Equality, Diversity and Inclusion Policy

Why a sailing product needs this policy

Sailing is not an easy sport to get into. A boat costs money. Getting to the events that matter costs more, and it costs a weekend. Clubs can be welcoming and can also be closed shops. The water is a long way from most people. Pontoons and boats are hard to use if you cannot walk easily. And the number of girls racing dinghies falls away sharply through the teenage years, in a way that the number of boys does not.

None of that is our doing. But we are building a product that records who trains, who races, and how well — and a product like that can either soften those patterns or harden them. This policy is about which one we intend to be, and about the places where the honest answer is that we are not there yet.

It applies to Innovology Ltd, trading as SailCoach: to how we build the product, how we behave towards the people who use it, and how we work as a company.

What the law requires of us

The Equality Act 2010 makes it unlawful for a service provider to discriminate against or victimise someone because of a protected characteristic, and to harass them — except on grounds of religion or belief or sexual orientation, which Parliament excluded from the services harassment provision (section 29(8)). There are nine characteristics.

Two of them are carved out of the services part of the Act by section 28(1). Marriage and civil partnership is excluded entirely. Age is excluded so far as it relates to under-18s — which is most of the people who use this platform; it still binds us for our adult users. We do not treat either exclusion as permission. Everything in this policy applies to all nine, to everyone. A statutory carve-out that makes age discrimination against a child lawful in a service is not a reason to design for it, and we would rather say that out loud than quote a duty we do not actually owe.

Two specific duties matter most to us:

• We must not discriminate in how we provide the service, directly or indirectly. Indirect discrimination is the one that catches products: a rule or a design that applies to everyone but puts one group at a particular disadvantage, and cannot be justified.
• We must make reasonable adjustments for disabled people, and that duty is anticipatory. We are supposed to think about it before a disabled sailor signs up, not after they complain. Our current position on that is set out in our Accessibility Statement, including the parts we have not done.

How this shapes the way we build

The fairness problem in our ratings, named rather than buried

SailCoach calculates a rating for a sailor from race results — a statistical skill estimate, updated race by race. It is described in full in our AI and Automated Analysis Policy.

Here is the problem with it, stated plainly.

The rating is fed by evidence, and evidence costs money. Official results from events carry full weight. A sailor's own self-reported results carry a quarter of that, because nobody has verified them. On top of that, the one normative judgement the product makes — the message that a sailor may have outgrown their current fleet — will not fire at all until there are at least five official results, a settled rating, and a race with a genuinely big fleet in it.

So a sailor whose family can drive to six open meetings a season accumulates the kind of evidence our system counts, faster than a sailor who races at their own club on a Sunday. A sailor whose club posts results online is visible to us; a sailor whose club does not is close to invisible. Most of the results we hold are from the Optimist class in the UK, which is neither the cheapest way into the sport nor a fair picture of who sails.

A ranking system built on that data, presented without care, tells a child that money is talent. We are not prepared to ship that. So:

• A rating is a position among the people who happened to race, not a measure of ability or potential. The product frames it that way — fleet position within your own circuit, with the number of scored races and how many were self-reported shown beneath it. It does not yet carry that sentence in full next to the number, and adding it is on the list below.
• Uncertainty is always shown. A sailor with few results gets a wide band, not a low score. Thin data reads as we do not know yet, because that is what it means.
• The bands stay unnamed — no "club", "national" or "international" labels — until we can show from our own data that the bands predict anything. Position can be honest immediately; names have to be earned.
• Sailors in the lower part of the range get no verdict at all. No headline, no judgement, no "needs to improve". That is a deliberate silence and it is tested for, not a matter of tone.
• After a sailor steps up a level, their first events are held in a neutral state, so that a promoted sailor's first mid-fleet result at a bigger regatta cannot read as going backwards.
• Self-reported results count. They are weighted lower because they are unverified, but they count — deliberately, so that a sailor whose club results never reach a results website is not treated as though they never raced.
• A rating must never be the sole basis for a squad place, a team pick or an event entry. The reasoning is meant to be visible precisely so a coach can argue with it. Today that is only half true: a "ready to step up" verdict lists the races it was built on for everyone, but the full workings behind the number — the skill estimate, the uncertainty, the size of the circuit — are shown only on an adult sailor's page. Showing them on a child's page too, which is the case this policy is actually about, is on the list below. Any club that wires our output straight into a selection decision is misusing it, and we will say so.
• Before we ever name a band, we will check whether our ratings track how much a sailor travels rather than how well they sail. If they do, we fix the model or we do not ship the names.
• We will not sell rank. No paid plan will ever buy a better rating, a higher position, or more prominence in any listing.

What we do not collect, and what follows

We do not ask anyone for their ethnicity, religion, disability, sexual orientation or gender identity. We do not collect a sailor's sex or gender at all. Gender appears in our records only where it was already printed in a published results table that we read, and we do not use it to rank, group, match or gate anything.

That is the right call for a product used by children, and it has a consequence we should be honest about: we cannot measure our own reach across most protected characteristics. We cannot tell you what share of our users are disabled or from a minority ethnic background, because we do not know and we are not going to start asking children for that information in order to produce a statistic. Where we need to understand a gap, we will look for it in the design and in what people tell us, not by profiling our users.

Other choices already made for this reason

• There is no browsable directory of sailors and no name search. You can find an account only if you already know its exact email address or phone number. One lookup does not yet meet that standard: a signed-in user can search the public results corpus by sail number and see the names and race history attached to it, whether or not they have any connection to that sailor. Restricting it to the caller's own sail numbers and the sailors they coach is a dated commitment in our data protection impact assessment. A searchable, ranked index of children is a thing we will not build.
• Head-to-head rivalry framing is switched off for under-13s. Younger sailors see their own sailing, not a list of children they are beating.
• Body weight is optional, private by default, recorded by a parent for under-13s, and never shown as a trend line. Weight matters in dinghy sailing and a progress chart of a teenager's body is a harm, not a feature.
• A coach's public page never names the children they coach.

Where our data is thin, and where our model does not fit

Our results coverage is strongest in UK dinghy racing, particularly junior classes, because those are the sites that publish. It is thinner in the rest of Europe and much thinner in North America and Australia, where we already have users. We will show a sparse record as sparse, never as a poor one.

We have also not modelled adaptive or para sailing. Our boat-class families and comparisons assume mainstream fleets, and we would rather tell you that than quietly produce numbers that mean nothing. If you coach adaptive sailing and the product does not fit what you do, please write to [email protected] — that is genuinely useful to us.

The accessibility of the product itself

A policy about inclusion that ignores whether a disabled person can use the software is not worth much.

We target WCAG 2.2 level AA. We do not currently meet it everywhere, and we have not had an independent audit. The specific things we know are wrong — race maps with no text description, charts without an alternative, colour carrying meaning on its own, video without captions — are listed with dates in our Accessibility Statement, together with how to tell us about a problem and what happens when you do.

If something about the product makes it unusable for you, tell us at [email protected]. We would rather make a change than defend a design.

What we expect of coaches, clubs and everyone else using SailCoach

We do not run sessions, employ coaches or admit anyone to a club. What we can do is set the standard for behaviour on our platform and act when it is broken.

Everyone using SailCoach is required to:

• treat sailors, parents, coaches and club volunteers fairly, whatever their background;
• not harass, abuse, demean or exclude anyone because of who they are — in messages, in session notes, in feedback, in a squad list, or in anything written about a sailor;
• write about a child as though the child and their parents will read it, because they may;
• not use the platform's records, ratings or comparisons to put a sailor down.

The detail is in the Code of Conduct and the Acceptable Use Policy. Breaking these rules can cost you access to the platform, and where a child is involved it goes to safeguarding rather than to a conduct process.

Clubs and squads using SailCoach keep their own responsibilities under the Equality Act for what happens at the club and on the water. We are not a substitute for a club's own equality policy, an RYA process, or a governing body's complaints route.

If you experience or witness discrimination

A child can raise something themselves. We will not turn a young sailor away because they did not ask a parent first.

You can also go to the Equality Advisory and Support Service (EASS) at equalityadvisoryservice.com for free, independent advice about your rights under the Equality Act — at any point, including before you come to us.

Us as an employer

We are a very small team, and we should not dress that up.

Innovology Ltd does not run a recruitment programme or a graduate scheme, and it has no HR department. Publishing workforce diversity figures for a handful of people would be meaningless and would identify them. We are also well below the 250-employee threshold at which gender pay gap reporting is required, so we do not publish one.

What we commit to, and will hold ourselves to when we do bring people in:

• Roles are advertised openly, with the pay range stated. "Someone we know" is not a hiring process.
• We describe the job, not a personality. We do not list requirements we would not actually enforce, because vague and inflated criteria filter out good candidates who read them honestly.
• Part-time, flexible and remote working are considered for every role from the start, not granted as a favour afterwards.
• We make adjustments at any stage of a process on request, and we ask everyone whether they need any — nobody has to disclose a disability to be accommodated.
• We do not ask candidates for unpaid work, speculative project work or long take-home exercises.

If you apply to work with us and something in the process disadvantaged you, say so at [email protected]. We will look at it, and we will change the process if you are right.

What we have not done yet

Stated plainly, because a policy that lists only successes is not credible:

• No independent accessibility audit has been carried out.
• We have not yet tested whether our ratings correlate with how much a sailor travels. That test is a commitment above, not a completed piece of work.
• The sentence that a rating is not a measure of ability or potential is not yet printed next to the rating itself.
• The full workings behind a rating are shown to adult sailors but not on a child's page.
• The sail-number lookup in the results corpus is open to any signed-in user, not restricted to their own sail numbers and the sailors they coach.
• We have no data on how our users break down across protected characteristics, and by design we are not collecting it.
• Adaptive and para sailing are not modelled in the product.
• Our results coverage is heavily UK-weighted and heavily weighted towards one junior class.

Review

We review this policy at least once a year, and sooner if the rating system changes materially, if we launch paid plans, or if something happens that shows this document is wrong. The next scheduled review is 17 September 2027.

Questions, disagreements and better ideas: [email protected].

Related documents

• Accessibility Statement
• Code of Conduct
• Acceptable Use Policy
• Safeguarding and Child Protection Policy
• Complaints Policy
• AI and Automated Analysis Policy
• Data Protection Impact Assessment
• Privacy Policy