I've been checking the Google Search Console numbers obsessively for the past two months. Not because I love staring at dashboards (well, maybe a bit — I did build a dashboard platform after all), but because the data is telling me something I didn't expect.
This month, Nydar hit 9,638 search impressions. In February, we were barely scraping 300 a day. By mid-March, we're consistently hitting 600+. That's not a hockey stick, but for a solo-built trading platform that's been live for a couple of months, I'll take it.
What's more interesting than the number itself is what people are searching for when they find us.
The queries tell a story
The biggest surprise has been order flow trading. Our learn article on it is sitting at position 7.3 in Google with 136 impressions this month. That's page one for a competitive term. People searching that aren't beginners asking "what is a stock" — they're traders who already know what a candlestick chart is and want to go deeper. They want to understand tape reading, delta divergence, and how institutional orders move through the book before they show up on your chart.
We built the order flow suite because I used these tools at Cowen and couldn't find anything decent outside of a Bloomberg terminal. The institutional platforms charge five figures a year for order flow visualisation. The retail ones give you a basic order book and call it a day. There's a massive gap between "here's the top 5 bids and asks" and "here's how liquidity is actually being consumed across price levels in real time." That gap is where footprint charts, volume delta, and aggregated book depth live.
Turns out a lot of other people were looking for the same thing. The search data confirms it — traders want institutional-grade tools without the institutional price tag. Not a simplified version. The actual tools.
Market microstructure is another one sitting at position 11.5, just off page one, with 54 impressions. That's a niche topic — bid-ask dynamics, latency, maker-taker models, how your order actually gets filled and why you sometimes get worse execution than you expected. The fact that people are finding our content on it tells me there's a real gap in plain-English explanations of how markets actually work under the hood. Most educational content stops at "supply and demand" and never gets into the mechanics. We went deeper because the mechanics are where the edge is.
What I didn't expect
Custom indicators and Pine Script — 40 impressions at position 11.6. I built the Pine Script engine because I personally wanted to run custom scripts against live data without paying for a premium charting subscription. Didn't think many people would search for it specifically, but apparently there's a crowd of traders who've outgrown the built-in indicators on their current platform and want to write their own.
That makes sense when you think about it. Anyone serious about technical analysis eventually hits the ceiling of what RSI and MACD can tell you. They start combining indicators, building composite signals, testing ideas. And then they realise they need somewhere to actually run those ideas against live data without writing a full trading bot from scratch. That's what the Pine Script engine does — it bridges the gap between "I have an idea" and "I can see if this idea works on today's market."
The AI signals page is pulling 100 impressions at position 10.0. That's interesting because AI trading signals is a space absolutely drowning in scams and snake oil. Every other ad on social media is some bloke promising his AI bot will turn £500 into £50,000 by next Tuesday. The fact that our honest breakdown of how the models work is ranking suggests Google is rewarding transparency over hype. We're not promising "guaranteed returns" — we're showing you the XGBoost confidence scores and letting you decide. You can see the model's reasoning, its historical accuracy, and where it gets things wrong.
We published a piece on meta-labelling a while back that explains how we filter out low-confidence signals before they ever reach you. That post didn't get much organic traction, but the people who found it stayed on the page. Quality over quantity — which is the whole philosophy, really.
The CTR problem (and what we're doing about it)
Here's the uncomfortable number: 21 clicks from 9,638 impressions. That's a 0.2% click-through rate. Brutal.
But context matters. Most of those impressions are from pages sitting at positions 15-60 in search results. Nobody clicks on page 6 of Google. The pages that are on page one — the homepage, the order flow article, the AI signals page — have much better CTR. The homepage alone has a 29% click-through rate at position 4.4.
The fix isn't to write more content. It's to make the existing content rank higher. We've been rewriting title tags and meta descriptions across the site — the kind of thankless work that doesn't make for exciting screenshots but moves the needle over weeks. When someone searches "order flow trading" and sees our result on page one, the title and description need to make them click instead of scrolling past.
We've also been dealing with some cannibalization issues — multiple pages competing for the same search term. The glossary is comprehensive (200 terms), and sometimes a glossary entry like fear and greed index competes with the full blog post explaining the same concept. We're sorting that with canonical tags and internal linking rather than deleting content.
What the data is shaping
Nothing dramatic. We're not pivoting based on search queries. But it is shaping priorities.
The pages sitting just outside page one — market microstructure, footprint charts, correlation matrices, funding rate and open interest — those are the ones getting attention right now. More internal links, deeper content, better explanations. The organic traffic will follow if the content deserves it. If it doesn't, at least users who do land on those pages are getting something genuinely useful rather than thin SEO bait.
I've also been spending time on things you can't see in search data. The automated trading bot running internally has been consistently profitable over the past few weeks. It's not public yet, and I'm not going to rush it out until the backtesting is thorough enough that I'd trust it with real size. Which I already do, quietly, in small position sizes. The current strategy uses a trailing stop system that we've been grid-searching to find the optimal parameters — the kind of work that takes weeks of data to validate properly.
Paper trading remains the core of what Nydar does. Everything we build gets tested there first, and it's what most users interact with. The bot, the signals, the order flow tools — they all feed into the paper trading engine before anything touches real money.
A side project is paying the bills
I'm also building something on the side — a separate product that generates revenue from day one. It made $30 before 7am this morning, and it's still early. That matters for Nydar because it removes the pressure to monetise prematurely.
Most indie trading platforms die the same way. They launch free, get some traction, panic about server costs, then slap a paywall on everything and watch their users disappear. Or worse, they start selling their users' order flow data to market makers. That's not happening here.
Having a separate revenue stream means Nydar can stay generous with what's free. It means I can take three months to build the premium features properly instead of shipping a half-baked subscription tier because the AWS bill is due. No "free trial ends in 3 days" pop-ups. No artificial limits on features that cost me nothing to serve.
That's a luxury most indie projects don't have, and I'm not going to waste it by shipping half-baked features just to hit a revenue target.
The content machine
One thing I'm genuinely proud of is the depth of educational content we've built. 42 learn articles covering everything from scalping strategies to liquidation cascades to iceberg orders and spoofing detection. 200 glossary terms, most of them linking through to deeper learn articles and help pages. 33 blog posts in two months.
None of it is generated filler. Every learn article is written by someone who's actually used these concepts on a trading desk. When I write about order flow, it's because I've sat in front of a Bloomberg terminal watching institutional orders sweep through the book. When I write about position sizing, it's because I've seen what happens when people don't do it.
The content strategy isn't "rank for keywords." It's "explain things properly, and the rankings will follow." So far, that's working. The pages that rank highest are the ones with the most substance, not the ones with the most keyword density.
Taking a breather
I'm off tomorrow for St Patrick's Day. The irony of an Irishman building a trading platform and taking the one day a year the entire world pretends to be Irish as a day off isn't lost on me. Markets are closed anyway for the weekend, so the timing works out — no FOMO about missing a trade while I'm three pints deep in a pub somewhere.
The platform will keep running — it managed a full week without me in hospital, so it can handle a bank holiday. The bot will keep trading crypto (those markets never sleep, even when the Irish do). The data feeds will keep flowing. The API quota management will keep doing its thing, backing off outside market hours and spinning back up when the bell rings.
When I'm back on Wednesday, the focus is on pushing those near-page-one pages over the line. Market microstructure, footprint charts, custom indicators — all sitting between positions 11 and 13. A few well-placed internal links and some content improvements should be enough to nudge them onto page one. There's also work to do on options flow — it's at position 19 with 58 impressions, which means people are searching for it and Google thinks we're relevant. We just need to prove we deserve a higher spot.
The numbers
For anyone tracking along:
- 9,638 search impressions this month (up from ~5,700 last month, ~2,800 the month before)
- 21 clicks (low, but trending up — was 3 a month ago)
- Average position improving — from 71.4 to 25.5 over the past 28 days
- 33 blog posts published since January
- 200 glossary terms — all interlinked with learn articles and help pages
- 42 learn articles covering every major trading concept
- 63 help pages — one for every widget and feature
It's still early. We're nowhere near profitable from the platform itself, the click numbers are still small, and there's a lot of work to do on CTR. But the trajectory is in the right direction. The content is ranking, the platform works, people are finding us through genuine searches for topics we actually know something about. That's the foundation.
Everything else gets built on top of it.
Curious what all the fuss is about? Try the dashboard — no sign-up required, start paper trading in seconds.
Originally published at Nydar. Nydar is a free trading platform with AI-powered signals and analysis.
Top comments (0)