Why run a growth audit?

Run a growth audit when growth has slowed, several explanations are competing inside the team, and you are about to commit a budget or a six-month plan. The audit tells you which constraint to work on and which projects to drop.

After 11 years running the growth agency Deux.io, and now building Derrick, the first thing I look for is the gap between where the real lever sits and where the team spends its effort. On an audit I ran for a B2B SaaS, traffic already had good bases and conversion had been left untouched, so that is where the analysis went. A standard checklist would have returned a long list of SEO fixes and missed the point.

Throughout this chapter, say you run a B2B SaaS at 2M ARR. Signups are still going up, the board deck looks fine, and net new ARR has been flat for three quarters.

The signs of a growth ceiling

A ceiling usually shows up as more effort for the same result. These are the signals I would check, and how.

Sign	How to check it	What it often hides
Net new MRR flat for several quarters	Split it into new, expansion, contraction and churn	New business that only replaces what leaves
CAC rising on your main channel	CAC per channel and per quarter, salaries included	A saturated channel, or targeting that drifted
Signups up, qualified leads flat	Qualified leads per source, next to signups	Content that attracts the wrong company size
One channel carries most of acquisition	New customers by first source, then by last source	A dependency that last-click reporting makes look healthy
Two teams, two answers	Ask marketing and sales for last quarter’s customers by source	Attribution nobody trusts, so decisions rest on opinion
A lot shipped, nothing attributable	List six months of projects and the metric each one moved	No prioritization and no measurement

One of these deserves arithmetic, because it turns a worry into a number. If you add a steady number of customers each month and lose a steady share of your base, your customer count converges on a limit:

ceiling = new_customers_per_month / monthly_churn_rate

With our illustrative SaaS adding 40 customers a month and losing 2.5% of its base, the ceiling sits at 1,600 customers. At 1,400 customers it loses 35 a month against 40 added, a net gain of 5, and that margin keeps shrinking. Halving churn moves the ceiling to 3,200, exactly what doubling acquisition would do, and the two projects have very different costs.

To see where you stand, split your monthly MRR movements. This PostgreSQL query assumes a mrr_by_account table with one row per account per month, dated on the first of the month:

WITH movements AS (
  SELECT
    COALESCE(cur.month, (prev.month + INTERVAL '1 month')::date) AS month,
    COALESCE(prev.mrr, 0) AS mrr_before,
    COALESCE(cur.mrr, 0) AS mrr_after
  FROM mrr_by_account cur
  FULL JOIN mrr_by_account prev
    ON prev.account_id = cur.account_id
   AND prev.month = (cur.month - INTERVAL '1 month')::date
)
SELECT
  month,
  SUM(mrr_after) FILTER (WHERE mrr_before = 0) AS new_mrr,
  SUM(mrr_after - mrr_before) FILTER (WHERE mrr_before > 0 AND mrr_after > mrr_before) AS expansion,
  SUM(mrr_after - mrr_before) FILTER (WHERE mrr_after > 0 AND mrr_after < mrr_before) AS contraction,
  SUM(mrr_after - mrr_before) FILTER (WHERE mrr_after = 0) AS churn,
  SUM(mrr_after - mrr_before) AS net_new_mrr
FROM movements
WHERE month <= (SELECT MAX(month) FROM mrr_by_account)
GROUP BY month
ORDER BY month;

Ignore the first row, where every account counts as new. If churn and contraction eat most of the new MRR, more acquisition spending will buy you very little.

Why you can’t see it from the inside

Each function reports on its own slice, usually with a model that favors that slice. A last-click report credits the conversion to the last thing that happened, which is often a brand search, so brand looks like your best channel and the origin of the conversion disappears. Content is judged on traffic, sales on deals, product on usage. The constraint usually lives between those dashboards.

A classic case: a handful of pages carry a large share of the traffic and attract the wrong company size, because their topics are too operational for your buyers. The content team sees a win. Sales sees weak leads. Neither sees the link until someone compares who reads with who buys, which is the work of the funnel chapter.

A rising curve can hide the next constraint too. Derrick got 42.4K Google clicks in 12 months from 19.2M impressions, a click-through rate of 0.2%. The growth is real, and the same numbers say the site is seen far more often than it is chosen. The next gains sit in that gap.

When an audit pays off

Four conditions need to hold together:

• A decision is coming: a budget, a hire, a new channel, a fundraising plan.
• You have enough history for the data to mean something.
• Someone can execute over the next six months.
• You are ready to hear a verdict that contradicts the current plan.

Then do the math on your own numbers. Say our SaaS spends 600K a year on acquisition and the people around it. If the audit shows that 15% of this goes to work that brings signups and few qualified leads, 90K a year can move to the real constraint. Compare that with the full cost of the audit, including your team’s time for interviews, access and the review of findings.

What you should walk away with is a clear, actionable roadmap over six months to break through the ceiling, with an owner on each line. The most valuable lines are often the things you stop doing.

When it is too early

An audit optimizes a machine that exists. In these situations, spend the money elsewhere.

Situation	Why an audit won’t help	What to do instead
Retention curves keep falling toward zero	No product-market fit yet, and acquisition would only fill a leaking product	Interview customers, pick one segment, work until a cohort flattens
Cohorts of a dozen customers	One departure moves the cohort by 8 points, so conclusions are noise	Talk to every customer and keep a simple log
The founder still closes every deal	There is no repeatable sale to measure	Write down what wins, then hand deals to someone else
Cause and fix are both known	You would pay to hear what you already know	Ship the fix and measure it
Nobody can execute	The report ages in a folder	Free up capacity first
A pivot is under way	The findings describe a business that is going away	Audit once the new offer has some history

Add one more: an audit commissioned to validate a decision already taken. You will either ignore it or resent it.

Find out in one afternoon

Before you pay anyone, gather this:

1. Net new MRR, split as above, over the last eight quarters.
2. Your ceiling, from new customers per month and monthly churn.
3. Qualified leads per source over four quarters, next to signups.
4. The projects shipped in the last six months and the metric each one moved.

Then ask your founder, your sales lead and your marketing lead to finish this sentence separately: “Our growth is limited by ___, and the evidence is ___.”

If the three answers match and the evidence holds, act on it, with a specialist if needed. If they differ, you have your case for an audit, and the next chapter sets out what one should cover. Send that sentence to those three people today.