Meet Nicolas
I’m Nicolas Lecocq, French, currently based in Switzerland with my wife Maria and our son Alexandre. I’ve been writing code for over fifteen years, mostly living in the WordPress and SaaS worlds. One of my better-known builds along the way was OceanWP, a WordPress theme that grew to more than half a million active installs and passed $1.5M in revenue. These days I’m fully focused on LinkedGrow with Maria, and I spend a strange amount of my time building AI agents for our own marketing, which is how I ended up obsessed with this whole space. I’ve got over fifteen of them running right now, handling everything from LinkedIn posts to outreach emails to Reddit research, and that obsession is honestly the reason LinkedGrow exists at all.
What inspired you to build LinkedGrow?
It started from pure frustration. I was trying to get consistent on LinkedIn and tested pretty much every tool on the market, and every single one wanted somewhere between $39 and $99 a month while also capping how many posts I could generate on top of that. The math never added up, because I knew the real AI bill underneath was closer to $2 or $3 dollars, and the rest was just a tax on people who don’t know how the plumbing works. So I flipped the model entirely. With LinkedGrow you bring your own AI key, you pay your actual usage (which stays tiny for most people), and we charge a small flat platform fee on top. The total ends up being roughly 96% cheaper than the competition, with zero generation caps, which was the whole point of building it.
A handful of things that I think actually matter when you use it day to day. The BYOK model is the obvious one, because it kills the per-generation anxiety that every other tool relies on to make you pay more. Then there’s a voice training layer that takes a small sample of your own LinkedIn posts and teaches the model to sound like you instead of sounding like every other ChatGPT draft floating around the feed. And finally, the product is built specifically for LinkedIn rather than trying to serve ten platforms at once, which means native features like a Reddit-to-LinkedIn idea importer, a hooks library based on what genuinely travels on the platform, a carousel generator, and scheduling that respects timezones and optimal windows. Generic multi-platform tools simply can’t go deep enough on any single network, and it shows.
Why focus specifically on LinkedIn instead of multi-platform tools?
Because LinkedIn is quietly the most underpriced attention in B2B right now. On almost every other major platform, organic reach has collapsed to the point where you basically need ad budget to get noticed, whereas on LinkedIn a single good post can still generate tens of thousands of views with no paid distribution at all. The audience composition is also very different from Instagram or TikTok, the people scrolling LinkedIn have budgets, decision power, and a real reason to follow up on what they read. For founders, freelancers, consultants, and B2B teams, LinkedIn is where pipeline actually starts. Splitting your energy across five social networks might sound like smart diversification, but in practice it usually means being forgettable on all of them instead of memorable on one.
How do you avoid the “robotic AI” feel most content tools suffer from?
The dirty little secret of AI writing tools is that they almost all produce the same output because they feed the model the same generic system prompts. We went the opposite direction from day one. Before you generate anything on LinkedGrow, we ask for a handful of your past posts, your business context, the audience you’re writing for, a preferred tone, and a list of topics or competitors you never want to see mentioned. That personal context layer is what actually transforms the output from a lifeless ChatGPT draft into something that feels like you wrote it. When someone complains that their AI posts sound robotic, nine times out of ten it’s because they haven’t given the AI anything personal to work with in the first place. If you feed the model thin inputs, you get thin outputs on the other end, and that hasn’t changed since the early days of generative models.
What AI models work best with LinkedGrow’s BYOK system?
It depends heavily on what kind of writing you’re doing and what you personally care about. Claude tends to be the strongest writer in my experience, especially on posts where voice, nuance, and structure matter. GPT is faster and great for high-volume brainstorming when you want a lot of angles quickly. Gemini is particularly strong for research-heavy posts where you want to weave in recent information, and Grok has a cheekier, more irreverent tone that suits certain personal brands. The advantage of BYOK is that you’re not stuck with one provider forever, you can test different models on the same prompt and keep whichever one feels most like you. For my own content I tend to run Claude for drafting and bring in Gemini whenever I need something with current data.
What’s the biggest misconception around AI-generated LinkedIn posts?
The biggest misconception is that AI content automatically equals slop, when the reality is most of the AI slop you see on LinkedIn comes from humans who pasted a lazy prompt into ChatGPT and hit publish on the first draft without editing a single word. AI is not the actual problem, the workflow behind the AI is. When someone uses AI as a real drafting partner, edits ruthlessly, adds their own stories, opinions, and data points, and reworks the hook until it genuinely lands, the final post is often sharper than what the same person would have produced alone at midnight after a long day of real work. The people loudly complaining that AI is ruining the feed are mostly reacting to low-effort posts, but high-effort AI-assisted writing is already almost impossible to tell apart from great human writing.
What’s your advice for someone trying to grow from zero followers?
My strong advice is to comment before you post, which sounds unintuitive because most people assume growth requires a flood of original content. The fastest compound growth I’ve seen on LinkedIn comes from people who pick 30 to 50 creators in their niche and show up in those comment sections every single day with a real perspective, not “great post” or a generic emoji. Within a month or two you start picking up follow requests from those creators’ audiences, and by the time you do publish your own posts, you’re no longer broadcasting into an empty room. The worst thing a new creator can do is post into silence for six months and conclude that LinkedIn is dead, when actually the platform is wide open for anyone willing to engage with other people first.
How do you see LinkedIn’s algorithm evolving?
The direction is pretty clear if you’ve been paying attention to what LinkedIn has quietly rolled out over the last year or two. Dwell time and comment depth are being weighted more heavily than raw likes, outbound links are being throttled more aggressively every quarter, and posts that are obvious AI clones of viral formats get flagged and suppressed harder than most people realise. Long-form, substantive posts with real opinions and distinctive voices are the content that keeps winning, and I’d bet that over the next year LinkedIn will also start penalising engagement pod behaviour more explicitly, because those coordinated like-and-comment groups are detectable at the data level. Overall the platform is rewarding depth over gaming, which is good news for anyone actually trying to build something real.
Will AI-generated thought leadership become the norm?
Yes, it’s already happening in most industries, and paradoxically I think that’s exactly why a strong personal voice will become the biggest differentiator of the next few years. When everyone has access to tools that can generate passable thought leadership on demand, what stands out is the opinionated, slightly uncomfortable, genuinely personal perspective that an AI won’t default to on its own. AI isn’t going to replace thought leaders, it’s going to flood the middle of the market with competent but forgettable content, and the top tier of voices (the people who take real stances backed by real experience) will matter even more by contrast. The gap between great creators and average ones is about to widen significantly, not shrink.
What’s next on LinkedGrow’s roadmap? Where do you see LinkedGrow in the next few years?
The biggest thing I’m pushing hard on right now is our Engagement page, which is the natural extension of the advice I gave earlier about commenting being the fastest way to grow on LinkedIn. Right now if you want to follow the creators who matter in your niche, you have to hop between LinkedIn tabs, scroll past people you don’t care about, get pulled into algorithmic noise, and eventually forget half the profiles you meant to engage with. The Engagement page solves that problem directly by letting you build your own curated lists of creators and prospects (your target accounts, your competitors, the people whose audiences you want to tap into), and it then pulls all of their recent posts into one clean feed inside LinkedGrow, with the ability to like, react, and comment without ever leaving the dashboard.
On top of that we’ve built an AI comment assistant that can draft a genuinely good comment based on the post content and your personal voice settings, so instead of writing twenty generic replies a day, you can engage meaningfully in a fraction of the time while still sounding like yourself. My vision for the next couple of years is that LinkedGrow becomes the control centre for everything a serious LinkedIn user does, from generating and scheduling content, to tracking analytics, to actively growing their network through targeted engagement. Similar to what Buffer was for multi-platform scheduling a decade ago, but rebuilt around how LinkedIn actually works today and around user-owned AI keys so that no one ever pays a hidden margin on the AI itself.
Did you enjoy the interview? Anything to say to the community?
Really enjoyed it, thanks for having me on. If there’s one idea I’d leave your audience with, it’s that LinkedIn rewards people who show up consistently far more than people who try to publish the perfect post once a month and disappear between attempts. You don’t need a big audience to start, you don’t need a budget, and you certainly don’t need a brand kit or a fancy setup. You need a real perspective, a daily habit, and a willingness to genuinely engage with other people’s work before expecting anyone to engage with yours. If any of this makes sense and you want to try LinkedGrow for yourself, we have a free tier so you can see whether the workflow fits before spending a cent.
