For software companies

You named a category. Your buyer named a problem.

The phrases on your homepage are the ones you invented. The phrases your buyers actually type describe the problem in their own words, at many times the volume, and you appear on almost none of them. We build the pages that meet them there — which is also what makes you citable when somebody asks an assistant the same question.

Monthly searches · click a bubble
Searches you target720/mo
Searches they actually make38,400/mo
Click a phrase You rank for the words you invented. The big bubbles are people describing the problem you solve, in their own words, at a volume 50 times larger than your category name — and almost none of them will ever type it.

What is actually going wrong

Three failures that look like a demand problem and are not.

01

The monthly number hides the month that broke

Signups are reported as one figure, so a cohort that churned out before it repaid is averaged against one that did fine. The blend stays flat and reassuring for two quarters while the newest intakes get quietly worse, and by the time the average moves the cause is six months behind you.

02

Activation is somebody else's metric

Marketing is measured on signups and product on retention, so the step in between — whether a new account ever reaches the moment your software becomes useful — belongs to neither. It is the cheapest number in the business to improve and the one nobody owns.

03

Free trials are counted as though they were customers

A trial is a cost until it converts, and a channel can produce them endlessly at a flattering price. Optimising to trial starts selects precisely for the people least likely to pay, which is why the cheapest acquisition channel so often has the worst twelve-month retention.

The numbers we manage

What we would report on, and what each one is for.

You already watch some of these. The difference is that they are managed together, by cohort, so a cheaper signup can never quietly become a shorter-lived customer.

CAC payback by cohort, not blended

How long each monthly intake takes to repay what it cost. Blending them is what lets a channel fail for half a year without showing up in a report.

Activation rate

The share of new accounts that reach the moment the product proves itself. Every point here shortens payback across every channel at once, and it costs nothing in media.

Trial-to-paid, held apart from signup volume

Two channels producing identical trial numbers can differ by a factor of four here. It is the first place a cheap channel stops looking cheap.

Net revenue retention

Whether the accounts you bought last year are worth more or less this year. It decides what you can afford to pay for the next one.

CAC by channel, held apart

Costed separately so the moment one source turns is visible in the week it happens rather than in the quarterly average.

Contribution after CAC and cost to serve

What is actually left once acquisition and the cost of running the account are paid. Some healthy-looking ARR is not worth having.

Where shortlists start now

3 products get named. You are not one of them.

Asked in an AI assistant · 14 seconds ago
AI assistant · a buyer on your category
The 5 sources it read to answer. None of them is your site.
Click a source Nothing in this answer came from your website. Every one of the 5 sources is public, findable and, in 3 cases, something you could have written. The assistant did not choose against you — it never had you to choose from.

Before the price

Tick what is true. Three or more and this is your stage.

The stage you need is decided by what already repays, not by what you raised. Switch stage below and the list changes with it.

Tick the ones that are true.

Nothing is submitted and nothing is stored — this is here so you can rule the stage out as easily as rule it in.

What is inside

Foundation for SaaS, line by line.

Every line links to the service page it comes from, with its standalone price, so you can check the arithmetic rather than take our word for it.

$1,700per month

Excludes ad spend, which you pay directly to the platforms.

One channel with a cohort that visibly repays

The uncomfortable comparison

What this would cost you separately.

Media Buying — Starter, one platform$900/mo
Two ad sets a month (2 × $220)$440/mo
Fixed lines, bought separately$1,340/mo
CRO & Tracking Audit, if bought separately$950 one-time
Server-side tracking, one platform$600 one-time
Foundation for SaaS$1,700/mo

Read this honestly: at $1,340 of fixed lines against $1,700, you are not buying a discount here. You are buying the cohort and activation work that decides whether the media spend was ever worth making — and one person accountable for telling you when a channel has stopped repaying. Figures on the left are the closest equivalent line from the à la carte menu; a line inside a stage is never an exact copy of a standalone service, so treat this as a fair comparison rather than an identical one.

Deliberately not included

What this stage leaves out, on purpose.

Named here rather than discovered in month three. Each can be added, and each links to what it costs.

A second channel

Foundation proves one. A second before the first has a readable payback period doubles the spend and halves the clarity.

Media Buying — from $900/mo →

Lifecycle and onboarding sequences

Activation is measured here but not yet rebuilt. Improving it is the highest-return work available once acquisition is honest.

Lifecycle Management — $800/mo →

Organic and AI visibility

Search compounds, which is why it is rarely the first move. It belongs once you can afford to wait for it.

SEO Essential — $850/mo →
Where we draw the line

What we will not do, whatever you pay us.

01

We will not optimise to trial starts

Trials are a cost, not a result. Every campaign is judged on trial-to-paid and on payback, because a channel that produces cheap trials and expensive customers is the most common way a SaaS marketing budget is wasted.

02

We will not report a blended number without the cohorts behind it

Blended CAC and blended payback appear in every report next to the monthly cohorts that produced them. If one intake is dragging the average, you see which one in the same week we do.

03

We will not route your ad budget through our account

It goes from your card to the platforms. We never hold it, float it, or take a percentage of it as margin. Management up to $10,000 a month of spend is a flat $900 — scaling your budget does not inflate our invoice.

The first thirty days

Signature to first campaign: 5 to 10 business days.

What happens, and what we need from you at each point. The second column is the one most agencies leave vague.

Days 1–2
What happens

We take access to what already exists in your name and rebuild the last twelve months as cohorts rather than as a monthly line — what each intake cost, and what it has repaid so far.

What we need from you

Approve access requests from your own logins. No shared passwords, ever.

Days 2–4
What happens

Activation is defined in one sentence and instrumented. Not signup, not login — the specific moment your software starts being useful to a new account.

What we need from you

Thirty minutes with whoever knows the product best, to agree that sentence.

Days 4–6
What happens

Billing and product data are joined to the ad accounts so trial-to-paid and payback are read from your own revenue rather than from a platform's conversion event.

What we need from you

Point us at the billing system and the product analytics.

Days 6–10
What happens

First campaigns go live against a payback target agreed in advance, with the rule for ending an experiment written before it starts.

What we need from you

One round of comments, from whoever owns the number.

Day 30
What happens

First full report: CAC and payback by cohort and by channel, activation, trial-to-paid, and what we got wrong — because month two depends on month one being read honestly.

What we need from you

45 minutes, and a decision on what changes.

Terms, in plain language

What you are committing to.

The same terms apply to all three stages and every service on the site.

What software companies ask

The questions that decide it.

Our CAC looks healthy. Why start with cohorts?

Because a healthy blended CAC is compatible with half your intakes never repaying. The average is held up by the cohorts that did well, often ones acquired through a channel you have since scaled down. Splitting the number is usually the first time anyone can see which months actually worked.

It applies exactly as far as activation holds. Product-led growth means the product does the selling, which makes buying traffic above a weak activation step more expensive, not less. We instrument activation first for that reason, and if it does not hold we will tell you to fix it before spending.

No. What we will do is agree a payback target before we spend, report every cohort against it, and tell you in month two whether it is reachable. A guaranteed signup number is a promise to buy the cheapest available trials, which is the opposite of what you want.

Activation and trial-to-paid move in weeks. Payback moves on the schedule your billing sets — if customers repay in nine months, no report can honestly declare success in six. The monthly report separates what has actually resolved from what is still in flight, so you are never asked to take an early number on faith.

Every account, asset and piece of data created during the engagement is yours, including after we part ways. Paid channels stop when the spend stops; the instrumentation, the content and the tracking keep working. That is the arrangement from day one, not a courtesy at the end.

Fit

Who this works for, and who it does not.

A good fit if

Most of this work looks like the following.

The wrong choice if

Said plainly, so nobody spends a call finding out.

Next step

A paid strategy session, credited back.

45 minutes. $95, credited in full toward your first invoice.

Bring twelve months of signups and what they have billed since. We will rebuild them as cohorts and show you which ones have repaid what they cost — including if the honest answer is that most have not.

Tell us how you charge

See which cohorts have paid you back

    We reply within one business day. Your details are used only to arrange this session, never sold or shared.