Available for Q4 2026
I’m Ajay Kumar. Eight years of SEO, paid, brand and growth for Web3, real estate and fashion — and the marketer who actually builds the automation most consultants only recommend. n8n workflows, lead-qualifying agents, reporting that runs itself.
The idea
Most marketing plans don’t fail. They die in the queue of manual work behind them.
The strategy is sound. Then somebody has to route the leads, chase the follow-up, pull the report and tag the campaign — by hand, every week, because nobody budgeted an engineer to automate it.
I build that layer myself. n8n workflows that move data between your ad accounts, your CRM and your CMS without anyone copying a spreadsheet, and AI agents that qualify a lead, draft a reply or flag an anomaly before a person has to look at it.
I am not a software engineer and I won’t pretend to be one. What I build is the automation layer between the plan and the result — the part most marketing consultants can only tell you that you need.
Where I’ve operated
Each one taught me a different half of the job. None of them let you get away with marketing that ignores how the thing is actually built.
Token launches, exchanges, NFT marketplaces and DeFi protocols. Community-led growth, paid media that survives platform policy, and SEO in a category where almost nobody bothers to do it properly.
Worked with
Developers, brokerages and proptech. High-intent local search, lead quality measured over lead volume, and site-visit economics that still hold up at the booking stage.
Worked with
D2C labels and boutique retail. Brand-first positioning, creative that earns the scroll, and lifecycle work that turns a single purchase into a wardrobe.
Worked with
What I do
See it run
This is the shape of workflow that actually ships: a lead moves through qualification, CRM and follow-up without anyone touching a spreadsheet.
Selected work
One from each market. Full write-ups — including what didn’t work — on the case studies page.
A layer-2 protocol had a loud X following and 90,000 Telegram members, with no idea which of them were real. We tied X and Telegram activity to on-chain wallets with an agent-run qualification pipeline, killed the bot-farmed airdrop tiers, and rebuilt the launch funnel around proof-of-use rather than proof-of-presence.
Leading marketing for a luxury developer with nine live residences, we built one brand that buyers trusted to hold its price. We owned the brand identity and storytelling, invite-only launch previews and experience-centre events, high-net-worth and NRI targeting across premium media and paid channels, editorial content and SEO, a CRM-led concierge journey from first enquiry to handover, and wealth-manager and channel-partner programmes. Every booking fed back into the plan, so spend went to the buyers who mattered, not the most clicks.
Strong acquisition, almost no second purchase, and a lifecycle calendar nobody trusted. We built an n8n workflow that reads what someone actually wears and returns, and an agent that decides the next message and channel — replacing a fixed send schedule with one that runs on real behaviour, unattended.
How it works
Two weeks inside your analytics, ad accounts, search console and CRM. You get a written diagnosis of where growth is actually stuck — not a ninety-slide deck about your industry.
Every opportunity scored by impact, effort and time-to-signal. We agree on the three that matter this quarter and deliberately ignore the rest.
Weekly execution against a public backlog. Fixes, campaigns, pages and content go live in small batches so we learn fast and compound.
Monthly review against one north-star metric. Winners get budget, losers get killed early, and the playbook gets written down for your team.
Questions
I don’t write production software, and I’ll tell you that before you have to ask. What I do build, myself, is the automation layer: n8n workflows that connect your ad accounts, CRM and CMS, and AI agents that qualify leads, draft replies or flag anomalies without a human in the loop. I’m also fluent enough in your stack to hold a real conversation with your engineers when something needs to sit deeper in the product.
Because most marketing budget is spent on work a workflow could do for free. Routing leads, pulling reports, tagging campaigns, chasing follow-up — a marketer who cannot automate that will spend your retainer doing it by hand, or won’t notice it’s costing you at all.
Those are where most of the last eight years went, and they are useful range: one technical market, one data-and-CRM market, one platform-and-creative market. I take work outside them when the shape of the problem is familiar. If your category is genuinely unlike anything I have touched, I will say so on the first call rather than bill you for my learning curve.
Both, depending on what you need. Some clients want a fractional head of growth who sets strategy and manages an in-house team; others want me executing SEO and paid campaigns directly. Most engagements open with a diagnosis sprint and then move into retained execution.
Paid media usually shows signal in two to four weeks. Conversion work compounds from the first shipped test. SEO is slower and more honest: meaningful movement in three to four months, compounding returns from month six. If your timeline and your channel choice do not match, I will say so before we start.
Engagements are scoped by outcome, not by hours. A diagnosis sprint, a monthly retainer and a fractional growth-lead arrangement are the three common shapes. We agree the scope and the number on the first call, before there is a proposal.
Yes. Roughly half of the work is with teams in the US, UK, UAE and Singapore, all remote — Web3 especially tends to be distributed by default. I keep a real overlap window with wherever the core team sits.
Let’s talk
Free, thirty minutes, no deck. You leave with a written view of the three highest-leverage moves for your business — whether or not we end up working together.