AI Experiments and Algorithm Tweaks

As spring 2023 unfolded, April showered us with new developments in AI and search, stirring both excitement and debate across the digital marketing landscape. Perhaps the most talked-about move: Google began trialing AI-generated search results for select users in the UK​

From my vantage point, this was a call-to-action to double down on content quality and authority. If AI summaries are pulling from top websites, being in that top cohort (the top 10 results) became even more crucial. Interestingly, data from the trial suggested over 93% of these AI overviews included citations linking to the original sources​, and Google even reduced their length by 40% to ensure brevity​. So, for well-ranked sites, it could actually drive clicks by appearing as a cited source – but for lower-ranked content, it might mean invisibility. We advised clients to keep producing authoritative content and structured data (to help Google identify and feature them), but also to start thinking beyond just raw traffic: in an AI-assisted search world,brand recognition and trustmight matter more than ever. If a user sees your brand referenced by the AI answer, that halo effect is valuable, even if they don’t click immediately.

In another major AI-related update, Meta (Facebook’s parent) rolled out a new system to identify and label AI-generated content across its platforms​. They expanded their definition of “manipulated media” to cover more than just deepfake videos – now AI-created images and possibly text could fall under scrutiny too. For social media marketers, this was important. It means if you’re using AI to create say, images of a fictional person endorsing your product, Meta might label it as such. Transparency in advertising was clearly the direction. We see this continuing today with policies requiring disclosure of AI in political ads, for example. In April 2023, savvy brands began preemptively labeling or at least being mindful of AI-generated elements in their content to maintain trust. In our content strategy meetings, we weighed the pros and cons of using AI-generated visuals: “cool factor” vs. potential backlash if audiences felt duped. Often, we opted to use AI for ideation but stick with real photos or clearly artistic illustrations for published posts.

SEO saw another subtle but telling shift when Google’s John Mueller (and documentation updates) suggested that over-focusing on link building might be useless or even counterproductive​. This was an evolution of Google’s long-standing stance: they want sites to earn links naturally, not through schemes. The April twist was Google hinting that their algorithms were discounting manipulative link practices so much that chasing links could be a waste of time. This nudged marketers to allocate efforts elsewhere – think content quality, UX, and on-page improvements, or newer off-site signals like brand mentions and reviews. In practical terms, I advised clients to pivot their SEO efforts. For instance, a software company client had been spending on link outreach; in April we shifted budget to create a series of authoritative whitepapers and webinars (the kind of assets thatattractorganic links and also give E-E-A-T credibility for SEO). We also intensified our focus ontechnical SEOand site speed – areas fully under our control that definitely impact rankings.

E-commerce and web design trends in April also pointed toward more integration and personalization. There was buzz about a “Knowledge Graph CMS” that could generate its own content using AI (a experimental concept, hinting that one day websites might populate themselves dynamically based on a knowledge database). And Shopify, gearing up for their Summer Editions release, leaked some features involving AI (like commerce assistants) which fully materialized later in July. We caught wind of these in April and started planning how to incorporate them for our retail clients – like preparing product feeds and FAQs in a way that an AI assistant could easily use to answer shopper questions.

On social, Twitter’s evolution (on the cusp of rebranding to X) had an interesting footnote: by this time many major advertisers were still shy about returning due to brand safety concerns, which left space for smaller businesses to grab cheaper ad inventory. Meanwhile, Instagram introduced new ad placements (like ads in search results and reminder ads for upcoming events), quietly expanding marketing opportunities for brands to appear in more parts of the app. This “everything is an ad opportunity” trend would only grow – today we even see ads on Instagram’s Explore grid and elsewhere. Back in April 2023, it was becoming clear that omnichannel presence is key; relying on just one platform was risky. Diversification – a bit of Google, a bit of Meta, a dash of LinkedIn, maybe TikTok – was the recipe for stability and reaching different demographics.

One more exciting development: GA4 and Universal Analytics integration finally arrived, allowing some combined reporting as UA’s sunset loomed in July 2023. This helped marketers compare year-over-year data between the two analytics systems more easily during the transition period. We took advantage of that to reassure clients (“see, your traffic in GA4 isn’t actually down, it’s just counted differently; here’s the apples-to-apples view”). It smoothed the switchover and by April’s end most had confidence in reading GA4 reports.

In April, my team’s expertise shone by helping businesses interpret these changes and adapt smoothly. When Google tested AI in search, we didn’t panic; we strategized. We convened an “AI Search Task Force” internally to monitor how often these AI results showed up and which client verticals might be affected. With that intel, we updated SEO game plans – for example, focusing on long-tail keywords where AI summaries were less likely to appear, thus still driving direct clicks to client sites. For social content, we ran audits on any images we’d created or modified with AI (say, using DALL·E for a background or minor element) to ensure compliance with Meta’s new stance. Transparency and authenticity, as always, were the north star.

The playful yet professional tone of April came from balancing cutting-edge tools with time-tested marketing principles. Yes, we played with AI – generating a few experimental blog intro paragraphs with ChatGPT here, a product image variant with Midjourney there – but always with a critical eye before using it in production. And we communicated openly with our clients: “This infographic was partially AI-assisted” or “We used an AI draft to speed up our work on your article, then thoroughly edited it.” These candid updates built trust and positioned us as innovators who don’t sacrifice quality.

April 2023 demonstrated the ever-evolving nature of our field – where algorithms can pivot, AI can rewrite the rules, and user expectations constantly shift. For businesses, navigating this could be daunting. But by partnering with experts who live and breathe these trends, even small companies took advantage of the changes. Whether it was adjusting an SEO tactic, trying a new ad placement, or experimenting with AI content (responsibly), those willing to adapt found fresh opportunities. The core lesson we championed: stay curious and agile. In a month where the very interface of Google began to transform, curiosity is what kept marketing fun and effective. And agility – well, that’s what turned these April showers of updates into May flowers (and results) for our clients.