New in 2026
Anti-Racism at Work
Speak Up : what you say in the ten seconds after something racist happens at work.
Register your interestSomeone says something in a meeting and it lands badly. Nobody says anything, and the moment passes.
We've built Speak Up around those ten seconds, because that's the moment we keep hearing about. Not the policy, not the training day years ago. The gap where someone could have said something and didn't, and everyone in the room felt it.
The design's ready for the first pilot. What hasn't happened yet is running it with real people in a real room, and that's what we're looking for now. Three organisations to go first, at a reduced rate, with the measurement built in from day one.
What you get
About half a day in a room, for 8 to 20 people, in person.
Most of it is practice. People run real workplace situations out loud, get them wrong, and go again. That's the point. Nobody leaves having only talked about it.
Short pre-work beforehand, so the day itself opens on practice instead of definitions everyone already knows.
In the room
The session closes with your people putting requests to somebody who can actually act on them, and getting one of four answers: yes, no, I'll try, or whatever. The last two both count as no, and everyone in the room knows that going in. Every request and every refusal gets written down in the words it was said in, and both come back to you in the report.
Over three months
Then three follow-ups over three months, because a single session doesn't hold on its own. A one-question email at two weeks. A 90-minute group call at six weeks, working through whatever people have actually run into since. A short survey at three months, checked against a baseline taken the week before.
You get a written summary at the end. Whatever the numbers say.
What this is part of
This pilot is one piece of something bigger we're building. The full program works on three things: what people say to each other, how one decision actually gets made, and whether that change is still holding a year later.
Speak Up is the first piece, and it's the one strong enough to stand alone. If this works, we build the rest together with whoever's in the room for it.
Before you have to ask
The limits
Most people who've bought this kind of training before have either been burned by it or are bracing for it. Here's what that means for how we've built this.
We'll tell you what's actually evidenced, and what isn't.
Anti-racism training as a category is barely tested. A systematic review covering 2000 to 2022 found only 8 studies, and just 2 of those followed anyone up afterwards. What does hold up, well, is interpersonal confrontation as a skill and behavioural rehearsal as the way to build it. That's what Speak Up is built from.
Nobody walks out having been told they're the problem.
The session isn't interested in intent, motive, or politics. It's interested in what was said and what it did. That's a technique question, and technique is something people can actually get better at.
We've designed this assuming people who've experienced racism are in the room, because they usually are.
Nobody's expected to speak for an experience just because they've lived it, and anyone can step out of an activity with no explanation needed. Every scenario runs from the bystander's position, not the target's. If you'd like a short separate briefing for staff who might find the material personally difficult, just ask and we'll build it in.
If it doesn't work, you'll hear that from us.
The measurement runs regardless of what it finds. Flat numbers still get a report, with our honest read on why.
Speak Up teaches one skill, to the point people can actually use it under pressure. It won't change how decisions get made in your organisation, and on its own it isn't an anti-racism program. That's further down the road, once this piece is proven.
What we need from you
-
01
One person in the room who can actually change how something gets decided.
Not the whole leadership team, just one name. The session closes with people putting real requests to that person, and without someone able to say yes, the whole thing caps out at what individuals can do on their own, which is real but small. We'll need that name before we lock a date.
-
02
Attendance that people choose, or that's framed as development rather than compliance.
Mandatory framing triggers resistance and has been shown to produce worse outcomes than doing nothing at all. Choice removes that barrier so the content can actually take hold.
-
03
Permission to ask your people two short surveys, five minutes each.
It's the best way for us to measure the lasting impact of this work and see what actually stuck.
Three pilot places
Going first
The standard rate for this session is $5,200 plus GST. The first three organisations pay about half that.
What we'd like in return is permission to publish what the measurement shows, anonymised and in aggregate, with your organisation named only if you want it to be. If your people would rather we didn't, no problem, and the reduced rate still stands.
For us, validating the methodology with real data matters more than the margin. Three cohorts is roughly what it takes before either of us actually knows whether this works.
Expressions of interest
Register your interest
We are taking on three clients for this round, focusing on where this work will have the greatest impact rather than rushing the selection.
Share where you are currently at and what sparked your interest. We'll get back to you with an honest view on whether this aligns, and we'll be in touch to organise a quick discovery call to discuss scope and timing.
Why this matters now
Most People and Culture teams can already show the organisation holds the right values. The harder thing to show a regulator, a board, or someone who's just made a complaint, is that you took a reasonable step. Training on its own doesn't answer that. A session that actually changes what people do afterwards, with data behind it, gives you something real to put in front of them.
The same workplaces, described twice
"43% of racially marginalised workers said racism was common, against 18% of racially privileged."
Diversity Council Australia surveyed 1,547 Australian workers for its Racism at Work report.
88%
said racism is an issue in Australian organisations
93%
said organisations need to act on it
27%
said their own organisation was actually doing something about it before it became a problem
The duty
Safe Work Australia now treats racism at work as a psychosocial hazard, which puts a duty on employers under work health and safety law to do what's reasonably practicable about it, the same as any other workplace risk.
In Victoria, the positive duty under the Equal Opportunity Act 2010 already covers race. A positive duty means acting before someone complains, not just responding after. The Australian Human Rights Commission's National Anti-Racism Framework, released in November 2024, recommends the same duty apply federally.
What else?
- A clearer read on why the same workplace gets described so differently depending on who you ask
- A repeatable way to name conduct and its effect without accusing anyone of intent
- Real practice saying it out loud, including the attempts that go badly, because that's where the learning actually is
- A way to choose between speaking up in the moment, following up privately, or escalating
- Something they've committed to doing differently, inside their own authority, no budget or permission needed
- A request put directly to someone who can act on it, and an answer
The two do genuinely different work, and it's worth being clear about the difference when you're briefing participants.
Unconscious bias training
- Looks at how your own mind works
- Core skill: notice, interrupt, and replace your automatic response
- Best used before a decision, to sharpen individual judgement
Speak Up
- Looks at what gets said and done between people
- Core skill: name someone else's conduct and its effect, out loud, in the moment
- Best used after an incident, to change what a room tolerates
They're complementary, and neither replaces the other. We think Speak Up works well as a follow-on to unconscious bias training, though we haven't tested that sequencing yet. It works fine with no prior training at all too.
We'd rather know whether people actually speak up afterwards than how they rated the day, so the measurement's built into the design from the start rather than bolted on later.
Three questions, asked the week before the session and again at three months. How confident people feel about speaking up, how many times they actually did, and how many times they saw something and stayed quiet. Three months is where a peer-reviewed bystander study still found a real effect.
We'll be upfront about the limits. This is self-reported, there's no control group, so it isn't a controlled study and we won't call it one. Two things push the numbers up. People who reply are usually the people who acted. And anyone who's just spent half a day learning to speak up has a pretty good guess at what answer we're hoping for. We report the response rate alongside the results, and we read both with that in mind.
Satisfaction gets collected separately, on its own form. It doesn't sit next to the behaviour questions, because asking people how much they enjoyed the day changes how they answer everything else.
These first few cohorts are where we finalise the instrument properly.
We design this assuming people who've experienced racism are in the room, because in most workplaces, they are. Nobody's asked to represent an experience on anyone's behalf. Anyone can step out of an activity, no explanation required. Every scenario is played from the bystander's position, never the target's.
If you'd like a short separate briefing offered to staff who might find the material personally difficult, raise it when we scope the session and we'll build it in.
Is half a day enough to change anything?
On its own, no. Speak Up builds one skill to the point people can actually use it, with three check-ins over three months because a single session on its own doesn't hold. Changing how decisions actually get made is bigger work, and it comes later.
Does this meet our WHS duty around racism as a psychosocial hazard?
Not on its own. That duty covers your whole risk management process, and no single session discharges it. What this gives you is a genuine, documented, outcome-measured step you can point to as part of that process.
What if we can't get a senior person into the room?
Tell us anyway. It's the one condition we hold firm on, because the session lands a lot weaker without it. We're building a shorter briefing for exactly that situation, tell us where you're at and we'll figure out the right first step together.
What if people in the room get defensive?
We've designed for it. Nobody's told they're the problem, and the session works on technique rather than moral standing. Compliance-style, controlling framing tends to backfire, so we'll say something early if a brief starts heading that way.
Do you run this online?
No, not yet. The whole session is built around live rehearsal, reading a room and adjusting in real time, and we haven't designed or tested a version that holds up on a screen. If distance or travel is the actual blocker, tell us, and we'll work out what's possible.
What happens if the results are bad?
You'll hear it straight, including if there's not much to report. That matters more to us than looking good, because almost nobody in this market actually publishes three-month data at all.
We're looking for three organisations to build this with. Bring the incident, the audit finding, or the thing you can't quite name yet, and let's find out if it's a fit.
Register your interest