Using AI as a collaborator and sanity check

Welcome to Episode 27 of the Boagworld Show, where we dive into a side of web work that doesn't get nearly enough attention. This month, we're exploring life as a freelancer working with small businesses. We're joined by Paul Edwards, a fellow member of the Agency Academy who has spent two decades serving clients that don't have massive budgets or sprawling marketing teams. If you've ever wondered how best practice advice translates to the real world of limited resources and high stakes, this conversation is for you.

App of the Week: Baymard UX-Ray

Before we get into our main conversation, we need to talk about an extraordinary tool that just launched. Baymard UX-Ray is built on the Baymard Institute's 150,000 hours of ecommerce research. If you're not familiar with Baymard, they've been conducting rigorous usability research for years, building an enormous repository of what actually works in ecommerce design.

What makes UX-Ray remarkable is how it applies all that research. You can input your own site or a competitor's URL, and the tool scans it against Baymard's research database. It then provides specific recommendations for improvement, each one linked back to detailed guides explaining the research behind the suggestion.

Now, we'll be honest. Tools like this can feel a bit depressing when you first encounter them. Another thing that AI can do that used to be our job, right? But the reality is more nuanced. You still need expertise to ask the right questions, to know when to ignore advice that doesn't fit your situation, and to implement recommendations effectively. What UX-Ray really does is democratize access to quality research, allowing smaller teams and solo practitioners to benefit from insights that would otherwise require a massive research budget.

For anyone working in ecommerce, particularly if you're trying to compete with larger players, this tool is worth exploring.

Life as a Freelancer Serving Small Businesses

Our main conversation this month centers on something we don't discuss enough in the UX and web design community. Most of the advice you read online, most of the case studies and best practice articles, come from people working with large organizations. We're guilty of this too. Between the two of us, we've worked with clients like Doctors Without Borders, GlaxoSmithKline, and major universities. That shapes our perspective in ways we don't always recognize.

Paul Edwards brings a different lens. He's spent 20 years as a freelancer, and while he's worked with organizations of varying sizes, the common thread through his client list isn't scale. It's circumstance. His clients typically have small or nonexistent marketing teams. They're often time-poor and lack technical expertise. Most importantly, they have skin in the game in a way that corporate clients rarely do.

The Origin Story

Paul's freelance journey started dramatically. On November 5, 2005, he had a tantrum at his job as a commercial manager for a civil engineering company and quit on the spot. No savings, no business plan, no real idea what he was doing. He just knew he'd been teaching himself web design with Dreamweaver and Fireworks, and he thought maybe he could make a go of it.

What followed was the classic freelancer trajectory. He worked his friends and family network, which led him into academia and international development work. He found himself building sites for projects funded by the Bill and Melinda Gates Foundation, DFID, and the World Bank. These weren't necessarily well-funded projects despite the prestigious funders, but they gave him experience working with agencies across Europe and projects in Africa focused on critical issues like hygiene and sanitation.

What Makes Small Business Work Different

When you're working with a small business owner, the stakes are fundamentally different. As Paul put it, the number of clicks their campaign generates directly affects how much money they take home at the end of the month and the security of their family. That changes everything about the relationship.

This isn't to say working with large organizations is easy or that the work doesn't matter. But in a corporation, success and failure are distributed across many people and many factors. When you're working with someone who owns their business, your work has an immediate, visible impact on their livelihood. The opportunity cost of failure is enormous. The credit for success is also more direct, which can be incredibly motivating.

Paul's business has evolved toward more retainer and time bank arrangements over project work. This shift happened gradually but has been transformative. For clients, it guarantees access to his expertise when they need it. For Paul, it provides income stability. But there's another benefit that often gets overlooked. When you have long-term retainer clients, especially small ones with staff turnover, you become a point of continuity in their organization.

One of Paul's retainer clients had a marketing department of two people. Both left within a year. Paul was literally the only person who understood the history of their digital presence, their past campaigns, and their strategic direction. That kind of institutional knowledge is incredibly valuable, and it's something freelancers can uniquely provide.

The Budget Reality

We had to ask about budget because it's the elephant in every room. When you're working with smaller clients, you simply have fewer resources to work with. So how do you adapt all the best practice advice that assumes you have time for extensive user research, iterative testing, and comprehensive documentation?

Paul's answer was illuminating. He doesn't find himself frustrated by advice that doesn't apply to his situation. He just doesn't apply it. As a generalist, he's always picked and chosen what's relevant, learning what he needs for each specific job and disregarding the rest. He can't let his head explode trying to take in everything, so he focuses ruthlessly on what matters for the work at hand.

The reality is that best practice often needs to be adapted regardless of client size. A lot of what gets labeled as essential process work serves organizational needs as much as user needs. In a large organization, you might conduct extensive research partly to align compliance, get legal on board, and protect your client contact from political fallout. In a small business where you're talking directly to the decision maker, you can move faster and iterate post-launch.

That doesn't mean cutting corners on things that matter. Paul still does discovery and research work, but he structures it differently. Rather than one large project with research baked in, he often does pre-project discovery as separate billable work. This allows him to flex the scope based on what the client has in-house, what they lack, and what will actually move the needle for them.

Filtering Clients and Managing Risk

One of the most valuable parts of our conversation was Paul's approach to client selection. He's learned through hard experience that taking on a client who isn't a good fit costs far more in stress and lost time than the revenue is worth. Every single time he's taken someone on when his gut said no, it's been worse than if he hadn't brought that money in.

So Paul has developed a risk scoring process. He researches Companies House filings and financial accounts. He Googles potential clients thoroughly. He makes sure to be himself from the first conversation, explaining that he's blunt and tends to say what he thinks. Some people say they want that but really don't, and it's better to discover the mismatch early.

When things do go wrong, which is rare after 20 years, Paul offboards as quickly and graciously as possible. He sees it as partly his fault for misjudging the fit, so he tries not to burn bridges. He'll help them find someone else to work with and exits professionally.

We wondered whether this kind of risk management is more necessary when working with smaller organizations. After all, you know Oxford University will eventually pay their bills, even if slowly. Paul's experience is that payment risk exists at all scales, but small businesses can have more volatile finances. However, most of his clients pay within 48 hours, which is remarkable. The key is that by moving toward retainer and time bank models where time is paid upfront, a lot of payment anxiety simply disappears.

The Generalist Advantage and AI's Role

Our conversation kept circling back to the value of being a generalist, and how AI is amplifying that advantage. Paul described AI as helping him get out of his own way. If he knows 90 percent of what's needed to help a client but lacks that final 10 percent, he used to decline the work. The opportunity cost of getting it wrong felt too high. Now, AI helps him bridge that last 10 percent with confidence.

He shared a perfect example. A trade business client, selling into the architectural sector, wanted help with their Google Ads campaign. Paul had dabbled in PPC but wasn't an expert. The client was willing to pay him to learn, which was fortunate, and AI supported that learning process. It helped him analyze the massive amounts of data that PPC campaigns generate, identify trends, and fill knowledge gaps. The result was a completely new campaign with much lower spend, a huge increase in relevant clicks, and better funnel positioning. The client was so pleased they sent him a Christmas hamper, a first in 20 years.

This is what the return of the generalist looks like. AI isn't replacing expertise. It's allowing people with broad knowledge and good judgment to tackle problems that previously required specialists. You still need to know enough to ask good questions, to recognize when something feels off, and to verify AI's suggestions. But you can now say yes to opportunities that would have been too risky before.

What Large Organizations Can Learn

Near the end of our conversation, Paul made an observation that stuck with us. While he learns constantly from working with small businesses, he thinks there's value flowing the other way too. People working with large organizations, like us, often miss things that become obvious when the stakes are personal and immediate.

When you work with a business owner who's putting their family's financial security on the line, you can't hide behind process or best practice. You have to deliver real value. You have to be adaptable. You have to become genuinely invested in their success because they're so clearly invested themselves. That kind of clarity and accountability can be harder to find in large organizational work, where responsibility is diffuse and success has many parents.

This Month's Reads

Each month, we share a few articles, videos, and resources that caught our attention and sparked interesting conversations about the state of our industry.

Functional Personas: A Practical Guide

Following up on last month's discussion about AI-generated personas, Paul has now written a comprehensive guide for Smashing Magazine. The article walks through his method for creating functional personas using AI, explaining when this approach makes sense and how to implement it effectively. If you've been curious about whether AI-generated personas can actually be useful, this piece answers that question with practical examples.

Experience Design: The Return of the Generalist

Nielsen Norman Group has posted a video arguing for a terminology shift from "user experience design" to "experience design." Their reasoning is that UX has developed a reputation problem. People think they know what it means, but they're often wrong, associating it primarily with visual interface design.

We have mixed feelings about this. The problem isn't really the word "user." It's the word "design." When most people hear design, they think of visual design and interface work, not the broader strategic and research work that UX encompasses. Changing to "experience design" doesn't solve that fundamental misunderstanding.

That said, the video makes interesting points about the return of the generalist, which aligns with much of our conversation this month. As tools like AI make specialist knowledge more accessible, there's growing value in people who can work across disciplines and see the bigger picture.

Marcus's Joke of the Week

A perfectionist walks into a bar. Apparently it wasn't set high enough.

Find The Latest Show Notes