What Nearly Two Decades of Digital Change Has Taught Me About What Actually Matters

I started Darby in 2008. The iPhone was a year old and the App Store had just opened. Facebook was still figuring out what a business page was for. Google was already the front door to the internet, but ranking well mostly meant getting your keywords and links in order. Marketing automation was clunky, and nobody outside a research lab was talking about AI.

Since then I've watched search get rebuilt several times over, from Panda and Penguin to mobile-first indexing to AI Overviews. Websites went from digital brochures to systems wired into CRMs, ad platforms, and analytics stacks. Analytics got more powerful and then splintered; anyone who lived through the forced move from Universal Analytics to GA4 in 2023 knows the feeling. Social platforms rewrote how companies talk to people, then throttled organic reach and asked them to pay for it. Now AI is changing how people find and act on information in the first place.

In a field that moves this fast, it's tempting to think keeping up means always learning the newest thing. I've done plenty of that. But the question I keep coming back to lately is a different one: what hasn't changed?

The tools change faster than the problems

Clients in 2008 wanted the same things clients want now. They want the right people to find them, understand what they do, trust them, and eventually pick up the phone. They want to know if the money they're spending on marketing is working. And they want some confidence they're focused on the right things.

The answer to those problems keeps moving. For a while it was a better website. Then it was search. Then paid media, then social, then content, then automation. Today it might be AI visibility or how consistently a company describes itself across the web. The questions behind them have barely budged.

If you define the problem by the tool, you end up dependent on the tool. Define it by the outcome and you can swap tools as they come and go. I wish I'd understood that as clearly at the start as I do now.

Complexity is easy to create

One pattern I've seen over and over is how quietly companies pile up digital complexity. A new platform gets added for one campaign. The website gets rebuilt. A CRM comes in, then a second analytics tool because the first one didn't answer a question someone asked in a board meeting. Another agency takes over paid social. An internal hire owns email. Automation gets layered on top, and now everyone has a few AI tools too.

Each decision made sense at the time. The trouble shows up a few years later.

By then there's a lot of activity and very little shared understanding of how it fits together. Traffic is up, but nobody can say whether it's the right traffic. Leads come in, and three different dashboards take credit for them. Content keeps getting published long after anyone remembers what it was for. Each vendor optimizes its own piece without much idea what the others are doing.

More technology rarely buys more clarity. Often it does the opposite.

That has changed a lot of my client conversations. Early on, most of the work was about what to build, launch, or add. These days the more useful question is usually what deserves attention at all, and what can be turned off.

The disciplines stopped staying in their lanes

When I started, you could treat digital marketing as a set of separate trades: web design, SEO, paid search, email, analytics, social, content. Agencies were organized that way, and so were most marketing departments.

Those specialties still exist. The walls between them just don't hold up the way they used to.

Search visibility depends on content, and content depends on positioning. Ad performance lives or dies on the landing page, and the landing page depends on whether the visitor trusts what they're seeing. That trust is shaped by reviews, press mentions, and whatever they ran into before they ever reached the site. Now AI systems pull from many of those same sources and form their own view of a company before a person clicks anything.

Seeing marketing as a system has changed how I work more than any single technology has. It's a big part of why Darby looks the way it does today. I care less about delivering a list of tactics and more about the system they plug into, because a well-executed campaign can still flop if everything around it is pulling the other way.

Fundamentals outlast platforms

Digital marketing has a rhythm to it. Every few years something shows up that's supposed to change everything, and sometimes it changes a lot. Google changed how people find information. Social changed how it spreads. Smartphones put the internet in everyone's pocket. AI is changing how information gets interpreted.

But new arrivals rarely wipe out what came before. They change the conditions around it. When Google's mobile-friendly update hit in 2015, the companies that scrambled hardest were usually the ones with weak sites to begin with. The ones with clear, useful pages mostly just needed a responsive template.

Good writing, a solid reputation, and clear positioning still do the heavy lifting. So does knowing your customer and being able to explain what you do in a sentence. The mechanics change; those keep showing up in new forms.

That's worth remembering in an industry that runs on urgency. There's always pressure to react to the latest algorithm update or acronym. A few deserve it. Most are better watched for a quarter before anyone rebuilds anything. New and important aren't the same thing.

Experience is pattern recognition

When I was younger, I assumed expertise meant having the answers. I believe that less every year. In a field that keeps rewriting itself, nobody gets to keep the same answers for long. What builds up instead is a feel for patterns.

You start to notice when a company is shopping for tactics before it has defined the problem. You can tell when a dashboard is producing confidence without producing understanding. You learn to separate a technology that's changing how people behave from one that's mostly changing the vocabulary. And you get better at spotting when a "marketing problem" is really a positioning problem, a technology problem, or an org chart problem.

Maybe the biggest shift is that I'm much more comfortable saying, "I don't know yet. Let's look at the evidence." I care less about having the fastest answer and more about getting to the right question. Experience doesn't let you predict what's coming. It gives you more past situations to hold up against the one in front of you.

AI is different, and the old lessons still apply

I don't think AI is another passing marketing trend. It's starting to sit between people and information in a way earlier technologies didn't. People now ask a chatbot to compare vendors, summarize reviews, and recommend a short list before they ever open a browser tab. We're early in figuring out what that means.

Still, the principles look familiar. An AI system needs evidence it can parse. It needs consistent information, corroborated by sources other than you, and some signal that separates you from the five other firms that sound just like you.

If a company can't clearly explain what it does to a person, it won't have an easier time explaining itself to a machine. AI raises real new questions, including how it's changing the value of expertise. It also exposes old weaknesses. Vague positioning stays vague when a model reads it. Thin proof stays thin. A company that hasn't nailed down what it does, who it serves, and why anyone should trust it won't get clearer because the thing evaluating it got smarter.

The interface is new. The need to be understood is as old as marketing.

Change is the job

One of the harder lessons of running a technology-focused business is that things never settle down.

For years I treated change as temporary. We'll get through this algorithm update. We'll learn this platform, rebuild the site, adapt to mobile, sort out the analytics. Then things will calm down.

They don't. Something else always arrives.

At some point I stopped seeing change as an interruption to the work and started seeing it as the work. That turned out to be freeing. You stop trying to predict everything. You test more, pay closer attention, build systems that can bend, and revise your assumptions without taking it personally.

You also stop confusing certainty with competence. Knowing what will happen is mostly out of reach. Knowing how to respond when it does is something you can actually get good at, and it's worth more.

What I'd tell myself in 2008

If I could hand my younger self a short list, it would look something like this:

• Figure out the real problem before you pick a tool.
• Make things clearer before you make them more complicated.
• Watch how the pieces affect each other, not just how each one performs.
• Don't drop the fundamentals because a new platform showed up.
• Look for evidence, stay curious, and be willing to change your mind.
• Remember that strategy is as much about what you won't do as what you will.

The technology will keep changing, and it should. Some of it will be incremental, and some will reshape how we work and decide. But the job underneath stays familiar: understand what's happening, decide what matters, and make a better call about what to do next.

Nearly twenty years in, that's still the part of the work I find most interesting. I've given up on predicting every change. I'd rather get good at telling which ones actually matter.