VSA’s VP of Technology Thaddeus Ternes was recently published in Ad Age. His piece reckons with the challenges and opportunities AI poses to creatives, and how we can best harness this power while protecting the aspects that make agencies great.

‍

As agencies begin to market AI, there’s an edge of panic to their offerings. “We use AI!” is plastered across websites without any real substance as to how or why. It reads less as FOMO and more as fear of being trampled.

I’m skeptical of AI’s ability to deliver on the hype—but we’ll regret letting panic drive the choices we make.

AI has real benefits. I’m incredibly optimistic about the possibilities for human advancement, equality of health and lived experiences and how we might explore our world and our existence in new ways.

But I’m also terrified about the misuse of emerging technologies when it comes to upholding values such as democracy, justice and civilization—the essentials of a modern society.

Advertisers, marketers and creatives need to admit that using artificial intelligence embodies this same paradox. We need to navigate these challenges pragmatically, with eyes wide open. From ethical considerations to legal repercussions, these are the four questions we as an industry should be asking ourselves when we think about using AI in our business practices:

What are we trying to accomplish?

Right now, it feels like we’re all running to catch the AI train without asking where it’s going. Let’s slow down and figure out what we really want to accomplish with AI.

So much of what I hear when people tout the abilities of AI is “faster, cheaper, more.” Is that really what we are chasing as agencies? Is that what you want clients to think of you?

I don’t think so. Most of us got into this industry because of our love of the craft and the experience of working with others to create something bigger than ourselves. Beloved brands are rarely associated with doing things faster and cheaper. They’re based on qualities such as reliability, wonderful customer service and a great product. If you’re going to adopt AI, my recommendation is to think about how it bolsters the things that already make agencies great. Maybe it takes mundane tasks off creatives’ plates so they can spend more time doing deep, slow thinking. Maybe it helps writers or designers or animators flex into new skill sets more quickly, diversifying perspectives and breaking down silos.

There’s no one-size-fits-all solution here, and some of it will take exploration. But if we can step back and figure out what the end game is, and how we ultimately want to present ourselves, we can make more intentional choices that support the long-term health of our business and our industry.

How are we respecting human creativity?

The progression of technology in the last decade—particularly in data science, machine learning and generative models—is astounding. Advances in parallel computing, mass data storage and new learning algorithms have unlocked seemingly magical capabilities. The results of these advancements are truly remarkable.

Much of it is also undeniably stolen.

That’s toxic. What rights to fair use do automated technologies have? Who gave them these rights? (The answer: No one.)

At the heart of our industry is the belief that creators deserve credit and control over their work. This principle should guide our choice of AI platforms. We should invest in technologies that align with our values and ethical standards rather than those that undermine the roles of human creators. Moreover, platforms should be required to recognize, credit and compensate artists for their work. It’s only fair that creators be rewarded for what they create. In an era when personal brand is increasingly important—but also in which AI threatens to sap us of our ingenuity—human creativity needs to be uniquely recognized and rewarded.

The platforms we use should recognize the value of creative work, crediting and compensating artists fairly. Period.

At the heart of our industry is the belief that creators deserve credit and control over their work. This principle should guide our choice of AI platforms. We should invest in technologies that align with our values and ethical standards rather than those that undermine the roles of human creators.”

How are we protecting our businesses?

Given the uncertain legal landscape surrounding AI-generated content, we’re going to need tools that serve as guardrails.

For example, I can see the benefit of a tool that works as an “AI vault,” meaning a detailed audit trail of all AI-created content, the parameters and processes used to generate it and how humans were involved. A record like that would serve as a crucial accountability tool, providing evidence of how content was created. In the event of legal disputes or copyright issues, this archive can offer a clear trail of AI usage, helping to resolve conflicts and demonstrate compliance with ethical standards.

Another thing to consider is transparency with clients. If we’re really going to embrace using AI, it shouldn’t be a secret we’re trying to hide. Clients should know when we’re using it and give explicit signoff on AI use in their projects.

One way to do that is to include AI usage in scope-of-work documents. This practice helps protect agencies against liabilities and ensures that clients are fully informed and comfortable with the technologies being used. It also builds trust, as clients are reassured there are no hidden elements in the creation process.

How can we advocate for better industry standards?

Creative agencies should not wait for tech companies to self-regulate. Instead, we have a responsibility to advocate for regulations that protect artists and creatives. Business leaders in the creative industry wield substantial influence and can use their platforms to push for legislation and policies that ensure fair and ethical AI usage in their fields. This involves engaging with elected officials, participating in public discourse and collaborating with industry peers to promote standards that prioritize the rights and contributions of human creators.

I’m not trying to be a Luddite. Integrating AI into the creative industry could yield some seriously exciting possibilities. But there’s always the danger of letting the shiny new toy become our value proposition.

The next 18 to 36 months might bring major headwinds for agencies, as companies decide they can use AI to replace the work agencies have provided. But we’ve seen this movie before, i.e., when the move to in-house creative shook our industry. That worked for a while, but ultimately, companies realized the benefits of bringing in outside, flexible support, and they came back.

It’s fine to tout your usage of AI. Being on the cutting edge of technology is always appealing to clients and prospective employees. But let’s not lose sight of the real thing that makes clients come back to agencies time and time again: the people.

This piece originally appeared in AllBusiness.com; see the published piece.

Shiny and new vs. tried and true

At this point, anyone in marketing has seen (or been the accomplice to) an unnecessary website overhaul.

To be fair, there are very persuasive reasons to burn it all down to the ground. Our culture deeply values the new over the old—particularly in a corporate world where “newness” can get more attention and accolades than the much less flashy “improvement.” The “what sounds better on a resume” question is compounded by clients repeatedly being sold or given the impression that they need to start over entirely to solve their business issues. It’s a chicken-or-egg question: Which idea came first—the one that website overhauls are necessary for creating change or that it is easier to sell a website overhaul when it is critical to change?

Why website overhauls don’t work

Here’s the problem: website overhauls don’t work—at least not in the way they’re promised to.

Website redesigns are often initiated because there’s a business problem: low conversions or poor lead capturing mean the website isn’t driving a significant portion of revenue; an underinvestment in the website leads to an outdated and out-of-touch image; acquisitions have grown the capabilities of an organization, but those aren’t reflected on the website. The list goes on.

Sure, a website overhaul may solve those problems … in the same way that firing your entire staff and hiring new people may improve employee morale.

But, like that analogy, website overhauls are expensive and hopefully unnecessary. In both the time they require and the financial investment needed, website overhauls demand the dedication and focus of large teams working for weeks, months, and sometimes even years. This time commitment poses problems of its own, adding in a layer of “well, we’ve spent so much time on it, now it really has to perform.” In other words, the perfect recipe for stakeholder disappointment, delivered late and way over budget.

It can also create more problems than there were to begin with. From structure changes to broken links, a great deal of complexity is involved in successfully migrating a site without decreasing SEO rankings. Data and content migrations are often underestimated herculean efforts, not to mention external integrations. A misaligned overhaul can also turn users off to the platform, creating frustration or confusion and causing a loss of trust and familiarity.

Lastly, it’s just not the best way to work. A website isn’t a static billboard—it’s a selling tool that should be responsive to customer needs. Responsiveness means it adapts as customers or the business adapts. Every month you spend on an overhaul is time that could have been put into making the site 5% better for the customers currently interacting with it. By the time you finish, the insights that you were implementing may no longer be relevant.

A website isn’t a static billboard—it’s a selling tool that should be responsive to customer needs.”

Why incremental change does work

When we spread the gospel of incremental change, we cite three reasons to believe: speed, sustainability, and CMO satisfaction.

Incremental change is fast, allowing you to respond to customer needs in real time. Let’s say you have a news website where an increasing number of readers are complaining about readability on mobile devices. Rather than adding this to a potential list of significant redesign efforts, a quick analysis reveals that this affects phones with folding screens. The team can consider how to handle multiple viewport segments on those devices instead of assuming there’s a larger problem.

Just because a change is fast doesn’t mean it can’t have a big impact. You wouldn’t believe how radically different a site can look by simply changing a typeface or how much more engagement you can drive by making minor improvements to a website’s user interface.

An incremental approach is also sustainable. Rather than burning out your teams with a monumental overhaul, short sprints with tangible results keep everyone happy.

Lastly, it meets the needs of CMOs today. Top marketers are increasingly reporting that they are being asked to do more with less and prove ROI all along the way. Incremental change is a way to accomplish both. You can make a tangible impact quickly with a nimble team—testing, measuring, and improving as you go.

Top marketers are increasingly reporting that they are being asked to do more with less and prove ROI all along the way. Incremental change is a way to accomplish both.”

How to get started

1. Prepare your team for success

Leadership might have concerns. We’ve been indoctrinated for so long that a website overhaul is the only way to remove legacy baggage. So, the idea of forgoing a splashy rollout for minor improvements over time can be a tough sell. When advocating for an incremental approach, answer this question: What is the one change we can make that will prove to leadership that this approach can work?

Another common complaint is that implementation or approval will take too long. If that’s the case for your business, rectifying this is the first step. If an organization cannot quickly and confidently change copy, images, or other basic elements of an experience, it signals internal misalignment about how digital experiences are best utilized. The incremental approach doesn’t work if things don’t keep moving incrementally.

2. Choose to be boring

An incremental approach should also include a commitment to stop chasing trends. While a fashionable typeface, parallax scrolling, or aggressively minimalist (or maximalist) website may seem like a good idea at the time, this user experience quickly becomes outdated because it’s tied to what everyone else is doing, not what your brand stands for.

It also behooves teams to pick boring technology. Industry staples are staples for a reason. Sometimes, a start-up disruptor can bring real value to the table, other times it brings a host of problems and minimal support to fix them.

3. Run short cycles with small teams

Keep your core teams small, ideally only 4–6 people. Choose experts over junior staff that may need a lot of oversight, and reduce meetings wherever you can by using asynchronous communication (like Slack) to keep everyone up to date.

You should also time-box every activity—don’t allow any single effort to run indefinitely. Always have a timeline in any hypothesis, and track it ruthlessly.

4. Know what matters and measure it

Even though you’re running with a tight team, that doesn’t mean you shouldn’t be getting input from across the organization. A successful incremental approach will be one where there’s buy-in from all stakeholders, and their needs are balanced with those of their colleagues and the audiences. Have listening sessions with different departments affected by the website. What do they wish was different? What are their pain points? This tactic helps you figure out what will have the biggest impact and how to prioritize.

Once you know what matters to your internal and external stakeholders, commit to measuring it meticulously. You should continually assess what works and what doesn’t and prioritize changes that can be measured against some kind of progress metric. If its impact can’t be measured, don’t do it.

5. Stay simple and move fast

As you begin to design, always reduce and simplify before expanding. Prototype with the lowest effort possible to achieve your goals—finding ways to consolidate changes and reduce redundancies in design can have substantial time-saving benefits over the life of an initiative. This tactic will mean starting with components you already have—ideally ones that exist in a design tool—and current code. The most optimized teams can prototype ideas in staging and run no-regrets experiments in production for quick feedback. They also create reusable assets that can be used for final delivery. Adaptable designs will lend themselves to adaptable prototypes and implementations.

When you reach something that feels good, make it happen. Move to the real product as quickly as you can. If it’s better than what you currently have, ship it.

6. Repeat

The incremental approach comes with a hard truth: a website will never be “done.” At least not as long as you’re in business! There will always be new technologies, customer needs, and insights to incorporate into your digital experiences. Always measure, adapt, and revisit previous changes to make them a little bit better. It’s like Boy Scout’s honor: Leave it a little better than you found it.