot of cash on manpower, skills and tools.

The reason is that great websites do not just "happen". Great web teams make them happen.


Your website is not an art project. It's a machine for doing things.

An internet meme from The Simpsons. A sign on a bus says: Your website is not an art project. It's a machine for doing things. The bus driver says don't make me tap the sign!

Think of a public art project, say, for a new sculpture. Consider how it progresses.

First a tender is advertised and an artist is chosen. After agreeing a brief, the work begins. A year or so later the new art work is announced to the world.

And then what happens? Absolutely nothing.

Web as 'art project'

A red-velvet rope is hung around the sculpture so people can admire it from a respectful distance. Apart from the occasional dusting by a contract cleaner, it never changes. In any case, all the budget has been used up so nothing can change anyway.

(Read my article: Why "have budget ... must spend!" leads to bad websites.)

That's why 'art project' is such a good analogy for many websites.

These websites are not expected to do anything. They are aesthetic objects. They are valued primarily for their 'look-and-feel' based on corporate-approved imagery and colours. Little else.

Web as 'machine'

In contrast, consider a machine.

Unlike an art project - which people step away from - people step towards a machine. They need to, so they can use it to do things.

A good machine earns its living. It hums with activity as skilled machinists endlessly tweak, tune and refine it. It even looks nice because the same professionals know how to craft elegant form from function.

That's why many websites cannot be thought as 'machines'.

Even if they work at first, they typically seize-up after a short time because they lack the fulltime specialists needed to keep them going, e.g. user researchers, content designers, UX designers, analytics specialists, developers, etc.

Don't make me tap the sign!

It's clear that a lot of websites - especially public websites - are still in the 'art project' phase of digital maturity. They are stuck in endless loops of "redesign" in the misplaced hope that just-one-more will fix their problems.

Nope. Don't make me tap the sign! It'll never happen.

Instead a change of mindset is needed. These organisations need to understand what a website really is. It's not an art project. It's a machine for doing things.

When the penny finally drops, this new approach to web will seem so obvious that they'll wonder why they ever did it any other way.

(Read my article on Why digital government is failing - and how to fix it.)


The new "O3 Digital Capacity Model" can score your web team's ability to deliver across 3 levels of activity.

A multicoloured poster with the chemical formula for Ozone O3

The O3 (Ozone) Digital Capacity Model assesses your web team's ability to deliver across 3 levels of activity:

  • Online
  • Operations
  • Organisation

(Read a detailed explanation of categories at the bottom of the page.)

The model works by benchmarking key inputs and outputs, and exposes holes in your capacity and resourcing.

Deploying the model is easy. Simply evaluate delivery against the criteria within each category and then grade the results: Red, Amber or Green. The grades indicates risks in your team's capacity.

I have been using this model for several years and it has been superb for grabbing the attention of senior managers. Not only are the results very clear, but they are easily presentable on a single slide (as illustrated below). This clarity just seems to draw senior managers' interest and compels them to focus on web.

(Find out below why it is called the 'Ozone' or 'O3' model below.)

O3 (Ozone) Digital Capacity Model

  • Online: This category benchmarks your website against common delivery standards, e.g. UX, content design, accessibility, performance, security, etc. The basic question is - "does your website meet the minimum standards of quality and experience your users expect?" If not, why not? Where are the holes?
  • Operations: This category benchmarks web processes, people/skills, tools, management practices, etc. The basic question is - "does your team have the capacity to carry out all essential maintenance activities based on the scale of your site?" If not, why not? Where are the holes?
  • Organisation: This category benchmarks strategy, governance, financing, etc. The basic question is - "does your organisation have the capacity to get the most from its web investment, i.e. is it digitally literate?" If not, why not? Where are the holes?

A list of the detailed scoring criteria is at the bottom of this page.

Example of results from the O3 (Ozone) Model

Download a Powerpoint slidedeck with an example of results (PPTX 65KB).

An example of results from the Ozone Model, showing Online graded as Green, Operations as Amber and Organisation as Red

For example, if a review of Operations reveals that your team has inadequate skills or follows shaky processes, you may grade it as Amber. This grade indicates that your current standard of delivery is poor and that your capacity to deliver is at risk.

Senior managers must then decide whether they accept that risk (and the impact it entails) - or if they will invest in the resources needed to raise standards.

Why digital capacity matters

The scale of demands placed on many web teams is utterly overwhelming. There is simply too much to do and not enough people, time or funding to do it all. (Read my post 'Why digital government is failing - and how to fix it'.)

Yet outside of digital teams there is little to no awareness of the enormous gulf between expectations and reality. Part of the problem is that we 'digerati' are very bad at making the case for increased manpower and funding.

We are typically so flexible and accommodating that - rather than looking for justified increases in resourcing - we simply find ever more creative ways to make do with what we've got.

Guess what happens? Colleagues keep piling on the work, making a bad situation worse.

The truth is that many web teams are stretched so far and have so little redundancy, that almost any issue could bring things crashing down, e.g. staff illness.

This has to stop.

As professional practitioners we need to get much, much better at communicating shortfalls in capacity and how this puts delivery at risk.

The O3 Model can help.

Why the O3 model works

The O3 (Ozone) Model works because it is a convincing.

The 3 categories (the 3 Os) in the model encompass all the elements upon which good digital delivery relies and grades them in a clear way.

It's main strength is how easily it can communicate the totality of your digital position in a straightforward and meaningful way to non-experts.

There is no need for senior managers to understand anything about code, content, design, UX, analytics, technology, etc, etc. The model's 'traffic light' scoring highlights digital risks so clearly that its findings are hard to ignore.

The 3 levels of the model also reflect a useful cascade of dependence.

That is, to be good at Online you need good Operations and to be good at Operations you need a digitally literate Organisation. Fail in any one of the 3 Os and you will likely fail overall.

That's why the model is so useful. Not only does it expose holes in your current capacity, but it can also predict future risks (if things are not fixed).

How to use the results

The most obvious way to use results from the O3 Model is to create a roadmap for improvement across each of the 3 categories, for example:

Inevitably, this all assumes that senior managers will give you the resources you need. In that sense, the best way to use the results is as a business case for web investment.

As noted above, the O3 Model exposes risks so clearly that it gives senior managers nowhere to hide. They have everything they need to make informed decisions about web.

If they accept the model's results but then choose to deflect requests for resourcing, they do so with eyes wide open. If demand continues to grow, quality will decline, risks will increase and ultimately, the site will fail.

Of course, web teams will also continue to go above-and-beyond the call of duty to keep the show on the road - e.g. by prioritising core content and ignoring everything else - but the problem is that many teams are running out of road.

O3 (Ozone) Digital Capacity Model - Detailed Criteria


The category of 'Online' represents your ability to deliver a good website, i.e. the digital property itself. The basic question here is - does our website meet the minimum standards of quality and experience our users expect? Key factors to assess include the following:

  • Quality Assurance: Does your site meet minimum accepted standards of quality. This includes typical things like broken links, typos/grammar issues, consistent language (style guide / controlled vocabulary), metadata, image optimisations, etc
  • User Experience (UX): Does your site deliver a good experience for users. In particular, do you have good content design, i.e. is content easy to Find, Read, Understand and Action?
  • Accessibility: Does you site meet WCAG 2.0 guidelines.
  • Functionality: Does your site operate predictably without error
  • Performance: Does your site respond with accepted time limits
  • Security: Does the side meet standards for Confidentiality, Integrity and Availability.

Explore a complete list of criteria in the Website Manager's Handbook.


The category of 'Operations' represents your capability to deliver a successful website. The basic question is - are you carrying out all the essential activities of web management using robust and repeatable processes based on the scale of your site.

Key factors to assess include the following:

  • People: Do you have the resources in place to adequately cover all essential tasks to the right level of granularity and detail. In particular do you have
    • Manpower: The right number of people (whether full-time, consultants, contractors, part time, support from colleagues)
    • Skills: Do your people have the expertise and experience to meet required standards
    • Teams: Are people organised such that everyone knows what they need to focus on, with well defined roles an responsibilities
  • Processes: Do you have documented procedures and standards that allow delivery to happening consistent and reliable way - and also be repeatable and auditable. Can you do it whilst maintaining good relationships with your internal stakeholders
  • Tools: Does your team have access to the right tools and other supports needed to do their jobs. This includes everything from CMS, to graphic editing tools to IDEs to style guide and design patterns. Basically, anything your people need to deliver.


The category of 'Organisation' stands for the capacity of your organisation to support digital development. The basic question is - is your organisation 'digitally literate' enough to get the most from its web investment?

In my Web Masterclass, I have written that the detachment of senior managers from web is one of the biggest risks to successful delivery.

Many senior decision makers have little awareness of the effort needed to maintain a sophisticated, modern website. Yet, they consistently demand the highest of standards and are often amazed when their expectations cannot be met.

This mismatch places huge pressure on web teams. The only thing that often keeps the show on the road is the flexibility and adaptability of staff.

The Ozone Model helps expose this incoherence and brings senior managers closer to web.

Some key factors to consider are:

  • Digital strategy: Do you have one? How clear is it? Is it achievable or just a set of vague aspirations?
  • Governance: Do you have robust and informed systems to support decision making, e.g. which goals to focus on, how this impacts resourcing?
  • Authority: Do you have the authority needed to direct web without undue interference - especially in UX and design decisions (to help circumvent the endless time wasted debating homepages).
  • Funding: Do you have access to the minimum funds necessary to deliver on your strategy?

Why "Ozone"?

The model assesses digital delivery and capacity across 3 interrelated categories, each of which happens to begin with the letter 'O'- Online, Operations, Organisation. Thus we get 3 Os.

It also happens that O3 is the chemical symbol for Ozone - a substance known to have developed several unfortunate holes over the years.

Putting it all together, the name seemed to suggest itself.

This model developed from a framework I created about a decade ago. See my article 'How to weaponise your web governance' for more.


To create government services in plain English, the 'State' must delete itself. Here's why.

A picture of Kathleen Ni Houlihan (Lady Lavery) with the face being erased

What does 'the State' mean to you?

In Ireland (where I live), it's not one thing but many.

'The State' refers to our institutions of officialdom - courts, ministries, agencies, etc.

It's also the area of Ireland's geographic jurisdiction - our dry land.

Lastly it describes a sort-of spirit of Ireland-ish-ness. An administratively sanctioned Kathleen Ni Houlihan if you will.

'The State' is used with such frequency in governmental communications here that I never thought much about its multiple meanings - until I had to.

Find, Read, Understand and Action

I'm a web guy. My job is making digital information and services easier to Find, Read, Understand and Action. I'm good at it.

From 2016-2018 I worked on a content design project for the Irish Immigration Service. We had a lot of success. Rewriting complex procedures in Plain English was at the heart of it.

It was there that the opaqueness of 'the State' and its varying semantics became clear to me.

(Learn more about this project in 'Effective content for better government' on Slideshare.)

"Apply to the State for permission to study in the State"*

This sentence makes complete sense in official-speak

But using one word with several meanings in a context like this just confuses normal people - especially immigrants who may not speak English as a first language.

With a little work we can clarify the difference between 'the State' as an institutional entity and 'the State' as a geographic area and reduce the cognitive burden on service users. 

How about...

"Apply to the government for permission to study in the State"?

Or maybe...

"Apply to the State for permission to study in this country"?

That's a start. But we can go further...

"Apply to the government for permission to study in Ireland"?

Even better. But again we can do even more.

Let's make some assumptions about users who visit an official immigration website for Ireland:

  • They know who we are (an official agency)
  • They what we offer (official immgration services)
  • They know what country we are from (Ireland!)

So maybe something very simple will work.

"Apply to us for permission to study here."

'The State' has been overthrown! As least terminologically.

To create government services in plain English, this needs to happen wholesale across the civil service ... and perhaps it is.

Terminological putsch!

During the last census, the 'State' was replaced with 'country' in most information campaigns.

That's progress.

More recently, you may have heard the public service announcements on the radio: "With the ongoing war in Ukraine, the government has put in place measures to help with the increased cost of living ..."

The government?! 'The State' is out. Government is in!

And yet, I don't think this anachronism will entirely disappear. It's too popular.

I mean, it feels good to say or write 'the State'. It has weight to it.

"Sit up straight, you're dealing with the State dammit!"

It lends officialdom to officiousness.

It also suggests 'the State' has agency. It has rights and interests. It does things. Not the government, not civil servants or people.

So when mistakes happen ... well, 'the State' is to blame. Which means no-one is to blame.

Except perhaps Kathleen Ni Houlihan.

*OK, this precise sentence did not occur, but many came damn close.


GA4 is looming as a 'Great Filter' - and many web teams won't make it to the other side

A drawing of a guillotine against the backdrop of an email about GA4

GA4 will be a 'Great Filter' for web.

It requires a level of expertise that many institutions simply don't have. Come July, they will crash and burn.

The best thing they can probably do is to switch to a much simpler analytics tool. Statcounter, Matomo et al - this could be your moment!

Let's be honest with ourselves.

Most institutions never used data from Google Analytics in a meaningful way anyhow. Beyond some (often misleading) page view stats, the vast power of GA Universal was never tapped.

So why bother with GA4? It's a huge and unnecessary operational burden.

Simpler will be better.

Yes, there will be denial, tears and gnashing-of-teeth but the writing has been on the wall a long time.

Web teams are in penury.

Without massive investment in skills and manpower, the expansive expectations of senior stakeholders just can't be met.

For many, the looming failure of GA4 is just a symptom of this.

No, your website does not belong to just the Communications Team. It's much too important for that.

A motorway exit sign pointing to a town called Boring

Most government websites are not built for communications.

Seems weird doesn't it? I mean many such websites are actually owned by Communications Departments, so of course they are about communications!


I have created many government sites (local, regional, national). Much of the content they contain is of little interest to their communications teams.

Manuals for information and services

You see, good government websites are manuals for citizens who want to do things.

It’s no criticism to say that communications teams are just not interested in manuals. They're interested in reputation management, image management, launches, interviews, events, campaigns, etc.

The findability of the 'pothole reporting service' simply does not align with their primary concerns.

Apart from 'About Us' pages or 'Director Bios' communications teams are quite detached from the content and services on the websites they manage. On the rare occasions when their interests do align, it is for very specific reasons, e.g. Brexit, COVID. When those issues pass, things go back to normal.

That's why it's so important for those who manage web to make a point of getting close to - and spending time with - the business owners who produce the content users want. It may not be sexy, but it's what the website is mainly about.

I referred to other aspects of the potential misalignment between web and communications in a previous blog post “Why web teams don't care about web traffic”.

A new home for web?

So where does web belong? There are no easy answers.

It started out in IT, because technology was so important back then. Over time it migrated to Marketing/Communications as tech was commodified and the justifiable need for better online image/message management emerged.

But web has not stood still.

It is much more than a communications channel - it is a service channel. In fact, most of web is now about service delivery ('what do you need?'), not communications ('look at what we are doing'). 

By way of example, the core skills of web (interaction design, content design, accessibility, etc.) are skills for creating and managing digital products and services - not managing communications.

Yet many communications teams don't know that they don't know how to manage a modern website - partly because they won't hire the people who could tell them

That's one reason for the disconnect described above.

Such teams continue to manage web as if it was for communications, when in fact the majority of user engagement is for content and services the communications team knows nothing about.

Organisations that have tackled this issue seem to have converged on a model that establishes a 'digital support service' at a corporate level. Under this model, relevant services and skills are managed centrally and made available to all business units as needed, including communications.  

(Image credit: 'Boring' Chris Murphy on Flickr.)

Web managers don't care about web traffic. But we do care about something else.

An antique TV set

It's a common misunderstanding. Web managers often have trouble explaining it to colleagues.

You see, we (mostly) don't care about traffic.*

That's for Marketing or Communications teams to worry about.

Our focus is delivery.

Delivery über alles!

A web team's number 1 responsibility is maintaining a stable online presence.

We do this by ensuring the site meets minimum standards (UX, content, loading, accessibility, etc) and is supported by effective operations (publishing, QA, analytics, etc).

Think of us like the production team in a TV studio.

We do everything we can to make sure things go smoothly - the cameras are in place, the lighting is on, the scenery is up, the scripts are proofed, the teleprompter is loaded.

Yet, we mostly don't care if anyone is watching.

That's not our concern - our job is delivery.

Business-as-usual is enough

Modern websites are so large and complex, that simply keeping the show-on-the-road is a huge undertaking.

Any sensible web team will have its own strategy to steer activity. For example:

  • UX: Improve the readability standard for the top 5 content topics to grade 6 by end of year.
  • Resourcing: Hire x2 full-time content designers by mid-year.

This web strategy is agnostic on marketing or communications priorities. It is only concerned with optimising delivery for users and the business.

Confusion about web strategy

Did you know many websites (far, far more that you might think) operate in the complete absence of marketing or communications goals?

It's true.

They may have vague aspirations about being "world class", but they are totally silent on actions, resources or outcomes.

Annoyingly, web teams are often asked to fill-in and to produce goals/targets for sites like this. That exposes the confusion about web strategy.

Goals for what? Targets for what?

At its core, web strategy is about delivery.

"But what about traffic, reach and engagement? What about communications, marketing and business goals?"

Sure, they are also important and they can be included in the strategy. But first someone needs to decide what they are - and that is not the web team's job (or at least not their job alone).*

* In practice, web teams are usually highly involved in developing communications goals for web and are very interested in traffic. But that is only because web is such a dominant channel. No-one expects the print department to set marketing goals - they focus only on high quality printing.

'Sleeping in your car?' Web teams are in penury and that's no accident.

Man sleeping an old car

You might know the story.

You are introduced to someone at a party.

They are very engaging, seem amazingly successful and drive an immaculate car.

Only later do you discover that they regularly sleep in that car and the tank is usually empty.

The car looks great because it's all they have - and they work very hard to keep up appearances.

It's a wretched state to be in.

In a strange parallel, the situation for many web teams is the same.

You probably agree that most government and institutional websites look great.

They have slick designs, beautiful images and gorgeous, florid "about us" text.

At first glance they seem amazing!

Only later do you discover that's all they have. There is nothing else.

It's almost impossible to do anything useful with them.

Their web teams are in penury. There is no budget for meaningful development.

The tank is empty.

The philosophy of web governance. The real reason web teams keep going.

Statue of a philosopher


That's why we do it.

Just consider the incredible complexity involved in managing a modern website.

Consider the array of activities, people, skills, teams, tools, technologies, processes and procedures. Consider the ever changing needs, moving parts, critical dependencies, growing volumes, evolving drivers and competing demands.

The reality is that in the absence of a system of governance, things soon fall apart. The examples are endless.

So the purpose of website governance is to deliver stability.

Stability means a web team doesn't have to waste time on fights about ownership, unclear priorities, dodgy processes, etc.

Instead, you have everything you need configured in the right way.

You can just get on with things—and focus your effort on pursuing online goals.

A utopia? You bet it is!

Learn more about it in my Website Management and Governance Masterclass.

(Yes, I love philosophy and ways it can connect with digital.)

(Image credit: Dove on J-J Rousseau's hand. Knoten2010 on Flickr.)

Off-topic... Consolations of philosophy? What a joke!

Bleak street

A while ago I reshared a link on Twitter to an excellent article about free will.

The article stated what I have often felt, "...the free will problem is really depressing if you take it seriously. It hasn't made me happy..."

I have been reading philosophy part-time for about a decade now (see my post about great web resources for starting philosophy). The truth is that many of the most important questions have very bleak answers.

Ummm, nope...

Consider the following topics, where many very smart thinkers have converged on similar answers (with of course, many dissenters and counter-arguments).

  • Free Will: You don't have it. All your actions have "causally sufficient predetermined conditions" as they say in the trade.
  • The Self: Sadly, there is no 'you'. 'You' are concatenated experience. Take experience away and—poof!—you're gone.
  • Moral Luck: You're locked in. People do as they do because they do as they do.
  • The External World: Utterly unknowable. (I am a particular fan of Hilary Lawson on this.)

And on it goes...

Philosophical about philosophy?

I reflected on this again yesterday, watching an interview with Maria Balaska on the brilliant Institute of Art and Ideas (IAI) YouTube channel

She described philosophy as both a domain of knowledge and a method of analysis—similar in structure to science.

In philosophy, the starting point of analysis is to abandon what you take for granted. Instead, ask why is it that you think-you-know what you think-you-know, and then keep digging.

Knowledge should follow (hopefully)—though, much of it is unsettling.

Some point to Boethius' "Consolations" as a philosophical salve to suffering in the world. That may be true, but do his consolations apply to the practice and knowledge of philosophy itself!?

The only true consolation is that, like David Hume, we are masters at finding ways to ignore ourselves and simply enjoy the sunshine. At least, that's how try to I do it.

Children are a reminder of this. In general, I find my nephews are somewhat skeptical when I try to tell them the external world is an illusion :)

Some interesting observations on children's philosophy in yesterday's video too.

(Image credit: 'Bleak' by Tim Green from Flickr. Image cropped. Creative Commons.)

Stop worrying. Your website will never be right! Zeno's paradox proved it thousands of years ago.

Child on a tortoise

Stop worrying. Your website will never be right!

5th century BC Greek philosopher, Zeno of Elea, knew that all along.

The lesson is in his famous paradox of "Achilles and the Tortoise".

The ancient paradox

Achilles is in a footrace with a tortoise.

Being a good sport, Achilles gives the tortoise a head start of 100 metres at point A.

"Go!" shouts Zeno and off dashes Achilles. He soon reaches point A.

But, uh-oh, the tortoise has already moved on to point B. On rushes Achilles.

But again, when he arrives at point B, the tortoise has moved to point C.

When Achilles gets to point C, the tortoise is now at point D.

And on it goes.

Whenever Achilles arrives at a point the tortoise has been, it has already gone.

In short, he never catches up.

(In reality, Achilles would catch the tortoise, but as a thought experiment it is still intriguing and useful.)

The digital paradox

Now, let's update the paradox.

Substitute yourself (as a digital manager) for Achilles.

Substitute your users for the tortoise.

The point is that, no matter how hard you try, you will NEVER catch up with your users.

They will ALWAYS be some distance ahead of you—and probably a lot further than the tortoise from Achilles!

The reason is that every website is really just an educated guess.

You can never know your users' needs with 100% accuracy. There will always be a gap.

You only find out how big the gap is after you launch and begin to track engagement.

Any issues you find then need time to be fixed, which creates a window for user needs to change even further.

So, stop fretting.

Your website will never be right :)

All you can do is narrow the gap as best you can.

The slooooowwwww evolution of digital (and cycling) infrastructure in Ireland.

Topics as a standard content set

2 domains.

Both with huge demand for higher standards and meaningful delivery.

Both with long-standing, international examples of how to do things right.

Both hobbled by years of wasteful, half-arsed, non-solutions that were doomed to fail from the beginning.

As a casual cyclist and web professional, the parallels between the achingly slow evolution of cycling and digital services from government in Ireland has long perplexed me.

Consider cycling

For 2+ decades it has been irrefutably and utterly obvious that only by creating good quality cycling infrastructure could our major cities deliver the minimally safe environment that existing bike users need—and also encourage more people to start.

Everyone has been in total agreement on this.

It has been discussed endlessly over and over and over again. For years, the necessary solutions have been completely clear from international experience.

And yet millions of euro have been wasted on expedient non-solutions.

Typically this involved smearing a narrow band of red paint into a pot-holed gutter, calling it a "Cycle Lane" and then being amazed (and somewhat hurt) that cyclists ignored it because it was almost unusable.

Only recently have all these non-solutions finally been exposed as complete failures.

Only now are senior decision-makers recognising what was completely obvious from the very beginning—that they need to invest in dedicated, high-quality infrastructure that is designed, built and maintained by skilled professionals.

And so on to digital...

For 2+ decades it has been irrefutably and utterly obvious...

Do I need to continue?

Sadly, unlike cycling, government in Ireland (with a few honourable exceptions) has still not fully exhausted its inventory of expedient non-solutions for digital.

A review of current delivery and governance quality suggests that at least 1 or 2 more rounds of useless web "redesigns!" are likely to happen before the penny finally drops.

That means several more millions of wasted euros before government finally admits it needs to invest seriously in digital manpower, skills and capability.

It's a shame, but somehow this long journey always seemed inevitable.

Don't agree? Read this.

How to manage 'unmanageably' high volumes of website content? Ignore most of it.

I bet you want to know what content your website users are most interested in. Right?

Of course, you do. Me too.

If your situation is anything like mine, you likely have very little time to spend on useful things—like making your content easier to find, read and understand. So, when you do get this time, you want to make sure you're prioritising the right content.

Not the stuff no-one really cares about.

Excel icon

Content Groups analysis spreadsheet (XLS 100KB)

Use this spreadsheet to identify the Content Topics that users of your website are most interested in. Watch the video above for set-up and usage instructions.

But, here's the problem.

Not only are you likely to have vast amounts of content on your site to track as it is—I bet more and more is being added all the time.

And more and more.
And more and more and more.
And more and more and more and more.
And more and more and more and more and more.
And more and more and more and more and more and more ... you know what I mean!

So, volumes are unmanageably high and, yes, I mean "unmanageably". There are very few web teams that could—hand on heart—say they are "managing" their content. We're not managing. We're coping!

So, not only that...the basic metric of content engagement—Views per page—is at best highly suspect for tracking user interests across an entire site. At worst it's misleading and often meaningless.

Read the full article "How to analyse data from Content Groups in Google Analytics using Excel".

Do the boring important stuff well. This editable Quality Assurance (QA) checklist shows how.

Some time ago, I developed a Web Quality Assurance (QA) checklist (xlsx 18KB) and I have been meaning to share it since.

Quality Assurance is one of those web tasks that is widely ignored. I mean, who wants to check lots of old pages against a series of booorrrring checkpoints, when there is sexy new content to add?!

If Gerry McGovern is right about this—and I think he is—much, much more QA is needed to help reduce the "cult of content volume".

Website Quality Assurance (QA) checklist

Excel icon

Web Quality Assurance (QA) checklist (XLS 18KB)

Use this spreadsheet to track and manage Quality Assurance (QA) on your website. Feel free to modify this list as you see fit.

QA is utterly fundamental to a good user experience—and your organisational reputation.

Recall the last time you visited a website that had lots of broken links, massive unoptimised images, documents with unintelligible filenames, bloated useless code, etc. It's shameful how it still continues.

Although much QA is now done using sophisticated tools, manual review remains common. Hence this checklist.

Where to start with QA

As well as checking new content for QA, I also recommend reviewing older content in a series of rolling sprints.

Each month, take one of the top content topics (not pages) on your website and prioritise it for attention.

Do that on an ongoing basis and you'll be amazed how things improve over the course of a year.

Download the editable QA checklist

Feel free to re-use or modify this list for your own site. Admittedly, it is somewhat tailored to the site I created it for (and I have sanitised it to remove many custom checkpoints), but it will likely prove useful to build on.

Download the editable Website Quality Assurance checklist (xlsx 18KB).

You can even have fun applying it to my own website and get points for telling me off about the many checkpoints I break myself :(

If you are interested in other useful web management resources, I published a Website Content Calculator a long time ago that you may also find helpful. Try it and other free web management resources on my 'Downloads' page.

Forget web pages. Make better web decisions using this new system of content analysis.

We already know that ranking pages across a site based on Views is a very bad idea. But, it's much worse if you use them to decide what content to prioritise for attention.

The reason is that there is no such thing as a "standard" web page. Comparing or prioritising them based on Views just leads to bad decisions.

Web pages are far too varied in length, scope and density for such simplistic analysis.

Topics as a standard content set

You can make far better decisions by tracking activity using topics as a new 'Standard Content Set'.

Read the full article "Using topics as a new 'Standard Content Set'".

What to know what your users are *really* interested in? This analytics feature shows how.

As you probably know, measuring aggregate web activity based on Page Views in Google Analytics is not a good idea.

For instance, imagine a site with 2 content topics: Economics and Politics. There are 5 pages about Economics and 10 about Politics.

You look at your analytics. Economics gets 500 views per month. But wow! Politics gets 1000.

Its obvious, therefore, that Politics is much more popular and should get most attention in terms of UX, optimisation, etc.


Look deeper and you discover that each topic gets exactly the same number of visits: 100 each.

It is simply because Economics has twice as many pages that it appears twice as popular using Page Views.

(Why it has twice as many pages is a separate question. Perhaps there is much more information on that topic? Or maybe it has the same volume of information but uses a different content design approach? Or perhaps the information has been poorly planned and arbitrarily separated among too many pages?)

Content in topic groups

This shows that measuring or comparing activity using Page Views (including Unique Page Views) at an aggregate level is a bad idea. Visits are far better.

The problem is that the Google Analytics' default Behaviour report does not count total Visits at a topic level, not even in Content Drilldown. (Content Drilldown does show total Unique Views, but that is not the same as total Visits.)

This is where Content Groups step in

Content Groups flatten issues caused by varying numbers of pages and show what users are really interested in.

Read the full article "Content Groups and Google Analytics".

Funds and financing for start-ups in Ireland

Some moons ago, I ran a small company (a 'start-up' as they tend to be called). It didn't work out - but I learned a lot. 

Over its lifetime, I built a list of funds, financing and other supports available in Ireland. This includes from local government, state agencies, angel funds, seed funds, P2P, crowdfunding, debt funding, etc.

Download the list

I have been meaning to share this list for a looonnng time. Admittedly the information is a few years old now, bit may still be useful to someone starting out. So here it is - finally.

Among the sources listed are:

Money in a bucketLocal Enterprise Office

  • Priming Grant
  • Feasibility Grants
  • Business Expansion Grants
  • First Time Exporter’s Grant

Enterprise Ireland

  • Competitive Start Fund
  • Innovation voucher
  • Innovation Partnership Grant Programme
  • HPSU Feasibility Study Grant
  • Innovative HPSU Fund (Equity)

InterTrade Ireland

  • SeedCorn
  • Acumen
  • Elevate
  • Trade accelerator vouchers


  • SURE tax back
  • R&D tax credits
  • Tax relief for StartUps
  • Exemption from VAT register
  • Use of the cash basis of accounting for VAT
  • Deferral of the Corporation Tax preliminary due date
  • EIS Scheme
  • Start Your Own Business Relief

Department of Social Protection

  • Jobs Plus
  • Wage Subsidy

As noted above, the information is a few years old. No doubt things have changed, especially in light of the pandemic. More supports may be available, so please conduct your own research and due diligence.

Ádh mór ort!

(Image credit: FoodImage on Flickr. Creative Commons.)