Set the Hyrax Marketing Budget

1 of 8

Three decisions: how much, where it goes, and what it has to return. Everything after this slide is the evidence for those three.

What this deck decides

The number. One annual figure for Hyrax marketing, phased by month.
The line items. Which channels and programs get funded, and at what level.
The return. What each line is expected to produce, and how we will know.

Where we are starting from

$1,500 a day on paid media, ramped from $15K to $45K a month over four months.
Paid media is the entire budget. Content, events, PR and measurement are funded at zero.
Signups are getting cheaper as spend rises, which is the case for funding the rest.

All figures in this deck are Hyrax account actuals from Google Ads, Reddit, LinkedIn and X, plus PostHog project 436309, pulled 2026-09-02.

Spend and Signups

2 of 8

Spend doubled from July to August. Signups went up 5x.

Bars are paid media spend, left axis. Line is signups, right axis. Spend from Google Ads, Reddit, LinkedIn and X account actuals. Signups are wizard.started in PostHog project 436309, the only signup event instrumented continuously since June; account_created was not added until 2026-08-10 and shows 245 in August. September is the configured run rate rather than actuals, so the signup line stops at August. Pulled 2026-09-02.

Cost Per Signup

3 of 8

Down 76% in two months. Every dollar added has bought signups more cheaply than the dollar before it.

$451June

→

$264July

→

$107August

June$451

July$264

August$107

Total paid spend divided by wizard.started in the same month. On account_created, which only exists from 2026-08-10, August comes to $80 per signup on 245 signups. Neither number is a cost per paying customer; that cannot be calculated yet, which is slide 6.

The Ramp

4 of 8

Authorized budget has tripled. Actual spend has never once reached the ceiling.

Period	Authorized / month	Actual spend	Used	What ran
May	$15,000	$3,306	22%	Google only
June	$20,000	$17,588	88%	Google, Reddit, LinkedIn
July	$25,000	$9,768	39%	Google, Reddit, LinkedIn
August	$45,000	$19,625	44%	Google, Reddit, LinkedIn, X
September, to date	$45,000	$1,500/day configured	—	Google-weighted, search-led

Authorized levels per the approved monthly caps. Actuals are account-reported spend for the calendar month. The gap between authorized and actual is not underspending by choice; it is the channel set running out of inventory that converts at target before the cap is reached.

Where August Went

5 of 8

Four channels. Google took over half.

Channel	Spend	Share	Clicks	Cost per click
Google Ads	$10,513	54%	25,229	$0.42
X	$3,868	20%	8,376	$0.46
Reddit	$2,834	14%	13,575	$0.21
LinkedIn	$2,410	12%	3,455	$0.70
Total	$19,625	100%	50,635	$0.39

August 2026 account actuals. Google clicks include 23,619 from two Demand Gen campaigns at $0.07 and $0.33 per click, which is why the blended Google cost per click reads far below its search cost per click of roughly $3 to $7. Meta ran under $10 in the month and is excluded.

The Line Items

6 of 8

Everything Hyrax marketing currently buys, and everything it does not.

Line item	August actual	Status
Google Ads	$10,513	Running, search-led, the scale lever
X	$3,868	Running, no attributable conversions to date
Reddit	$2,834	Dark since late August, destination failing
LinkedIn	$2,410	Thought-leader ads plus one boost post
Meta	<$10	Paused, gated on pixel verification
Paid media total	$19,625
Committed but not yet in a budget
Out of home, SF	$17,000	Blindspot proposal, quoted, undecided
Commercial production	$5,000	Noldio, capped
Product Hunt launch	no number	Social Growth Labs, assets due Sep 7
Influencer campaign	no number	In negotiation
Not funded at all
Content and SEO	$0	No writer, no agency, no cadence
Events and community	$0	No conference presence of any kind
PR	$0	No agency, no launch programme
Video production	$0	Beyond the single capped commercial
Attribution and measurement	$0	See slide 6
Design and creative	$0	No contractor

Rows marked "no number" need a quote before they can be budgeted. Rows marked $0 have no line at all today.

We Cannot Say Which Channel Works

7 of 8

Every signup arrives with no source attached. Channel budgets are currently being set on cost per click, not cost per customer.

What is broken

Source is empty on every signup. The handoff from the site to the app drops the tag, so all 245 August signups are untagged.
The completion event has gaps. wizard.completed ran at 38 in August against 245 accounts created.
Reddit's conversion feed is down. Failing continuously for roughly four weeks.

What it costs us

No channel can be ranked against another. Any cut or increase made today is a guess.
No cost per paying customer exists. Which means no route to a payback number, and no way to argue efficiency to finance.
This is the cheapest line in the budget and it gates the value of every other line.

PostHog project 436309, source breakdown for August and September signups. Reddit conversion destination status per the daily optimizer log, 2026-09-02.

The Ask

8 of 8

Signups are getting cheaper as spend rises. Fund the channels that are missing, and fix measurement first.

What the data supports

Cost per signup fell 76% in two months while spend rose. That is the case for more budget, not less.
The cap is not the constraint. Actual spend hit 88% of the cap once and 44% last month. Paid media runs out of inventory that converts before it runs out of money.
So the next dollar cannot go to more of the same. It has to buy channels that are currently at zero.

In this order

1. Attribution. Smallest line, gates everything else. Nothing else should be funded until a signup carries a source.
2. Content and SEO. The only channel whose cost falls over time instead of rising.
3. Events. Where this buyer actually evaluates tools.
4. Then more paid. Once cost per paying customer is measurable, scaling paid becomes an arithmetic decision rather than an argument.

All figures from Google Ads, Reddit, LinkedIn and X account actuals and PostHog project 436309, pulled 2026-09-02.