Why “Best Practices” in Digital Strategy Age Faster in the AI Era 

For enterprise organizations, one of the biggest strategic risks right now is not failing to adopt new technology. It is continuing to operate from assumptions that no longer reflect how users behave, how AI systems surface information, or how digital trust is established. Practices that once felt stable across SEO, UX, accessibility, personalization, and content strategy are now shifting under the pressure of AI mediated discovery, evolving regulations, and changing expectations around transparency and credibility.

The challenge is not that previous best practices were wrong. Many still matter deeply. The challenge is that in a fast moving environment, static guidance ages quickly. Organizations that treat digital strategy as a living system rather than a fixed playbook are better positioned to adapt to AI search, changing discoverability patterns, and growing trust expectations.

If you know me, you know I’m a reader. Ask me a question and there is a very real chance I have already read a book about it, highlighted half of it, and dog-eared the pages that made me rethink how I work.

Recently, I read Think Again by Adam Grant, and it nudged me back to something I have felt for a while in digital strategy: the phrase “best practices” can get rigid very quickly. It sounds comforting. Proven. Efficient. Almost permanent. But in a world shaped by AI, changing search behavior, shrinking patience for friction, and rising expectations for trust, many so-called best practices are no longer wrong so much as incomplete. They need to be revisited far more often than we used to admit.

“Best practices” can be useful, but they can also become a trap. The phrase sounds solid and reassuring, as if the thinking has already been done and all that is left is execution. But in fast-moving fields like digital strategy, that mindset can quietly turn into rigidity. What worked well even a year or two ago may now be incomplete because the context has changed: technology evolved, user behavior shifted, platforms changed their rules, and people’s expectations moved with them. 

That is why I have been thinking more carefully about how often we treat established guidance as permanent truth instead of temporary usefulness. In strategy work, confidence is valuable, but curiosity is often more valuable. The strongest practitioners are not the ones who cling hardest to old frameworks. They are the ones willing to re-examine their assumptions, question inherited advice, and update their thinking before the market forces them to. In that sense, good strategy is less about defending what used to work and more about staying mentally flexible enough to notice when it no longer does.

As digital strategists, we are expected to connect disciplines that are often treated separately: UX, UX research, content strategy, accessibility, privacy, SEO, and now GEO. But the reality is that all of them are being reshaped by the same forces. Users are not just visiting websites anymore. They are asking AI tools to summarize, compare, recommend, and decide what is worth their time. That changes what “good” looks like.

1. SEO best practices are no longer enough on their own

For years, digital teams were taught to optimize for rankings, keywords, click-through rate, and organic landing pages. That foundation still matters, but it is no longer the full picture. Google now explicitly advises site owners to think about how their content appears in AI features like AI Overviews and AI Mode, and notes that users are asking more complex questions in these environments.

 

In other words, discovery is becoming more answer-mediated, not just search-result mediated.

That means a modern strategy cannot stop at SEO. It has to consider whether your content is structured, credible, and useful enough to be surfaced, summarized, and trusted in AI-assisted experiences. I still believe in SEO deeply, but increasingly I think in terms of discoverability, retrievability, and answer readiness. Classic optimization is still necessary. It is just no longer sufficient.

2. Publishing more content isnot the same ascreating more value 

One of the most outdated interpretations of content strategy is the idea that scale alone wins. More pages. More keyword variants. More coverage. More production.

That playbook was already weakening, and AI has exposed the weakness even more clearly. Google has said that using AI is not inherently a problem, but content created primarily to manipulate rankings remains a problem. What matters is whether the content is helpful, original, and people-first.

This matters because generic content is now easier than ever to generate and easier than ever to ignore. If an answer can be synthesized by any model in seconds, then commodity content loses strategic value. The content that stands out now tends to include lived expertise, clear perspective, useful examples, real evidence, and substance that goes beyond “what everyone already says.”

The bar has shifted from volume to distinctiveness.

3. Content should not just answer a query. It should support a thinking process

A lot of content strategy advice has historically focused on query matching. Find the question. Answer the question. Keep it tight.

That still matters, but it ignores how people increasingly behave when they use AI-assisted search or research tools. Google says people are using Search more often, asking more complex questions, and exploring a wider range of sources in AI-enhanced experiences. Nielsen Norman Group research similarly shows that generative AI is reshaping search behavior, even as some long-standing habits remain.

The implication for content strategy is big. We should design content not just as an answer, but as a progression. A strong page today often needs a fast summary, clear headings, comparison points, definitions, caveats, proof, and next steps.

That is not bloat. That is support for how people actually build understanding.

Users still scan, but they also go deeper when the topic matters and when the structure earns their attention. NN Group has long found that even highly educated readers want content that is digestible, concise, scannable, and credible.

4. “People have no attention span” is a lazy conclusion

I hear this constantly: nobody reads anymore. People want everything shorter. Make it bite-sized. Simplify it until there is almost nothing left.

But that is not really what the best research says.

NN Group’s work on the attention economy frames the issue more accurately: digital products are competing for limited user attention. The problem is not that people are incapable of depth. It is that attention is scarce, selective, and easily lost when experiences create unnecessary effort.

In practice, that means the better principle is not “make everything shorter.” It is “make value obvious faster.”

People still read when they trust the source, care about the task, and can navigate the content efficiently. NN Group’s findings on long-form content and expert audiences reinforce this. Users scan first, and credibility heavily affects whether they continue.

Good strategy today means designing for fast entry and optional depth at the same time.

5. The website is no longer the whole experience

Another outdated assumption is that the website is the main stage and every other channel exists mostly to drive traffic back to it.

That model feels increasingly incomplete.

The first brand interaction may now happen in a search summary, an AI assistant answer, a forum thread, a social post, a product review, or a third-party comparison before the user ever decides whether your site deserves a click. Google’s guidance on AI features makes this shift impossible to ignore.

This is one reason digital strategy has to become more distributed. Your content model, metadata, knowledge base, author credibility, structured information, and consistency across channels all influence discoverability and trust.

The site still matters enormously, but it is no longer the only container of your digital experience. In many cases, it is now one node in a larger answer ecosystem.

6. Accessibility is not just a compliance checkpoint anymore

A lot of accessibility work still gets framed like a once-a-year audit followed by a remediation backlog. 

That mindset was never ideal, and it feels especially outdated now.

World Wide Web Consortium’s WCAG 2.2 added new and updated criteria that reflect more realistic interaction patterns, including target size, focus visibility, alternatives to dragging, redundant entry, consistent help, and accessible authentication. W3C has also published specific guidance on applying WCAG 2.2 to mobile applications.

The practical takeaway is that accessibility should be treated as interaction resilience, not just compliance hygiene.

If your strategy includes complex forms, mobile patterns, design systems, authentication flows, AI-assisted tools, or account experiences, accessibility cannot live in a silo. It has to be built into how experiences work from the start.

7. Less friction is not always better

One of the most persistent UX beliefs is that every extra click, every pause, every review step is bad.

But as AI becomes more embedded in digital experiences, that principle needs more nuance.

Some friction is wasteful. Some friction is protective.

National Institute of Standards and Technology’s Generative AI Profile emphasizes risks such as inaccurate output, harmful content, privacy impacts, and broader trustworthiness concerns. NN Group has also found that users struggle to error-check AI outputs, partly because polished responses create a false sense of confidence and verification takes real effort.

This is why good digital strategy cannot optimize only for speed. It also has to optimize for confidence.

In higher-stakes moments such as payments, health, identity, legal information, or AI-generated recommendations, a review step, a disclosure, or a confirmation can be a feature, not a flaw.

The new question is not “how do we remove all friction?” It is “where do we remove friction, and where do we design for trust?”

8. UX research should use AI, but should not outsource judgment to it

AI has absolutely changed the research workflow. It can help summarize notes, cluster themes, generate hypotheses, speed up desk research, and support analysis. Ignoring that would be impractical.

But overtrusting it is just as risky.

NN Group has been very clear on this: synthetic users and AI-moderated approaches can support research in limited ways, but they do not replace the depth, empathy, and nuance of real human research. Their recent work also warns that methodological blind spots in tools become even more concerning when AI starts helping plan and analyze studies.

To me, this is one of the clearest examples of a best practice needing revision.

The updated version is not “never use AI in research.” It is “use AI as acceleration, not authority.”

Real participant research should still anchor meaningful decisions. AI can help us move faster, but it should not become a shortcut around actual user understanding.

9. Personalization needs governance, not just relevance

For a long time, personalization was treated as an obvious good. If we can make the experience more relevant, why would we not?

But AI-driven personalization introduces different kinds of risk: opacity, creepiness, hidden bias, over-inference, and manipulative experiences that users never knowingly signed up for.

NIST’s guidance on generative AI is useful here because it frames AI systems in terms of trustworthiness, risk management, and governance rather than raw capability alone.

That is the shift I think strategy teams need to internalize.

Personalization should not just ask, “Can we do this?” It should ask, “Should we do this? What data powers it? How visible is it to the user? And what happens if we get it wrong?”

Relevance without explainability is not always good UX. Sometimes it is just polished discomfort.

10. Privacy is much bigger than banners and policies

Cookie banners and privacy policies were never the whole story, but AI has made their insufficiency much more obvious. 

The European Data Protection Board’s 2024 opinion on AI models underscores that organizations must think carefully about when models can be considered anonymous, how legitimate interest applies, and what happens when personal data is used unlawfully in development.

The opinion explicitly frames these questions as part of responsible AI and GDPR compliance. 

For digital strategists, this means privacy can no longer sit off to the side as a legal afterthought. If your organization is using AI-powered search, personalization, chat, summarization, or content generation, privacy questions now touch the full lifecycle: inputs, training data, storage, vendor access, model behavior, outputs, and retention.

A privacy strategy that only shows up in the footer is not a strategy.

11. Trust signals have moved from “nice to have” to strategic infrastructure

In an AI-shaped web, trust signals matter more, not less.

When content is increasingly summarized outside its original context, users and systems both need stronger clues about credibility. Google continues to emphasize helpful, reliable, people-first content, and NN Group’s research on expert audiences reinforces that credibility is not secondary for serious readers. It is central to whether they continue engaging at all.

That means authorship, evidence, dates, methodology, citations, source clarity, and visible expertise are no longer decorative. They are part of the product.

They help users decide whether to trust what they are reading, and they help your organization stand apart from the flood of generic content that now fills the web.

In a world of abundant answers, trust becomes a competitive advantage.

Final Thought 

What Think Again brought back to the surface for me is that strategic maturity is not about clinging to best practices longer than everyone else. It is about knowing when to revisit them.

In digital strategy, “best practice” should never mean fixed doctrine. It should mean “the most useful guidance we have right now, based on current behavior, technology, and risk.”

And right now, the best digital strategies are the ones flexible enough to admit that users have changed, systems have changed, and our frameworks need to change with them.

Not because fundamentals are dead. They are not.

But because fundamentals only stay useful when we keep rethinking how they apply in the world we actually have.

As organizations navigate AI-assisted discovery, evolving trust expectations, accessibility requirements, and changing user behavior, the companies that adapt fastest will not necessarily be the ones chasing every new trend. They will be the ones willing to continuously reassess assumptions, modernize their digital frameworks, and align strategy with how people actually experience the web today.

At RBA, these are the kinds of conversations we help organizations navigate every day, from SEO and GEO strategy to accessibility, AI readiness, content governance, analytics, and digital experience transformation. The goal is not to abandon fundamentals. It is to evolve them thoughtfully so they remain effective in a rapidly changing digital landscape.