Canada has already decided it wants to build more of what it needs at home.
You can see that in the Defence Industrial Strategy, the Defence Investment Agency, and the growing list of programs meant to connect defence and security problems with Canadian companies. Those are useful steps. They suggest Ottawa understands that sovereignty, resilience, readiness and productivity now belong in the same conversation.
Budget 2026 is the chance to tie those pieces together. What is still missing is a procurement path that lets government get capability into the field faster and lets Canadian-controlled firms grow.
In the pre-budget work I have been doing, I keep coming back to one question: what happens after the pilot?
That is where the next round of policy has to focus.
Pilots are useful. So are test ranges and innovation programs. But a pilot is not a purchase. A test is not a deployment. An “ecosystem” only becomes real when there is demand behind it — contracts, repeat customers, and someone with a budget who can buy.
A pilot can show that a technology works. It rarely shows that Canada has a way to turn that technology into something operators can use, or to turn the company behind it into a supplier that lasts.
Eliot Pence made a good point recently about Canada’s Build-Partner-Buy framework. It should work as a ranking, not a menu. Build comes first. If Canada is serious about sovereign capability, new money cannot keep flowing through the same old channels and produce the same supplier base.
The hard part is making “Build” mean something on the ground.
It is not enough to open more doors for innovators. It is not enough to validate a technology and then send the company back into a procurement system built for incumbents. And it is not enough to praise Canadian innovation while Canadian-controlled SMEs stay too small to prime a contract, too short of capital to wait years for a decision, and too often stuck as subcontractors under larger firms.
If Canada wants to build, it has to bridge the gap between pilot and deployment.
We are not starting from scratch. The Defence Industrial Strategy, the Defence Investment Agency, DDI, IDEaS, DIANA, DISH and Borealis all show that government knows it needs to connect defence problems with Canadian innovation. That work matters.
Budget 2026 can build on it by fixing the next problem: linking pilots, test environments and innovation programs to the people who actually hold procurement authority, budgets and contracts.
That is not a knock on what is already underway. It is the next step if those efforts are going to work.
For a lot of defence and dual-use technology, pilot money might cover a trial, a prototype, some limited integration or an early operational look. It usually does not cover deployment, production readiness, ongoing support, supply-chain work, certification, training, export prep, or the working capital a company needs to become a real supplier.
That matters because companies do not scale on encouragement. They scale on customers, contracts and repeat demand.
So Budget 2026 should treat procurement itself as the main tool for building Canadian industrial capacity.
First, set a clear target: at least 25 percent of eligible federal procurement should be reserved for Canadian-controlled SMEs.
A 25 percent set-aside would not mean lower standards. It would mean admitting that procurement rules shape markets. If government wants Canadian-controlled firms to grow, it has to create demand those firms can actually win. Without a target, departments will keep going back to the usual suppliers, the usual contract structures and the usual risk profile.
A set-aside would force a better question: where can a Canadian SME lead?
That is the right question. Too often SMEs are treated as a source of ideas, not as the industrial result we say we want. They get asked to demonstrate, advise, partner or subcontract. They rarely get to prime. If Canadian-controlled firms never own the customer relationship, never lead delivery, never keep the strategic intellectual property and never manage scale, they will not become the companies Canada claims it wants.
Second, create a real middle lane between small pilots and major procurements.
Canada needs something between a small pilot and a major program: rapid pilots and demonstrations up to $5 million; operational trials and limited deployments between $5 million and $25 million; and scale-up buys between $25 million and $100 million where Canadian-controlled SMEs can prime, co-prime or lead a consortium.
That would let officials manage risk without moving at a crawl. Government could test early, learn quickly, buy what works and stop what does not. Canadian firms would have a believable path from proof of concept to production, deployment and export.
In areas that move fast — counter-UAS, cyber defence, AI-enabled decision support, sensing, autonomy, secure communications, Arctic surveillance — the issue is often not whether a perfect product exists on day one. It is whether government can test, adapt, buy, improve and scale faster than the threat changes.
Traditional procurement is a poor fit for that. It tries to write the full requirement up front, run a long competition, and buy a finished product years later. That can work for some big platforms. It does not work well for software, autonomy, electronic warfare, data systems or anything else that keeps changing.
This is where “ecosystem” should mean something practical.
It should not mean more roundtables or branding. It should mean lining up the firms, integrators, operators, test sites, capital and government buyers needed to solve a specific problem — then using procurement to move the best Canadian-controlled options from pilot to deployment to export.
Counter-UAS is a good example. No one company is likely to deliver the whole thing. A real Canadian capability would pull together sensors, radar, electronic warfare, artificial intelligence, command and control, interceptors, operator training, test ranges, sustainment and integration with what defence and security already use.
The goal should not stop at meeting Canadian needs. It should be export-ready Canadian solutions that Ottawa can credibly back in allied markets, including through government-to-government channels.
That is what an ecosystem should look like. Not more panels. Not more pilot announcements. Not more innovation theatre. Firms, operators, integrators, investors and buyers connected by real demand and repeat procurement.
Third, define “Canadian-controlled” in a way that actually matters.
Canada will not build sovereign capacity if procurement treats a Canadian-controlled firm and a foreign-controlled subsidiary as the same thing. Having an office here matters. Control matters more. If the point of spending public money is to strengthen domestic capability, then ownership, who makes the decisions, who holds the IP and data rights, and how resilient the supply chain is should show up in the evaluation criteria.
This is especially true in defence and dual-use work. Canada should know whether public money is building firms that can make decisions here, hold IP here, hire here, export from here and put earnings back into Canadian capacity.
Fourth, connect test environments to buyers and budgets.
Companies need ranges, operators, secure digital testbeds, Arctic conditions and honest operational feedback. Those things only change the industrial base if they are tied to procurement authority. A successful test should lead to a defined next step: a limited deployment, a production contract, a scale-up competition, or a clear no.
Without that link, Canada can end up with excellent validation infrastructure sitting on a weak path to market.
Budget 2026 should measure success that way. The question is not only how many pilots launched, how many firms showed up, or how many technologies got tested. The question is what happened next.
Did Canadian-controlled firms win contracts? Did they prime? Did they keep the IP? Did they hire? Did they build production capacity? Did they export? Did government users get capability faster? Did the Canadian supplier base get stronger?
Those are the outcomes that count.
Canada does not have to choose between value for taxpayers and support for Canadian SMEs. Done properly, a better SME procurement path can do both. Smaller firms can move quickly, specialize and adapt faster than many incumbents. Government can manage risk by starting small, scaling on performance and shutting down what does not work.
The aim is not more process. It is faster capability, stronger firms and more Canadian control.
Budget 2026 can help officials get the results these programs were supposed to deliver. It can turn entry points into pathways, pilots into purchases, and Canadian innovation into Canadian capability.
Build-Partner-Buy asks the right first question: can Canada build this?
Budget 2026 should ask the next one: are we giving Canadian-controlled firms a real path from pilot to deployment?
A 25 percent procurement set-aside for Canadian-controlled SMEs is a concrete place to start. A scale-up lane past the pilot stage would make that commitment real. And a procurement system that rewards Canadian control, Canadian IP and Canadian delivery would turn ecosystem talk into industrial capacity.
Canada should not just fund Canadian pilots.
It should build Canadian companies.