Enterprises, banks and public sector organisations trust evidence more than vendor claims. So I start with the buyer's real problem, then decide which markets to win, what to build first and how to price and package it, and build the team that scales it.
Buyers now meet vendors through AI before they meet them in person. At the same time, every vendor claims AI: Gartner estimates only about 130 of the thousands selling "agentic AI" are genuine. For an enterprise, a bank or a government, that means more claims and less certainty. The vendors who win are the ones whose evidence holds up when the buyer, or the buyer's AI, checks.
Lead with something useful: original research, a peer forum, a readiness assessment. It reveals the buyer's real problem and creates the evidence that buyers and their AI tools will find.
For example, at Radware, bot-threat research cited by BBC and TechRadar made the company a go-to voice on bot fraud. At Merkle Science, RegWatch regulatory intelligence became the primary growth engine.
Shape what gets built, how it's packaged and priced, and how it's positioned around what buyers reveal, so the product reads as the answer to their problem, not another AI claim.
For example, most recently at Mozn, a readiness assessment for new counter-fraud rules became a new revenue line, and a rule simulator let banks test changes against historical data before going live.
Independent validation is what separates real capability from agent-washing, so analyst relations and peer advocacy are core to go-to-market.
For example, at Radware, Forrester made intent-based behaviour analysis a requirement in its Wave and named Bot Manager a Leader. Most recently at Mozn, the team earned Chartis AML & KYC Leader status and 25 analyst recognitions in 2026 so far.
Run one operating model across product, sales, marketing and pre-sales, from release management to enablement, and use AI to make the team faster without lowering the bar on accuracy.
For example, most recently at Mozn, that engine supported 2.5× ARR growth in a single year.
The playbook has worked in enterprise AI, financial crime, cybersecurity and crypto compliance. Technology keeps changing. The buyer's need for evidence doesn't, and AI has only raised the bar.
Four stories from my career, with company names left out. Each shows the situation, what we did, what happened, and what I took from it.
This is one of my favourite stories from my career. Banks had one worried question, and answering it properly turned into a business of its own.
When sales has to invent a price for every deal, the problem usually isn't sales. It's the packaging. We rebuilt ours around how customers actually grow.
Many growing companies win deals through heroics: a brilliant salesperson, a founder stepping in, a deck rebuilt the night before. It works, until you need to do it a hundred times. We set out to make winning repeatable.
Buyers who don't understand a category compare on price. Original research and a clear point of view gave regulated buyers a better way to judge, and a reason to trust us.
How automated bots exploited public fear during the COVID-19 infodemic, from amplifying misinformation to enabling fraud.
How sophisticated bots mimic human behaviour to evade detection and commit online fraud, and what that means for detection strategy.
Read on Radware →Engineering research proposing a bio-inspired queen bee algorithm for designing communication networks between robots.
Read the paper →I'd like to use cookies from Google Analytics and LinkedIn to see how people use this site and which companies visit. They're off unless you agree, and you can change your mind any time from the link at the bottom of the page.
Netlify, which hosts the site, also counts visits from its server logs. That uses no cookies and doesn't track you across sites.