Clears the previous record — Q1 2022, $10,399,133 — by 16%, and it is the first quarter we have ever crossed $10M on recurring-heavy business rather than a one-off spike.

H1 2026 is the largest half in company history, and it beats a half that was itself a record.

The growth came from bigger, better deals — not more of them.

Same motion, larger deals, slightly longer to close. That trade is the half.

Renewals are 39% of H1 revenue, and that base now compounds.

Partnership gave back $1.66M — the routes to market shifted rather than shrank.

Every opportunity we took was worth 24% more when it entered, and 65% more by the time it closed.

Fewer meetings, and a larger share of them turned into something real — 70.9% accepted, up from 62.0%, and scheduled in 5.8 days instead of 7.1.

Every source except referral got more valuable. The H1 2026 cohort is still closing, so these are floors, not ceilings.

We are recovering, not setting a record.

Channel wins 84% of what it touches — the highest of any motion, on sixteen deals.

Last year one deal carried the tier. This year it is a habit.

Enterprise agreements went from 11 deals to 18, and from 13.4% of sales to 21%.

The 2,500+ band fell — on three deals, too small a base to read either way.

It is not yet true above 2,500 employees. Three deals is not a trend.

We got bigger and less concentrated at the same time.

More diversity. Bigger deals across all verticals.

Verticals above $1M went from two to nine.

Nine of the thirteen were sourced inside accounts we already had — cross-sell, channel and renewal, not net-new logos.

The biggest thing we sell is now something we own. Omar builds on that.

Cisco point products fell from 55% of revenue to 31% while suites and bundles went from 3% to 23%.

This is the recurring revenue we own outright — not resold, not renewed on somebody else's behalf.

Managed XDR alone added $2.31M of new sales across 36 deals. New this half: managed Splunk, managed NOC firewall, and the student-powered SOC.

Revenue retention is 85.9%, up from 79.5% a year ago and 76.6% two years ago. Only H2 2023 has ever been higher.

Two years ago we added 9.4% across a whole year. We did more than that in six months.

In H1 2026 Channel delivered $1.27M across 16 deals and ten partners, at an 84% win rate — the highest of any motion. Frank takes it from here at 2pm.

Ten partners produced revenue this half. A year ago the number was one.

Only 87 were lost to a competitor. The biggest available win in H2 is not a better pitch — it is not letting deals go cold.

We are not losing on price, and we are rarely losing head to head. We are losing to silence.

Kira's renewal book moved to the renewals team this half, so her total is not comparable to last year's.

Kim competes for the biggest deals in the company, so a single large loss moves her dollar rate more than anyone else's.

This is revenue that closed because somebody else set the meeting.

Charles Thompson sourced more won revenue for other people ($660K) than he closed himself ($385K) — including the second-biggest deal of the half.

Seven things that never show up in a revenue ranking.

One account produced $927,907 in six weeks because the renewals desk set up the cross-sell.

A year ago that first number was measured in hours. The work an AI can be left alone with is now measured in days.

It has stopped being a better way to find what we already knew.

When a technology becomes the thing holding up the economy, it also becomes the thing worth attacking. Derek.

Every one of those is offence. That is where this became real first, and it always is.

The window a defender has to work in has closed by two thirds in twelve months.

Vendors are buying their way into AI security because they cannot build it fast enough. Every one of those is a category that had no budget line eighteen months ago.

Almost everyone believes it. Almost nobody has operationalised it. That gap is a market.

The tooling to find a vulnerability now runs at machine speed. The work to fix one still runs at human speed. Everything in that gap is exposure.

Eighteen months, eight milestones, and a services business that now runs in two countries.

Expand the SIEM plane options so the customer's existing choice is never the reason we cannot serve them.

We are selling faster than we can onboard. These two hires are the fix.

Where the analyst sits is a contractual requirement for some customers, not a preference.

Three surfaces, three questions, three services. Nobody else is answering all three.

This is the second pillar of the plan Omar set out this morning, with dates against it.

There were dozens more, and most of them were bad. That is what a scratchpad is for.