Not long ago, the question was “will AI replace marketers?” Today, companies are asking something more practical: how do you build a team where AI takes over the repetitive work and people do what the model can’t do for them? There is no shortage of forecasts and bold announcements. First-hand accounts are much rarer: descriptions of what actually changed inside a team and what hiring looks like now.
So we asked marketing leaders how AI has affected the way they build teams and hire.
We asked which skills have become more important and which have lost value, what they look for in candidates, how they upskill their existing people, and what mistakes they see in companies restructuring marketing around AI.
Our respondents include SEO experts, growth and strategy advisors, marketing directors, agency heads and founders. Some have handed all of their marketing to AI agents, while others are doubling down on specialists. They don’t all agree, and that is what makes this collection interesting. Below you’ll find their answers, followed by conclusions you can apply to your own hiring.
AI has made raw production speed less valuable as a hiring differentiator. I now care much more about judgement, customer understanding, research ability, and whether someone can spot when an AI-generated answer is confidently wrong. I also look for marketers who can break a problem into a strong brief before touching the tool. When upskilling a team, practical use cases work better than generic AI training. The restructuring mistake I would avoid is assuming faster execution means you need less expertise. In practice, AI makes good judgement more important.
Which skills have become more important, and which have lost value?
In my opinion, the rise of AI has made skills like diagnosing technical SEO, understanding search intent, and identifying keyword cannibalization even more relevant. In my team, we use AI for things like content audits, keyword analysis and even suggestions for metadata. But still, having someone to understand our websites’ technical setup, find the root of an issue that impacts visibility and validate AI’s suggestions is still important. For example, on a recent e-Commerce project, we grouped more than 1,200 products into clusters of keywords. This allowed us to minimize competition between products (many being variants of the same product) and improve organic visibility and sales. Overall, I’d say repetitive tasks are now less relevant, but technical knowledge and the ability to pinpoint areas of improvement that will have the biggest impact on visibility (and therefore on revenue) is more relevant than ever.
How do you upskill your existing team for working with AI?
We upskill our team by converting repeatable tasks into AI workflows (not everyone writing different prompts each time). For example, I created this e-commerce SEO Skills Pack for Claude. This includes hreflang generation (super helpful when dealing with large e-commerce sites and multiple localized versions), content audits, competitor-gap analysis, metadata optimization, etc. This works because Google’s international SEO guidance reminds us how important the nitty-gritty is when dealing with country & language-targeted pages. Using this Skills Pack means Claude has instructions for each task. We want AI to help with repetitive tasks and provide a level of consistency. However, our team still always checks outputs for relevancy to clients’ sites. The idea is to create systems that distill our subject-matter expertise into workflows for AI without removing human-in-the-loop that makes SEO recommendations worthwhile.
I will start by saying something probably unexpected in today’s landscape about AI: AI is not the only tool, AI is not (yet) guiding employment decisions in the way media is trying to portray it. so in that context:
1. I am searching for my team based on marketing skills, rather than AI skills. The basic marketing skills still outweigh the AI skills. Yes, they need to know how to use it, but more importantly i’m looking for people who know how to channel it in the right direction rather than believing it blindly.
2. I place less value on their previous experience and more value on their skills, asking factual question, scenarios for interviews, rather than generic based experiences.
3. Multi-disciplinarity. Most of the times companies look to hire the tech marketer, or the hospitality marketer. I’m looking for the chameleon. Their most important skill today is adaptability. Not just because of the volatile environment, but also because customer acquisition is not straight forward anymore, it’s a more complex journey than ever before.
4. Hands on experience. That’s the best teacher. Use it, challenge it, get it wrong, fix it and learn from it.
5. this one is my favorite. Most of the companies I’ve worked for during this time they go about AI in marketing as they do with Social Media: try to wing it. One person cannot be marketer, photographer, designer, content director and agency in the same time that a team would do all these, even with the help of AI. Yes, AI is helping automate a lot of our previous challenges, but these are specific different qualifications that needs different functions. For example: i can be a great brand manager with a very detailed idea about how i want to communicate my brand, but my eye for photography will not be trained to spot the details of the b-roll in order to bring that idea to life. So we still need specialists to discern what AI is giving us.
AI has had a massive impact on marketing at my company. We don’t even have a marketing department anymore, and I don’t plan to build one anytime soon.
I’ve handed off all marketing functions to agents I built myself. In my view, AI has actually hit marketing harder than development. All the routine work a marketing department used to do is just gone. No need to spend time analyzing marketing campaigns, researching competitors, evaluating marketing tools, or putting together reports. An AI agent now does all of that in a couple of hours without any drop in quality.
I always wrote all the “authored” content myself, the marketing team just fact-checked it, and now AI does that fact-checking. Writing articles from scratch is still a core skill for me personally.
What I’ll look for when I do build a team again: not just someone who knows the industry, but someone who can actually automate the work. For me now, a marketer is a process manager, not an executor. I know I’ll need someone to manage and check the work of my agents, someone who understands the specifics well enough to correct them when needed.
I think that’s where the future is, because you simply can’t compare someone spending a week manually collecting data to someone who sets up an agent and gets it done in a couple of hours. It’s not competitive anymore.
On upskilling: it happened somewhat on its own. Part of the team started actively using AI in their daily work, and that alone gave us a real boost in results without losing quality. The people who clearly wanted to keep working the old way, unfortunately, we parted ways. The pro-AI part of the team fully backed the idea that going back to the old way wasn’t an option, and time proved us completely right.
One mistake I’ve seen: your team actually needs to understand AI, not just absorb its possibilities through LinkedIn posts. Since we’re a software startup, most of our team picks up new things easily, so AI was with us from day one, and adoption was smooth. But if a team is far from tech, you can’t just fire everyone and replace them with AI like a lot of people claim you can. Start with one business process, look at the results, and move forward step by step. Until your team understands AI the way they understand MS Word, nothing is going to work. So either commit to the deep dive, or don’t bother.
Group trainings on AI tools produced almost nothing in the rooms I run. People sat through the demo, nodded, and went back to doing the work the same way. What moves them is one small assignment with a deadline.
Take your weekly email or your ad copy, rebuild it with AI by Friday, come back and tell me what broke. Momentum first, mastery later.
I’ve worked with thousands of entrepreneurs through group consulting, and the ones pulling ahead aren’t the ones with the longest tool list. They can hold strategy, the build, and the creative in one head at the same time. That hybrid shape is the floor now, not a bonus. Teaching a strong writer to read a dashboard and steer a model is faster and cheaper than stacking three specialists who hand work across a wall.
So when I’m assessing someone, I want to see the conversation they had with the model. Show me the bad first output and the three corrections you made. That tells me whether you can interpret a signal and decide what to do with it.
The mistake I’ve watched happen is redrawing the org chart before anyone’s workflow changes. New titles, same process. I’d take three running experiments with measurable output over a reorg deck every time.
How do you upskill your existing team for working with AI? What has worked and what hasn’t?
Upskilling your team around AI starts with practice and feedback. Employees need to understand where they can save time with AI, when results aren’t accurate, and where human judgement is needed. And as AI continues to develop, training needs to be an ongoing conversation so employees can continually learn how AI can support them with their ever-evolving responsibilities. At LaGrande, we have a 60- and 90-day feedback component of the onboarding process to allow our new hires the opportunity to reflect on their progress and areas of opportunity. Similar to our careers process, upskilling your team around AI looks like having continual discussions around their work, areas of opportunity and confidence-building conversations to ensure employees are comfortable and confident with how to use AI responsibly.
What’s one mistake you’ve seen companies make when restructuring marketing teams around AI?
As companies continue to restructure around AI, it’s important to avoid making the mistake of focusing on getting things done faster instead of focusing on quality. Getting things done faster only becomes an issue down the line when others have to use their own time to fix errors or try to give some substance to empty results. It’s exactly why we value clear communication. That means a prompt response from us or regular, weekly meetings with each of our clients. Clear guidelines around quality control and checkpoints are key. We always ask: Who is looking over AI-based work to ensure it’s useful, accurate, and ready to pass along to the next phase? Similar to how we approach client interactions, employees can leverage AI to get things done faster, provided they’re making sure they’ve done things well.
The thing is, you can’t go around saying things like “AI saves 30 percent of our time, so 30 percent of the team has to go”. Yes, you do save time, but AI isn’t something you should follow blindly. Using it as a tool, but taking the output with a grain of salt. The time you’ve saved goes into fact-checking, editing and client review.
So I think learning to use AI to your advantage instead of thinking it’ll replace you, is the right approach. If you know how to prompt it, you’re already ahead of everyone else.
When I hire an SEO writer now I set a short task, take a live topic and show me the search intent and competitor research before writing a line, because AI can’t do that thinking for them. Stringing sentences together fast used to matter and it doesn’t anymore, most first drafts come from AI now anyway. What’s gained value is editing someone else’s draft hard, cutting whole sections and knowing why. It’s changed team shape too, I no longer hire someone purely for first-draft copy, one editor now directs the AI drafts instead.
The eight experts differ on plenty, but a few themes come up in almost every answer.
Judgment over speed. Production speed is no longer an advantage, because any AI tool delivers it. What matters is the ability to evaluate the output: spotting an answer that is confidently wrong, validating a recommendation, and identifying the actions that will actually move revenue. Several respondents point to quality control and to the question of who checks AI-generated work before it moves on.
Marketing fundamentals before tool fluency. An SEO expert and a strategy advisor say something similar: technical knowledge and classic marketing skills now matter more than AI proficiency alone, because they are what let people steer the tool in the right direction.
A split over team shape. Opinions differ here. Some argue you still need specialists who can recognize a good result in their own field. Others prefer people who combine strategy, execution and creative, or a “process manager” who supervises agents. One respondent has dropped the marketing department altogether.
Upskilling through practice. What works best is turning repeatable tasks into standard workflows instead of everyone writing new prompts each time, giving small assignments with a deadline, and solving real problems. Generic tool training produces little.
The most common mistake. Companies change the org chart before the way of working changes, or assume that faster execution means less expertise is needed. Our respondents advise starting with one process, checking the results, and scaling from there.
Takeaway for hiring: don’t ask whether a candidate uses AI. Ask how they check and correct its work.
