Sheridan, Wyoming  ·  Software & community

We build the software.
The audience came with it.

HORSX HOLDINGS runs two divisions. One ships apps — HelpVote is live on both stores, where people post two photos of a room and ask which sofa to keep. The other operates the three Facebook communities that grew around it: women, mostly 35 to 65, mostly in the United States and the United Kingdom, asking each other what to buy. Brands can now reach them.

21,213,571
Content views in 28 days,
to 2 October 2026
183,649
Posts and comments
in the same 28 days
See who they are → Our software

Content views are the number of times posts were seen, not unique people and not advertising impressions. Combined membership across the three is 4,464,400, which is a sum of three counts rather than a deduplicated audience. Both caveats are spelled out below, along with what last month looked like.

21,213,571
Content views,
28 days to 2 Oct 2026
183,649
Posts and comments,
same 28 days
4,464,400
Combined membership
across three communities
~88%
Women, mostly
between 35 and 64

Read this before you use any of those four numbers. Content views are the number of times posts in the groups were seen. They are not unique people, they are not advertising impressions, and they should not be divided by membership to produce a per-person frequency. Combined membership is the sum of three separate group counts, so a woman who belongs to two of our groups is counted twice in it; we do not know the deduplicated figure and will not estimate one. The percentages are unaffected by either problem. Sources: Facebook Group Insights — membership and demographics read 5 October 2026, activity from the Growth tab for the 28 days ending 2 October 2026.

One company, two divisions

Where does an audience
this size come from?

It is the first question a media buyer asks, and in this industry most of the honest answers are unflattering. Groups get bought from whoever built them. Groups get farmed with giveaways. Ours has a different answer, and it is the reason the company is worth a conversation.

We built an app. HelpVote asks you to post two photos and let other people vote on which is better — which sofa, which rug, which paint, which dress. We opened communities alongside it, and they filled with the same people: women who would rather ask before they buy.

That is the whole origin story, and it is unusual enough to be worth stating plainly. The audience was not acquired. It was attracted, by a product we made and still run.

So HORSX HOLDINGS has two divisions and each one feeds the other. Software builds and ships the apps. Community operates the audience those apps gathered, and is now open to brands.

Division one

Software

Apps designed, built and shipped in-house. HelpVote is live on the App Store and Google Play in eight languages, and the studio behind it continues to develop.

The software →

Division two

Community

Three Facebook communities we own and operate, open to brands as sponsored posts, native content, pinned placement, cover takeovers and campaign-length moderator access.

The audience →

Division one — Software

We build the apps that
start the conversations.

HORSX HOLDINGS designs, builds and ships its own software. HelpVote is live on the App Store and Google Play, it is maintained from here, and it is the reason the communities in the second half of this page exist at all.

HelpVote

Post two to four photos. The community votes. You get an answer.

LIVE ON BOTH STORES

The three communities grew around an app we designed, built and shipped ourselves: thirteen everyday categories, home and decor among the busiest, in eight languages, on iOS and Android. It is the reason the groups exist and the reason they are full of people who are used to answering a question about someone else's living room.

HelpVote is live, maintained and updated on both stores, and the studio that built it is still running. Future apps from HORSX ship on the same stack, through the same developer accounts, under the same release process. For a brand partner that matters in a practical way: when a campaign needs something made — a landing page, a voting mechanic, a way of collecting answers from an audience — the capability to make it is in this building, not on a subcontractor's roadmap.

Flutter Firebase Cloud Firestore Cloud Functions iOS and Android 8 languages 13 categories
Download on the
App Store
Get it on
Google Play

Division two — Community

The women in these
groups are mid‑decision.

That is the whole proposition. Not awareness, not a vague lifestyle affinity — a woman with a shortlist of two, asking for a second opinion before she spends the money.

The groups began as somewhere to settle small domestic arguments. This sofa or that one. Grey or green. Keep it or send it back. What grew out of that is a standing audience of women who are comfortable asking what other people think before they buy, and just as comfortable telling someone else what they think.

Most of them are between 35 and 65. Most are somewhere in the middle of a house they care about, working through it room by room, on a real budget. The things they post about are ordinary and specific: a rug that has to survive a dog and a toddler, a paint colour that looks like a different colour after six in the evening, whether a chair is worth what it costs.

They are concentrated in the United States and the United Kingdom, with meaningful numbers in Canada and Australia. In every one of the three groups, the 13–17 band reads zero. There is no teenage population here to write around.

And they turn up. Over the 28 days to 2 October they left 183,649 posts and comments between them, and group content was viewed 21,213,571 times. Those are pages with traffic on them, which is the thing a large member count on its own never tells you.

On how we know what they are interested in. Facebook's panel proves gender, age and location. It says nothing at all about interests, so we are not going to infer a shopping habit from a demographic and sell it to you as insight. The evidence is the app instead. HelpVote exists for one thing: you photograph two options and ask which is better. Which sofa. Which rug. Which curtains. Which colour on that wall. These communities formed around that question and are still built on it — which is a behaviour we can point at, not an affinity segment somebody bought.

Audience & activity

Three owned communities.
One very specific audience.

HORSX owns and operates three public Facebook communities. Below is what they did in the last 28 days, and who the members are — taken straight from each group's own Insights panel, with the limits of each figure stated next to it.

Source: Facebook Group Insights. Activity from the Growth tab, “Last 28 days”, period ending 2 October 2026. Membership composition read 5 October 2026.

Activity · 28 days to 2 October 2026

What the three communities did in four weeks

21,213,571
content views across the three groups
183,649
posts and comments in the same 28 days

“Content views” means exactly one thing: the number of times posts in these groups were seen, as Facebook counts it in the Group Insights Growth tab. It is not unique people, not reach and not ad impressions, and it must not be divided by membership to produce a per-person frequency.

These two totals are event counts, so summing them across the three groups is exact. The double-counting caveat below applies to membership only — not to views, and not to posts and comments.

HelpVote
Content views, 28 days 12,213,788
Posts & comments, 28 days 81,850
Members 2,500,000
HelpVote: Community Decisions
Content views, 28 days 5,311,635
Posts & comments, 28 days 62,576
Members 764,400
Helpvote.app
Content views, 28 days 3,688,148
Posts & comments, 28 days 39,223
Members 1,200,000
Combined
Content views, 28 days21,213,571
Posts & comments, 28 days183,649
Members4,464,400 combined membership
Within each column a bar shows that group's share of the column total. The three columns are different measures in different units — compare bars down a column, never across.

The members column is a sum, not a deduplicated audience. Somebody who belongs to two of the groups is counted twice. We do not know the overlap, so we claim 4,464,400 combined memberships — never 4.4M unique people, and never 4.4M reach. The two activity columns are not affected by this.

  • The activity order is not the membership order: Community Decisions is the smallest of the three groups and the second largest on both activity measures.
  • The two largest member counts are Facebook's own rounded displays (“2.5M”, “1.2M”), so the combined membership figure looks more precise than it is. The activity figures are exact as reported.
Gender

Overwhelmingly women — in all three groups

≈88%
weighted female share
across the three groups

A percentage is unaffected by group overlap — this number survives the double-counting caveat intact.

Each bar is 100% of that group's members. The panel reports 0% custom in all three groups; the remainder is male.

HelpVote 85% female
Community Decisions 95% female
Helpvote.app 89% female
Female (group colour) Male
Age

The weight sits from 35 upward

Share of each group's members, by age band.

  • HelpVote
  • Community Decisions
  • Helpvote.app
13–17

≈0% — the panel draws no bar here, in any group.

18–24
25–34
35–44
45–54
55–64
65+
Approximate share of that group's members; the axis runs 0–25%. Shaded rows = 35–64.

Read the shape, not the decimals. The Insights age chart carries no numeric labels, so these are heights read off the bars. They total ≈79%, ≈91% and ≈81% rather than 100% — that gap is our reading error, not a hidden segment.

  • 35–64 accounts for roughly 53% / 57% / 62% of the three groups (sums of the approximate bars above).
  • Under 25 is ≈2–3% of every group.
  • 25–34 lands on ≈11% in all three, independently.
View age data as a table
Approximate share of members by age band, read off unlabelled Insights bars.
Age bandHelpVoteCommunity DecisionsHelpvote.app
13–17≈0%≈0%≈0%
18–24≈3%≈2%≈2%
25–34≈11%≈11%≈11%
35–44≈17%≈17%≈25%
45–54≈20%≈19%≈22%
55–64≈16%≈21%≈15%
65+≈12%≈21%≈6%
Bars total≈79%≈91%≈81%
Geography

The United States is the audience

Members by country, summed across the three groups, each drawn against the full 4,464,400 combined membership.

United States 1,850,153 41.4%
United Kingdom 315,601 7.1%
Canada 190,750 4.3%
Australia 147,862 3.3%
Everywhere else, combined 1,960,034 43.9%
Bar length = share of the 4,464,400 combined membership. The five rows sum to 100%.

Why only four countries are named. The United States, United Kingdom, Canada and Australia each appear in all three groups' top-ten country lists, so their sums are complete. A country that appears in only one list — Germany, for instance, at 92,197 in Helpvote.app — cannot be summed honestly and sits inside “everywhere else.”

Same caveat as the members column above: these are combined memberships across three groups, not deduplicated people. A member of two groups is in these numbers twice.

2,313,616
US + UK + Australia combined memberships — ≈52% of the total
≈41%
of all combined membership sits in the United States alone
English
the top four countries are all English-speaking markets
Method & limits

What this data is, and what it is not

Where the numbers come from

  • Facebook Group Insights for three public groups owned and operated by HORSX HOLDINGS LLC. Activity from the Growth tab, “Last 28 days”, period ending 2 October 2026; membership composition read 5 October 2026.
  • Content views and posts & comments are event counts. Summing them across the three groups is exact, and nothing is double-counted.
  • Membership totals are not deduplicated. They are sums of three separate member counts; a person in two groups is counted twice, and we cannot tell you by how much. That is why every total on this page reads “combined membership”.
  • Member counts are the panel's displayed values. Two of the three are rounded by Facebook itself (“2.5M”, “1.2M”).
  • Age percentages are read off a chart that carries no numeric labels. Treat them as a distribution shape, not as measurements.
  • Country figures are summed from each group's top-ten list, and only for countries present in all three lists.

How to read the activity figures

  • This is a single 28-day window, not a trend line. It is a sharp increase on the preceding 28 days — between roughly +200% and +1,250% depending on the group — which means the earlier period was far quieter. We are not projecting that forward and we are not treating it as a run rate.
  • We do not divide views by members. Content views are not people, so any per-person figure derived from them would be invented.

Not in this data set — and therefore not claimed anywhere above

  • Unique reach — how many distinct people saw anything
  • Ad impressions
  • Click-through and conversion
  • Per-post performance
  • Member growth over time
  • A deduplicated cross-group audience

None of it is estimated, modelled or implied on this page. When those panels are read, they will be published the same way — as what the panel says, with its limits attached, and nothing more.

The three communities

Each one is a different woman.

Same instinct, three different centres of gravity. Ordered by content views over the last 28 days, not by headcount. Every figure below is read directly from that group's Insights panel.

Public group
12,213,788
content views, 28 days
HelpVote

The largest and busiest of the three, and the most international. A little over a million of its memberships are in the United States.

Activity, 28 days to 2 Oct 2026

Content views12,213,788
Posts and comments81,850

Members and composition

Members2,500,000
Women85%
Largest age band45–54 (~20%)
Then35–44 (~17%)
Then55–64 (~16%)

Top countries

United States1,053,958
United Kingdom108,664
Canada105,928
Australia60,317

Also in its top ten: South Africa (65,973), Philippines (56,548), Mexico (50,633), India (44,005), Romania (36,659), Egypt (31,734). Top cities are a genuine global mix — New York (12,240), Melbourne (10,610), Cairo (10,394), Dhaka (9,703), Sydney (9,649), London (8,594).

Public group
5,311,635
content views, 28 days
HelpVote: Community Decisions

The most American and the most female of the three, and the one to buy if a campaign is US-only. It also skews oldest.

Activity, 28 days to 2 Oct 2026

Content views5,311,635
Posts and comments62,576

Members and composition

Members764,400
Women95%
Largest age band55–64 (~21%)
Equal with65+ (~21%)
Then45–54 (~19%)

Top countries

United States636,152
United Kingdom34,950
Canada29,727
Australia11,521

Also in its top ten: South Africa (6,431), Ireland (3,008), Nigeria (2,870), Philippines (2,683). Top cities are almost entirely American — New York (4,804), Houston (3,208), San Antonio (3,035), Chicago (2,651), Phoenix (1,959), Jacksonville (1,840).

Public group
3,688,148
content views, 28 days
Helpvote.app

The British and Northern European one, and the youngest of the three. Germany and Australia carry real weight behind the UK.

Activity, 28 days to 2 Oct 2026

Content views3,688,148
Posts and comments39,223

Members and composition

Members1,200,000
Women89%
Largest age band35–44 (~25%)
Then45–54 (~22%)
Then55–64 (~15%)

Top countries

United Kingdom171,987
United States160,043
Germany92,197
Australia76,024

Also in its top ten: Canada (55,095), Netherlands (44,761), Sweden (42,777), Denmark (38,221), Italy (28,355), France (27,314). Top cities — Melbourne (11,947), Sydney (11,885), London (7,681), Auckland (7,502), Perth (6,655), Dublin (4,880).

Where the memberships are, summed across all three

United States1,850,153
United Kingdom315,601
Canada190,750
Australia147,862

Same caveat as every other total on this page: these are three numbers added together, not unique people.

On picking groups. If the campaign is United States only, HelpVote: Community Decisions is the one that matters: 83% of its membership is American and its top cities are New York, Houston, San Antonio, Chicago and Phoenix. HelpVote is the volume, but it is genuinely international — Cairo and Dhaka sit in its top ten cities, and we would rather you knew that from this page than from a screenshot.

On the age chart. Facebook does not put numeric labels on the age distribution, only bars. The percentages above are read off those bars and are approximate, which is why they carry a tilde. The shape is unambiguous even if the decimals are not: in all three groups the mass sits from 35 upward, and the 13–17 band is flat at zero.

On gender. 85%, 89% and 95% female across the three, weighting out to roughly 88%. That figure is a proportion, so the double-counting problem does not touch it.

Before you ask

What we can show you,
and what we cannot.

You are going to ask for numbers we do not have yet. Here they are, listed by us first, so you do not have to find out in the second email.

We can show you

  • Content views per group for the last 28 days
  • Posts and comments per group for the same period
  • Membership counts per group, straight from Insights
  • Gender split per group
  • Age distribution per group, as unlabelled bars
  • Top ten countries and cities per group, with exact counts
  • Screenshots of any of the above, on request, same day

We cannot, yet

  • Unique reach. Views are views, and we will not convert them into people for you
  • Impressions in the advertising sense. This is not an ad server
  • Per-post performance history or an engagement rate
  • Click-through or conversion data. None exists
  • Member growth over time. Not captured
  • Case studies or past campaign results. You would be the first

The 28 days we are quoting were a step change, and you should know that before you price anything. The month before them was much quieter. In the largest group, content views went from roughly 900,000 in the previous period to the 12.2 million reported for the period ending 2 October. The other two moved in the same direction by smaller multiples.

We are telling you this rather than leading with the percentage, because a chart like that is as likely to mean a quiet prior month as a permanent new level, and we do not yet know which. We will not project the curve forward and we would be wary of anyone who did. Price against the absolute 28-day figures, treat the trajectory as unproven, and ask us again in a month — by then there will be a second period to compare and we will publish it whichever way it goes.

A media kit that quotes a number it cannot produce on request loses the account at the first reply, so this page quotes only what is in the Insights panel today, with the panel and date attached to each figure. If being the first campaign here means you want it priced accordingly, say so in the first email. That is a reasonable position and we will treat it as one.

Placements

Five ways to be in
front of them.

Each is sold per group, so you can buy one community or all three. Nothing is bundled and nothing is auctioned. Tell us the format, the group and the window, and you get a written price.

Sponsored post

One post published by the group account, carrying your images, your copy and your link, labelled as paid. You supply the creative, or we draft it and you approve the final wording before anything goes up. It stays on the page after the campaign ends.

Unit: one post, one group. You supply: images, copy, destination URL. Lead time: three working days from approved creative.

Native content integration

Your range dropped into the format the group already runs: two or four of your products, side by side, posted as a question rather than an advert. Which rug for this room. Which of these three in a south-facing kitchen. Still labelled as paid, but it reads like the rest of the page.

Unit: one post, one group. You supply: two to four product shots. Useful for: range testing and shortlist products.

Pinned placement

Your post held at the top of the group for an agreed window — a week, two weeks, the length of a sale. One pinned slot per group at a time, so it is sold as an exclusive for the dates you book and nobody else sits above you.

Unit: priced per day, per group. Exclusive: one advertiser at a time. Pairs with: a sponsored or native post.

Cover-image takeover

Your artwork becomes the group's cover image for the run of the campaign — the first thing anyone sees at the top of the page, on desktop and on the app. One group or all three. We send you the exact dimensions and safe areas, and it goes back to ours on the agreed date.

Unit: priced per week, per group. You supply: artwork to spec. Best with: a seasonal or launch window.

Partner moderator access, for the length of a campaign

A named person from your team is given moderator rights on the group for the campaign window. You post on your own cadence rather than ours, reply to members in your own voice, and run your own threads without going through us for every asset. The ground rules — what gets posted, how often, what happens to member questions, what happens at the end — are agreed in writing before access is granted, and access is removed on the closing date. This is the deepest thing we sell and the one we are most careful about, because the audience belongs to the group before it belongs to anyone's campaign.

Unit: priced for the campaign window, per group. You supply: one named moderator. Requires: a signed scope before access.

On pricing. There is no public rate card, because the right number depends on which group, which format and how long. Send the brief and you get a price in writing, per placement, with the dates. No retainers, no minimum commitment, no media agency markup in the middle.

How we work

Four steps, and
no surprises in them.

A campaign is a short email thread, an approval, a publish date and a set of numbers afterwards. That is deliberately all it is.

Send a brief →
1

The brief

What you are selling, who it is for, and when it needs to be live. One paragraph is usually enough to tell whether we are a fit, and we will say so quickly if we are not.

2

A placement plan, in writing

Which of the three groups, which formats, which dates, what it costs. One page. If a group is wrong for your product we will take it off the plan rather than sell it to you.

3

Your approval, then we publish

Nothing goes live without your sign-off on the final image and the final line of copy. Paid placements are labelled as paid. Where you have moderator access, you publish on your own schedule inside the agreed rules.

4

What came back

After the run you get the post-level figures Facebook reports for your own placements, exported as they come, with no interpretation layer on top. If a placement does poorly we will tell you it did poorly.

A good fit

Rugs and flooring. Furniture and lighting. Paint, wallpaper and textiles. Kitchen and tableware. Storage and organisation. Garden. Homeware gifting. Women's apparel and everyday beauty. Anything where the buying decision is visual, discussed before it is made, and lands in a home.

A poor fit

Anything aimed at men under thirty. B2B software. Financial products and crypto. Anything that needs a verified cost per acquisition from day one, because we cannot yet give you one. If that is your requirement, we would rather tell you now than take a first order and disappoint you on the report.

Tell us what you are launching and when.

One paragraph is enough for a first answer. You will hear back from a person, and the price will be in writing.

[email protected] →

Contact

Who you are
dealing with.

Registered office

75 E 3RD ST STE 7
SHERIDAN, WY 82801
United States

Company profile

Legal nameHORSX HOLDINGS LLC
Entity typeLimited Liability Company
JurisdictionWyoming, USA
RegisteredMarch 2026
BusinessSoftware & community marketing
Apps liveHelpVote (iOS, Android)
Owned communities3 Facebook groups
Content views, 28 days21,213,571
Posts and comments183,649
Combined membership4,464,400
Websitehorsxholdings.com
BillingDirect, in USD

Activity figures cover the 28 days ending 2 October 2026. Content views are times posts were seen, not unique people. Combined membership is the sum of three group counts and includes people who belong to more than one group.