You’ve spoken recently on AI, efficiency, and productivity – how do you see AI-driven tools transforming digital marketing strategies? What opportunities do these technologies offer?
There are many things in marketing that have changed. Here is one example. Think of some of the questions that marketers ask
Will that headline work?
Will this topic resonate?
Will they click this ad?
Will they register for this webinar?
Will this pricing page convince them?
Before AI: Let’s test it.
After AI: Let’s give it to our AI persona and ask for feedback.
AI takes some of the guesswork out of marketing. It makes it possible to get quick feedback on every marketing effort before it launches. That’s because you can create a synthetic member of your target audience and talk to it first. So AI reduces risk and improves targeting.
The trick is to first make an AI persona and then keep it close by. Share it with your team. Or make it into a custom ChatGPT that responds to all questions from your prospects point of view.
AI and Privacy
Are there data privacy challenges that marketers should be marketers aware of?
A virtuous marketer cares about the data of their audience and avoids creating or storing PII (personally identifiable information) whenever possible.
But some marketers are excited to use real data to train their AI persona. It sounds great, right?
- Give the AI all of your customer reviews. Now the AI has the actual VOC (voice of customer)
- Give the AI transcripts of your sales calls. Now the AI has all your actual prospects questions.
- Give the AI all of your CRM data. Now the AI has the customer data and knows your target.
Of course, all of these are PII. And when you use it to train an AI, you are handing it over to a Big Tech company that will never give it back. It cannot be recovered. It cannot be deleted.
I don’t believe that the data is now in AI’s long term memory. I don’t think it’s possible for you (or a competitor or a hacker) to get the data back out again. But it’s still a violation of the ethos of the privacy-first marketer.
Don’t upload personal or confidential information to an AI unless you are confident that you have a secure and legal technical environment where your AI and data are hosted.
Probably, you don’t.
Banks and healthcare companies do have these specialized server setups. They need them to use AI to detect fraud and cancer. But that’s predictive AI, not generative AI. It’s not what we’re doing here as marketers. That’s not ChatGPT, Gemini or Claude.
Compliance Fatigue Among Clients
With evolving regulations like GDPR and CCPA, have you noticed any signs of compliance fatigue among your clients? How do you help them balance strict data privacy requirements with the need for innovative web solutions?
The challenge for marketing leaders is real.
On one hand… they want to get the most out of their efforts. If they’re investing time and money building a TikTok channel, they want to measure results. Adding the TikTok tracking pixel (sometimes called the “TikTok web beacon”) can help measure traffic and conversion from TikTok ads. Let’s add it and get some performance data.
On the other hand… they worry that this TikTok tracker wasn’t added properly. Is it in Google Tag Manager? What’s the trigger? What are the proper settings for the CMP (consent management platform)? They use good tools and get advice from experts.
Then two things happen at regular intervals:
- They are preparing a report and notice that they are missing data and can’t fully report on the results of their good work. The lack of data is costing them credibility.
- They see a headline or hear from a friend that someone is getting sued for making their visitor data available to the Chinese Communist Party (TikTok’s parent company stores data in Chinese data centers). The collection of data may cost them their job.
All marketing leaders with this tension, pulled between these two risks. One low and continuous (missing data and poor attribution), the other high but remote (lawsuits and enforcement action).
Most of us got into marketing because we had a spark of creative passion.
Then one day, we wake up finding ourselves studying international law. It’s weird.
Integrating Performance, Accessibility, and Privacy
Your agency emphasizes doing the right thing by focusing on factors like optimized page speed and enhanced accessibility. How do these best practices support your overall commitment to ethical data handling and privacy protection?
The key is to align with the client’s values and goals.
The conversation starts with the business goals. Websites are part of marketing and marketing has a job to do. An accessible, privacy compliant website is a failure if it doesn’t generate leads. So if the client doesn’t see that you are aligned on bottom-line business outcomes, they’ll go somewhere else.
The conversation is also about values. Some of the many little decisions involved in web design are driven by the beliefs of the client and what they see as important. For example, popup windows might be effective, but they might annoy the visitor. If I value my visitors’ experience, I choose not to use a popup window.
Accessibility and privacy protection are two of the considerations that align with values. Yes, there are practical reasons to do the right thing (avoid lawsuits) but these requirements are also about what we believe. If we value these things for all visitors, that means accommodating the minority. That’s values alignment.
- Even if a minority of users have a disability and require an accessible website, we make accessible websites.
- Even if a minority of visitors care about the protection of their own privacy, we allow visitors to opt-out of tracking.
Do we care about what’s right for all users? Are we an inclusive brand? Do we put people first? Everyone or just some?
This sounds lofty and idealistic. Really, it’s not a big deal. Doing the right thing for everyone isn’t that hard. And in the end, caring about 100% of your visitors is better for the bottom line.
Contrasting Privacy Perspectives: EU vs. US
In Europe, privacy is framed as a fundamental human right, whereas in the US the approach is often centered on individual choice and data as an asset. How do these differing perspectives influence your approach to web development and your clients’ privacy strategies?
This is a cultural difference between Europe and the US.
Americans have low expectations for privacy. It’s no surprise. Every day, we get privacy policy updates (we ignore them) and letters in the mail informing us of data breaches (we throw them away). So we just assume that it’s too late.
Americans believe that our personal information has already been sold or stolen. We assume that we’re being tracked on our phones. 60% of Americans think their phones are listening to their conversations. Yet …no outrage?
It may be just that trust (and distrust) manifest themselves differently in these different cultures. Globally, trust is down across all institutions, governments, businesses, NGOs and Media (source). But people generally trust businesses more than governments.
If you trust businesses more than your government, you may prefer to leave data protection in the hands of companies, rather than through new laws. If the government is less competent than businesses, why would you want their protection? Why make a law?
Apparently, Americans prefer privacy policies to government policy.
Adapting UX for an AI-Driven Web
As AI agents increasingly browse the web and influence user interactions, how is your agency adapting its UX design strategies?
AI is suddenly a new and important type of visitor. We are in the era of “agentic AI” where AI agents are searching, clicking, landing, reading and reporting back to their user (these are companies that can help you) or taking action directly (I bought it for you and it’s on the way).
So we are building AI-friendly websites. They work for AI users who are using virtual browsers, keyboards and mice. This means a few new considerations.
- Add detail to your key pages
Make your main service pages very detailed pages. Explain everything about the work you do, how you do it, who you do it for. AIs don’t care about clever copy or brand messaging. They don’t mind long pages with blocky paragraphs.
- Use simple code
Avoid anything that uses Javascript to display important information. If there’s an answer that can’t be read until the user interacts with the slider, or carousel or accordions, pull it out and just put it in the content. The curious human might click to expand that FAQ question, but the AI agent might not.
- Create an AI training page
Using all of the schema markup you can think of, create a new page that is specifically constructed to add information about your business to the AI’s training data. It’s one more way to make it more likely that the AI’s will recommend your brand. You can simply link to this from your footer, next to the links to your sitemap and privacy policy.
- “How did you hear about us?”
If this question is on your contact form, add “AI / ChatGPT” as an option.
Humans are using the web in new ways. So brands need to look at their websites in new ways. Think of all the website requirements, from page speed to mobile responsiveness. Now think about each in the context of AI agent visitors. Some requirements are more important, others are less important.
This chart is from our guide on building AI-friendly websites.
We suggest that privacy law compliance is less important for AI visitors, because AIs don’t have human rights. The laws don’t apply. Besides, AI agents don’t accept cookies. Even if they did, it would be useless to track them. AIs have no personal information. Why track an AI agent if you can’t see who the user they’re working for?